POSIX BS2000 file system bs2fs

© Siemens Nixdorf Informationssysteme AG 1995
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.vor
English
BS2000/OSD
POSIX
BS2000 file system bs2fs
User Guide
Valid for
BS2000/OSD V7.0/V8.0/V9.0
Edition November 2012
Comments… Suggestions… Corrections…
The User Documentation Department would like to know your
opinion on this manual. Your feedback helps us to optimize our
documentation to suit your individual needs.
Feel free to send us your comments by e-mail to:
manuals@ts.fujitsu.com
Certified documentation
according to DIN EN ISO 9001:2008
To ensure a consistently high quality standard and
user-friendliness, this documentation was created to
meet the regulations of a quality management system which
complies with the requirements of the standard
DIN EN ISO 9001:2008.
cognitas. Gesellschaft für Technik-Dokumentation mbH
www.cognitas.de
Copyright and Trademarks
Copyright © Fujitsu Technology Solutions GmbH 2012.
All rights reserved.
Delivery subject to availability; right of technical modifications reserved.
All hardware and software names used are trademarks of their respective manufacturers.
This manual is printed
on paper treated with
chlorine-free bleach.
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.ivz
12. March 2014 Stand 09:03.10
© cognitas GmbH 2001-2010
Dokuschablonen 19x24 Version 7.4us für FrameMaker V7.x vom 09.02.2010
Contents
1
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1
Objectives and target groups of this manual . . . . . . . . . . . . . . . . . . . . . 7
1.2
Summary of contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3
Changes since the last edition of the manual . . . . . . . . . . . . . . . . . . . . . 9
1.4
Notational conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2
Overview and embedding in POSIX . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1
Concept and embedding of the bs2fs file system in POSIX . . . . . . . . . . . . 11
2.2
2.2.1
2.2.2
2.2.3
Overview of the use of the bs2fs files system
Making files available in the bs2fs file system . .
Representation of files in the bs2fs file system . .
Interfaces for managing bs2fs file systems . . . .
2.3
Security concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4
2.4.1
2.4.2
Sample sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Introductory example for quick entry into the bs2fs file system . . . . . . . . . . . . 16
Application scenarios for bs2fs file systems . . . . . . . . . . . . . . . . . . . . . . 24
3
Use of bs2fs file systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1
3.1.1
3.1.2
3.1.3
3.1.4
Managing bs2fs file systems . . . . . . . . . . . . . . . . . . .
Creating and mounting the bs2fs container . . . . . . . . . . . .
Mounting bs2fs file systems . . . . . . . . . . . . . . . . . . . .
General aspects relating to mounting and unmounting operations
Automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
U41802-J-Z125-2-76
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
12
12
13
14
27
27
28
29
29
Contents
3.2
3.2.1
3.2.1.1
3.2.1.2
3.2.2
3.2.3
3.2.4
3.2.5
Working with the bs2fs file system . . . .
BS2000 files and library elements supported
BS2000 DMS files . . . . . . . . . . . .
PLAM libraries and elements . . . . . .
bs2fs naming conventions . . . . . . . . .
Access attributes . . . . . . . . . . . . . .
Security . . . . . . . . . . . . . . . . . . .
Various special features of bs2fs . . . . . .
4
Interfaces for supporting bs2fs file systems
. . . . . . . . . . . . . . . . . . . . 47
4.1
4.1.1
Administration interfaces . . . . . . . . . . .
Commands . . . . . . . . . . . . . . . . . . .
mount . . . . . . . . . . . . . . . . . . . . . .
mount a file system . . . . . . . . . . . . .
mountall . . . . . . . . . . . . . . . . . . . . .
mount file systems . . . . . . . . . . . . .
show_pubset_export. . . . . . . . . . . . . . .
show file system affected by pubset export .
start_bs2fsd . . . . . . . . . . . . . . . . . . .
start copy daemons . . . . . . . . . . . . .
umount. . . . . . . . . . . . . . . . . . . . . .
unmount a file system . . . . . . . . . . . .
umountall . . . . . . . . . . . . . . . . . . . .
unmount file systems) . . . . . . . . . . . .
Copy daemon bs2fsd . . . . . . . . . . . . . .
Administration files . . . . . . . . . . . . . . .
/etc/mnttab table of the mounted file systems .
/etc/vfstab table of the defined file systems . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4.1.2
4.1.3
4.2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
31
31
31
34
36
37
39
42
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
47
47
. .
48
. .
55
. .
56
. .
58
. .
59
. .
60
61
62
63
65
Information on shell commands and POSIX tools . . . .
df. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
eport free disk space) . . . . . . . . . . . . . . . . . .
dumpfs . . . . . . . . . . . . . . . . . . . . . . . . . . . .
dump file system . . . . . . . . . . . . . . . . . . . .
fsck . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
file system check . . . . . . . . . . . . . . . . . . . .
fsexpand . . . . . . . . . . . . . . . . . . . . . . . . . .
expand existing file systems . . . . . . . . . . . . . .
pathchk . . . . . . . . . . . . . . . . . . . . . . . . . . .
check pathnames . . . . . . . . . . . . . . . . . . . .
pdbl/posdbl. . . . . . . . . . . . . . . . . . . . . . . . . .
set up and manage user-specific/global program cache
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
67
. r
67
. .
67
. .
67
. .
68
. .
68
. .
68
U41802-J-Z125-2-76
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.5
Information for C program interfaces . . . . . .
errnos . . . . . . . . . . . . . . . . . . . . . . .
fstat, stat – bs2fs file structure . . . . . . . . . . .
fstatvfs, statvfs - file system information . . . . . .
sysfs - query information about the file system type
Further special aspects and restrictions . . . . . .
5
Access to bs2fs file systems via NFS . . . . . . . . . . . . . . . . . . . . . . . . 79
5.1
Releasing bs2fs files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.2
Mounting released bs2fs files . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.3
Security aspects and access rights . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.4
Converting file names
5.5
Further special features when working with bs2fs files on the NFS client . . . . 83
5.6
Recommendations for mounting bs2fs file systems on NFS clients . . . . . . . 84
5.7
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6
Diagnosis and enhancing performance . . . . . . . . . . . . . . . . . . . . . . . 87
6.1
Overview of the mount and unmount operations . . . . . . . . . . . . . . . . . . 87
6.2
Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
get_container_index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
assign bs2fs container index to bs2fs file system . . . . . . . . . . . . . . . . . 90
6.3
Correcting errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
bs2fs_recover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Performing recovery for the bs2fs_lost+found area . . . . . . . . . . . . . . . . 92
6.4
Measures to enhance performance . . . . . . . . . . . . . . . . . . . . . . . . . 97
Dokuschablonen 19x24 Version 7.4us für FrameMaker V7.x vom 09.02.2010
© cognitas GmbH 2001-2010
12. March 2014 Stand 09:03.10
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.ivz
Contents
U41802-J-Z125-2-76
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
70
71
72
74
75
76
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Contents
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k01
12. March 2014 Stand 08:57.50
1 Preface
The BS2000 file system bs2fs permits direct and transparent access to BS2000 files under
POSIX. Consequently both “simple” DMS files and PLAM library elements under POSIX
can be edited as if they were POSIX files.
The BS2000 file system bs2fs is supported in POSIX V7.0 and higher.
1.1 Objectives and target groups of this manual
The manual is intended for all bs2fs file system users. Use of this manual presupposes a
knowledge of the UNIX and BS2000 operating systems and assumes that you have the
manual “POSIX Basics” in your possession.
1.2 Summary of contents
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Chapters 2 and 3 provide an overview of bs2fs file systems, how they are integrated in
BS2000, and how to use bs2fs file systems.
Chapter 4 contains a description of the commands, the daemons and the administration
files.
Chapter 5 describes the accesss to bs2fs file systems via NFS.
Chapter 6 gives advice on troubleshooting and error recovery, along with measures you can
undertake to enhance the performance of bs2fs file systems.
The manual also contains a glossary, a list of related publications and an index.
U41802-J-Z125-2-76
7
Summary of contents
Preface
Readme file
The functional changes to the current product version and revisions to this manual are
described in the product-specific Readme file.
Readme files are available to you online in addition to the product manuals under the
various products at http://manuals.ts.fujitsu.com. You will also find the Readme files on the
Softbook DVD.
Information under BS2000/OSD
When a Readme file exists for a product version, you will find the following file on the
BS2000 system:
SYSRME.<product>.<version>.<lang>
This file contains brief information on the Readme file in English or German (<lang>=E/D).
You can view this information on screen using the /SHOW-FILE command or an editor.
The /SHOW-INSTALLATION-PATH INSTALLATION-UNIT=<product> command shows the
user ID under which the product’s files are stored.
Additional product information
Current information, version and hardware dependencies, and instructions for installing and
using a product version are contained in the associated Release Notice. These Release
Notices are available online at http://manuals.ts.fujitsu.com.
8
U41802-J-Z125-2-76
Changes since the last edition of the manual
1.3 Changes since the last edition of the manual
This edition of the manual contains the following changes with respect to its predecessor
(order number: U41802-J-Z125-1-76):
Support of access to bs2fs files systems via NFS.
Recovery of bs2fs files which were relocated as a result of errors when writing back to the
bs2fs_lost+found directory.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k01
Preface
U41802-J-Z125-2-76
9
Notational conventions
Preface
1.4 Notational conventions
The following notational conventions are used in this manual:
In body text
italics
all syntax elements and also other file names, path names and commands are
shown in italics.
i
This symbol identifies important information and exceptional circumstances which
you should be aware of.
In the syntax
plain type
Variables: These characters are replaced by other characters which you select and
enter.
bold face
Constants: These characters must be entered exactly as they are printed.
[]
Optional: Everything which is enclosed in square brackets can be entered, but does
not have to be. The square brackets themselves should not be entered unless this
is expressly required.
Ë
A blank which must be entered.
...
The preceding expression can be repeated. If blanks need to be entered between
the repeated expressions, and the blank is not contained in the expression, a blank
(Ë) will precede the ellipsis symbol.
|
The vertical line separates alternative specifications.
In examples
bold typewriter face
Inputs: With character-oriented terminals input lines are concluded with the
RETURN key, while with block-oriented terminals the same function is fulfilled by
EM DUE; the key specifications are therefore omitted.
normal typewriter face
Outputs
10
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
12. März 2014 Stand 08:57.50
2 Overview and embedding in POSIX
This chapter provides you with an overview of bs2fs file systems. You will learn how bs2fs
file systems are embedded in POSIX and which security mechanisms you can use to
protect files in bs2fs file systems.
The chapter also contains a collection of sample sessions.
2.1 Concept and embedding of the bs2fs file system in POSIX
The bs2fs file system enables you to access BS2000 files using POSIX interfaces
(commands and program interfaces).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
To do this, the user must specify the set of files with which they wish to work (using BS2000
wildcard syntax) and have these files mounted (by the system administrator) in POSIX as
a bs2fs file system. This mount operation makes these BS2000 files accessible to the user
in POSIX. These files can then be edited in the bs2fs file system using POSIX commands
or from POSIX programs.
To enable these accesses, when the first access takes place in the bs2fs file system (first
open), a background process (daemon) copies the files concerned from BS2000 to a
special ufs fie system in POSIX which was mounted solely for this purpose (bs2fs
container). Only the system may access this file which is stored temporarily in the bs2fs
container. Access by a user take place only to the file mounted below the mount point in the
bs2fs file system. The system redirects this access to the file stored in the bs2fs container.
In the case of write accesses, the file is locked for other users in BS2000, but the bs2fs file
is not locked for other POSIX users. After processing in the bs2fs file system has been
completed, a daemon transfers the file back to BS2000 again. After this has been done, it
can then also be accessed there by other BS2000 users. As long as only information
functions such as ls are executed, no copy function by a bs2fs daemon is initiated. The ls
command merely outputs the files determined in BS2000 using FSTST as POSIX path
names from the bs2fs mount point.
To summarize, use of the bs2fs file system therefore offers the advantage that the user no
longer needs to copy each individual file from BS2000 to the POSIX file system (e.g. with
bs2cp) in order to be able to edit them with POSIX means. The user need only define the
required BS2000 file set and have this mounted by the system administrator. The file set
U41802-J-Z125-2-76
11
Overview of the use of the bs2fs files system
Overview and embedding in POSIX
defined can consist either of files which already exist or ones which are to be created later.
Transfer between BS2000 and POSIX and in the opposite direction is executed invisibly for
the user of copy daemons as soon as a file is opened or when write processing has been
completed.
The use of bs2fs file systems offers, for example, the following options:
●
BS2000 files and PLAM library elements can be searched according to particular
patterns using the POSIX command grep.
●
make can be used to generate programs or program systems efficiently.
●
Nested procedures in which multiple switches between the BS2000 command level and
the shell take place can be replaced by pure POSIX shell scripts if the required BS2000
files are mounted beforehand in a bs2fs file system.
2.2 Overview of the use of the bs2fs files system
The bs2fs file system enables you to make BS2000 files available for editing under POSIX
and to edit these as though they were POSIX files.
2.2.1 Making files available in the bs2fs file system
The bs2fs file system is a hierarchical file system, such as the UNIX file system, which
consists of directories and files.
The system administrator must take the following steps to permit BS2000 files to be edited
in a bs2fs file system:
12
–
Create the bs2fs container
–
Mount the bs2fs container (mount)
–
Mount the bs2fs file system (mount)
U41802-J-Z125-2-76
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Overview of the use of the bs2fs files system
2.2.2 Representation of files in the bs2fs file system
How BS2000 files are represented in a bs2fs file system depends on whether “simple” DMS
files or PLAM libraries are concerned:
–
DMS files are represented as files in POSIX.
–
PLAM libraries are represented as directories with a permanently predefined two-level
hierarchy in POSIX.
The names of these files and directories are identical to the names of the corresponding
DMS files, library elements or PLAM libraries. However, in BS2000 the file names are not
case-sensitive.
The following applies for the notation for file names in the bs2fs file system:
–
File names are always output in lowercase notation.
–
Entry can generally take place in uppercase or lowercase notation or a mixture of both.
When wildcards are used in the shell or in the case of the find command, however, only
lowercase notation may be employed, otherwise no corresponding file names are
found.
–
As customary in the POSIX shell, special characters of the POSIX shell (e.g. ’$’ or ’*’)
must be explicitly escaped during input.
Examples of how to escape BS2000 file name BS2DAT$$1:
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
ls /home/bach/bs2fs1/bs2dat\$\$1
ls /home/bach/bs2fs1/'bs2dat$$1'
U41802-J-Z125-2-76
13
Overview of the use of the bs2fs files system
Overview and embedding in POSIX
2.2.3 Interfaces for managing bs2fs file systems
Commands, daemons and administration files are available for managing bs2fs file
systems.
Commands
The mount, mountall and umount, umountall commands are contained in POSIX-BC. These
commands can also be used for mounting and unmounting bs2fs file systems.
Daemons
Daemons are system processes which run constantly and mainly in the background and
which execute general tasks. The copy daemon bs2fsd transfer files from BS2000 to the
bs2fs container and vice versa. When the bs2fs container is mounted, two bs2fsd daemons
are started automatically.
Administration files
The administration files support the management of resources. They contain either information for the user which is output using commands or information for commands which
either the user or commands have entered in these files. The administration files /etc/mnttab
and /etc/vfstab are used for the application of bs2fs file systems.
14
U41802-J-Z125-2-76
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Security concept
2.3 Security concept
bs2fs is used to make BS2000 objects (DMS files, PLAM libraries, PLAM types, PLAM
libray elements) visible for POSIX interfaces in the form of a file system.
Each POSIX user should be entitled to exactly the same rights as they are entitled to as a
BS2000 user. Each BS2000 object should be provided with precisely the same protection
vis-à-vis POSIX functions as vis-à-vis BS2000 functions.
Only users who are permitted to modify protection attributes with BS2000 means should
also be permitted to do this in a bs2fs file system with POSIX means. Users who are
permitted to read, write or modify a BS2000 object with BS2000 means should also be
permitted to do this in a bs2fs file system with POSIX means. Users who may not do the
former should also not be allowed to do the latter.
In the case of bs2fs file accesses, the POSIX user’s rights may on no account exceed the
rights they have as a BS2000 user.
This concept means that when bs2fs files are accessed with POSIX means, the same
options are not always offered as in the UNIX world.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
An example: write permission is required to delete BS2000 objects with BS2000 means.
Consequently bs2fs files can therefore only be deleted with POSIX means (e.g. rm) if the
write attribute is set. In the UNIX world files can also be deleted without the write attribute.
U41802-J-Z125-2-76
15
Sample sessions
Overview and embedding in POSIX
2.4 Sample sessions
This section contains examples of the application of bs2fs file systems. You are shown how
you can configure and manage bs2fs file systems and how you can access the files. A few
examples of how bs2fs file systems are used are also provided.
Two bs2fs file systems are used in these examples.
2.4.1 Introductory example for quick entry into the bs2fs file system
This section describes how a session with two bs2fs file systems typically runs.
First the system administrator configures the bs2fs container and mounts it. Then the
system administrator mounts two bs2fs file systems for a user ID. Subsequently the user
can work with the files of the bs2fs file systems. Finally the system administrator unmounts
the bs2fs file systems again.
Configuring and mounting the bs2fs container using the POSIX installation program
The append function of the POSIX installation program (Administrate POSIX filesystems)
is used to create a new ufs file system with the bs2fscontainer option and to mount it immediately using Automount=Y. Creating a new ufs file system or overwriting an existing ufs file
system with the bs2fscontainer option causes this file system to be assigned the property
bs2fs container, which is required to permit the ufs file system to be used as a bs2fs
container.
Definition of BS2000 Container File
BS2000 filename:
:V70A:$SYSROOT.FS.BS2FSCONTAINER
BS2000 filesize:
1000000
PAM-Pages
POSIX filesystem? (y/n): Y
A new BS2000 container file has been created
=============================================================================
Definition of POSIX filesystem
Size of filesystem: 100000
PAM-Pages
Journaling? (y/n): N
POSIX mountpoint: /bs2fscont
Automount? (y/n): Y
Mountoptions:
Overwrite existing filesystem? (y/n):
16
bs2fscontainer
POSIX filesystem marker (y/n): Y
U41802-J-Z125-2-76
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Sample sessions
No filesystem in BS2000 container yet
=============================================================================
After the container has been mounted successfsully and the installation program has been
terminated, two bs2fsd copy daemons are started automatically.
# ps -efT|grep bs2fsd
ROOT
163 74MR
ROOT
162 74MQ
1
1
0 15:29:55 ?
0 15:29:55 ?
0:00 [bs2fsd]
0:00 [bs2fsd]
At this time the bs2fs container still has no content:
# ls -l /bs2fscont
Gesamt 0
#
Alternative: mounting the bs2fs container using the mount command
Once a bs2fs container file system has been created with the POSIX installation program
(property bs2fs container, see above), it can also be mounted using the mount command by
specifying the –o bs2fscontainer option, e.g.
# mount –F ufs –o bs2fscontainer /bs2fscont
Two copy daemons are also started automatically in this case.
●
Supplementary information on mounting the bs2fs container
The contents are deleted in all mount operations: mount command, automount in the
case of POSIX startup, automount by the installation program.
The property bs2fs container was introduced as a safety measure to prevent a ufs file
system from being emptied inadvertently when it is mounted as a bs2fs container.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The ufs file system which functions as a bs2fs container is expected to be an empty file
system. If it is not empty, its contents are deleted by the -o bs2fscontainer option when it
is mounted.
U41802-J-Z125-2-76
17
Sample sessions
Overview and embedding in POSIX
Mounting bs2fs file systems
After the bs2fs container has been mounted and the copy daemons have been stated,
bs2fs file systems can be mounted. As with nfs file systems, this is not done using the
POSIX installation program, but with the mount command.
The following files under the user ID BACH and with the catalog ID :V70A: should be made
available in bs2fs file systems:
1. All files whose names match the pattern ASS.*.S should be made available under
/home/bach/bs2.1
2. All files whose names match the pattern PLAMLIB* should be made available under
/home/bach/bs2.2
Two bs2fs file systems are therefore required, and these can be set using the following
commands:
# mount –F bs2fs –o ftyp=text ’:v70a:$bach.ass.*.s’ /home/bach/bs2.1
# mount –F bs2fs –o ftyp=text ’:v70a:$bach.plamlib*’ /home/bach/bs2.2
The mount command automatically extends the /etc/mnttab file by the following entries:
:V70A:$BACH.ASS.*.S
:V70A:$BACH.PLAMLIB*
/home/bach/bs2.1
/home/bach/bs2.2
bs2fs
bs2fs
ftyp=text
ftyp=text
1196152657
1196152666
The bs2fs container now contains directories (which are still empty) which are to accommodate copies of the two bs2fs file systems’ BS2000 files. Pure information functions such
as the ls command do not initiate copying. Only when an open() is applied to a bs2fs file
does the bs2fsd daemon become active. Details on the structure of the bs2fs container are
provided in section “Overview of the mount and unmount operations” on page 87.
# ls -l /bs2fscont
Gesamt 16
drwx--x--x
2 SYSROOT
drwx--x--x
2 SYSROOT
#
SYSROOT
SYSROOT
2048 Nov 27 11:40 V70A.BACH.1
2048 Nov 27 11:41 V70A.BACH.2
The protection bit attributes of the bs2fs container are automatically set by the mount
operation in such a manner that the container’s contents are not visible to nonprivileged
users (drwx--x--x).
18
U41802-J-Z125-2-76
Sample sessions
Working with bs2fs file systems
After the two bs2fs file systems have been mounted, the result of the mount operations can
be checked:
Ê
Display the content of the file system mounted under /home/bach/bs2.1:
# ls -l /home/bach/bs2.1
Gesamt 3620
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
.
.
.
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
-rwx-----1 BACH
OS315
#
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
12288
18432
6144
18432
2048
18432
139264
Nov
Nov
Nov
Nov
Nov
Nov
Nov
22
22
22
22
22
22
22
14:54
14:54
14:54
14:55
14:55
14:55
14:56
ass.consio.s
ass.lock.s
ass.lockadm.s
ass.lockwrk.s
ass.posasto.s
ass.posbs2fs.s
ass.posbs2fx.s
69632
51200
49152
2048
18432
32768
2048
Nov
Nov
Nov
Nov
Nov
Nov
Nov
22
22
22
22
22
22
22
14:58
14:59
14:59
14:59
15:00
15:00
15:00
ass.posutil.s
ass.posvert.s
ass.posvmm.s
ass.pprot.s
ass.slpwkp.s
ass.ttrap.s
ass.uaddr.s
So a number of simple BS2000 files is mounted under /home/bach/bs2.1.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
●
Supplementary information on the representation of the bs2fs files
Protection bits
map either the BS2000 standard attributes (USER-ACCESS, ACCESS) or the
BACL properties. In our case, for example, the standard attributes USERACCESS=*USER-ONLY and ACCESS=*WRITE.
Owner (USER)
is always the ID of the bs2fs mount
Group (GROUP)
is always the POSIX group of the owner ID. However, this group assignment is not
relevant for bs2fs accesses. The BS2000 group assignments with / without SECOS
apply. See chapter 3 for details.
File size
of unopened files is a multiple of 2048 occupied PAM pages. The smallest file
according to the ls command is therefore 2048 bytes in size.
U41802-J-Z125-2-76
19
Sample sessions
Overview and embedding in POSIX
File name:
Names are displayed in lowercase notation. This also applies for library and
element names. The reference to bs2fs objects with shell commands or C program
interfaces, on the other hand, can be in either uppercase or lowercase. Exception:
lowercase notation is required in wildcard constructs in the shell and with the find
command.
Ê
Now display the content of the second file system mounted under /home/bach/bs2.2:
# ls -l /home/bach/bs2.2
Gesamt 5224
drwx-----2 BACH
OS315
drwxrwxrwx
2 BACH
OS315
1562624 Nov 19 17:39 plamlib.1
1112064 Nov 19 17:30 plamlib.2
This file system consists of two PLAM libraries, which can be recognized from the representation as a directory.
Ê
Display the content of the library plamlib.1:
# ls -l /home/bach/bs2.2/plamlib.1
Gesamt 32
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
drwx-----2 BACH
OS315
48
48
48
48
48
4032
48
Nov
Nov
Nov
Nov
Nov
Nov
Nov
19
19
19
19
19
19
19
17:39
17:39
17:39
17:39
17:39
17:39
17:39
d
j
l
m
p
s
x
The library directory contains subdirectories whose names correspond to the element
types (D, J, L, M, P, S, X).
●
Supplementary information on the library and type directories
The directory structure and the directory attributes cannot be modified. Actions such as
deleting, renaming, creating and modifying attributes are always rejected when they
apply to library and type directories.
Consequently, for example, no new PLAM library can be created using mkdir.
20
U41802-J-Z125-2-76
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Ê
Sample sessions
Now display the type S elements:
# ls -l /home/bach/bs2.2/plamlib.1/s
Gesamt 6024
-rwx-----2 BACH
OS315
8192
-rwx-----2 BACH
OS315
8192
-rwx-----2 BACH
OS315
51200
-rwx-----2 BACH
OS315
51200
-rwx-----2 BACH
OS315
8192
-rwx-----2 BACH
OS315
8192
-rwx-----2 BACH
OS315
4096
-rwx-----2 BACH
OS315
4096
-rwx-----2 BACH
OS315
12288
-rwx-----2 BACH
OS315
12288
.
.
.
-rwx-----2 BACH
OS315
14336
-rwx-----2 BACH
OS315
14336
-rwx-----2 BACH
OS315
18432
-rwx-----2 BACH
OS315
18432
-rwx-----2 BACH
OS315
6144
-rwx-----2 BACH
OS315
6144
-rwx-----2 BACH
OS315
4096
-rwx-----2 BACH
OS315
4096
-rwx-----2 BACH
OS315
8192
-rwx-----2 BACH
OS315
8192
#
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
19
19
19
19
19
19
19
19
19
19
17:33
17:33
17:33
17:33
17:33
17:33
17:33
17:33
17:34
17:34
acct.c
acct.c+001
bio.c
bio.c+001
bitmap.c
bitmap.c+001
bitmasks.c
bitmasks.c+001
bs.c
bs.c+001
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
19
19
19
19
19
19
19
19
19
19
17:39
17:39
17:39
17:39
17:39
17:39
17:39
17:39
17:39
17:39
vm_pageout.c
vm_pageout.c+001
vm_pgout.c
vm_pgout.c+001
vm_subr.c
vm_subr.c+001
xmmu.c
xmmu.c+001
xsys.c
xsys.c+001
The library directory contains some type S elements. Each of these elements occurs in
the example with just one version, '001'. As a result, two entries exist for each element:
one entry for the element name without a version identifier and one with the (only)
version identifier.
Explanation:
In the bs2fs file system the element with the highest version can also be addressed as
an element name without a version (implemented internally as a hard link). The representation and the reference to elements with lower versions, on the other hand, always
include the version. See also chapter 3.
U41802-J-Z125-2-76
21
Sample sessions
Ê
Overview and embedding in POSIX
Now you can, for example, apply POSIX commands to the files of the bs2fs file
systems:
# grep ETPND /home/bach/bs2.1/*.s
/home/bach/bs2.1/ass.consio.s:
ETPND
/home/bach/bs2.1/ass.lock.s:
TITLE
/home/bach/bs2.1/ass.lock.s:
ETPND
&MODNAME,VER=&MODVERS,PATCH=200
'&MODNAME - ETPND/XREF'
&MODNAME,VER=&MODVERS,PATCH=200
# tail /home/bach/bs2.1/ass.posvert.s
hdrcheck unit=228,funct=(187,1)
@bend
@bend
@exit
@end
eject
******************************* etpnd *****
etpnd &modname,ver=&modvers,patch=200
*
end
#
Unmounting the bs2fs file systems
If the files in the bs2fs file systems are no longer required in POSIX, the file systems
concerned can be unmounted again.
Use the umount command to unmount individual bs2fs file systems, thus in this example:
umount /home/bach/bs2.1
and
umount /home/bach/bs2.2
You can use the umountall command to unmount all bs2fs file systems:
– either
umountall –F bs2fs
–
or
umountall –b
After you have unmounted the file systems using umount or umountall, the corresponding
entries in the /etc/mnttab will also have been deleted.
22
U41802-J-Z125-2-76
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Sample sessions
Automating the bs2fs mount operations
The following options, among others, are available to automate bs2fs mount operations:
1. Enter the mount commands in shell scripts and execute these scripts at any time.
2. Create entries in the file system table /etc/vfstab using Automount=Y.
The bs2fs mount operations are then executed automatically when the POSIX
subsystem is started. Furthermore, the /etc/vfstab table is read out by the mountall
command if no other file system table is specified.
You can use the mountall command to mount all the bs2fs file systems on the basis of
the entries in the /etc/vfstab table:
– either
mountall –F bs2fs
–
or
mountall –b
3. Create entries in a separate file system table (similarly to /etc/vfstab). This table can then
be used as an argument of the mountall command.
Example: create an entry in a mountall table (/etc/vfstab would be similar)
Ê
You can use the mount -p command to create a file which already contains a suitable
line in vfstab format:
# mount -p |grep /home/bach/bs2.1 >> /etc/bs2fstab
# cat /etc/bs2fstab
:V70A:$BACH.ASS.*.S - /bs2test/bs2fs bs2fs - no ftyp=text
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Ê
You must change the automount entry from no to yes to permit automatic mounting
with the mountall command, e.g. using edt /etc/bs2fstab
:V70A:$BACH.ASS.*.S - /bs2test/bs2fs bs2fs – yes
ftyp=text
You can use the mountall command to mount all bs2fs file systems on the basis of the
entries in the /etc/bs2fstab table:
– either
mountall –F bs2fs /etc/bs2fstab
–
or
mountall –b /etc/bs2fstab
U41802-J-Z125-2-76
23
Sample sessions
Overview and embedding in POSIX
2.4.2 Application scenarios for bs2fs file systems
Various application options for bs2fs file systems are presented in this section.
Example 1: searching CONSLOG files
The stored CONSLOG files for the month of June 2007 are to be searched for particular
contents. This search can be executed very simply using the grep command if the files are
available in a bs2fs file system.
This example requires that the bs2fs container must already have been created.
Proceed as follows:
Ê
Mount the BS2000 files which are to be processed:
# mount -F bs2fs ':V70a:$sysaudit.sys.conslog.2007-06*'
Ê
/home/bs2.conslog
Check the result of the mount operation:
# mount
/ on /dev/root read/write/setuid on Tue Nov 27 11:31:04 2007
.
.
.
/bs2fscont on /dev/dsk/23 bs2fscontainer/setuid/read/write/noquota on Tue Nov 27
11:35:23 2007
/home/bs2.conslog on :V70A:$SYSAUDIT.SYS.CONSLOG.2007-06* ftyp=text/nosuid on Tue Nov
27 13:52:23 2007
#
or
# df -k -F bs2fs
.
.
.
:V70A:$SYSAUDIT.SYS.CONSLOG.2007-06*
/home/bs2.conslog
Ê
2000000
331518
1668482
17%
Display the BS2000 files made available in the bs2fs file system:
# ls -l /home/bs2.conslog
Gesamt 11836
-r-x------r-x------r-x------r-x------r-x------r-x------r-x------
1
1
1
1
1
1
1
100
100
100
100
100
100
100
OTHER
OTHER
OTHER
OTHER
OTHER
OTHER
OTHER
151552
256000
75776
73728
77824
77824
5347328
Jun
Jun
Jun
Jun
Jun
Jun
Nov
12
13
13
13
14
14
14
12:32
13:17
16:26
17:25
12:36
14:42
11:12
sys.conslog.2007-06-11.007.001
sys.conslog.2007-06-12.007.001
sys.conslog.2007-06-13.007.001
sys.conslog.2007-06-13.007.002
sys.conslog.2007-06-13.007.003
sys.conslog.2007-06-14.007.001
sys.conslog.2007-06-14.007.002
#
24
U41802-J-Z125-2-76
12. März 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k02
Overview and embedding in POSIX
Ê
Sample sessions
In the files provided, search for file names with the prefix :V70A:$BACH.SEM:
# grep ':V70A:$BACH.SEM' /home/bs2.conslog/*
.
.
.
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MBW-000.163259 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MB0-000.163310 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MB0-000.163310 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MCH-000.170925 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MCH-000.170925 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MCL-000.170939 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MCL-000.170939 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MER-000.144635 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MER-000.144635 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0ME1-000.165231 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0ME1-000.165231 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MKI-000.133902 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MKI-000.133902 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MNC-000.130039 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: analyseresource after toupper <:V70A:$BACH.SEM*.C>
/home/bs2.conslog/sys.conslog.2007-06-14.007.002: <C %0MNC-000.130039 % POS1020 Message
the POSIX kernel:WARNING: bs2_subr.c: couldbebs2fsname <:V70A:$BACH.SEM*.C>
of
of
of
of
of
of
of
of
of
of
of
of
of
of
of
#
Example 2: converting lowercase notation to uppercase notation
In this example it is assumed that the file to be edited is made available under
/home/bs2.conslog, as in “Example 1: searching CONSLOG files” on page 24.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The lowercase notation in a BS2000 file is to be converted to uppercase notation. You can
use the tr command for this conversion if the files are contained in a bs2fs file system.
U41802-J-Z125-2-76
25
Sample sessions
Ê
Overview and embedding in POSIX
Display the BS2000 files provided in the bs2fs file system:
# ls -l /home/bs2.conslog
Gesamt 11836
-r-x------r-x------r-x------r-x------r-x------r-x------r-x------
1
1
1
1
1
1
1
100
100
100
100
100
100
100
OTHER
OTHER
OTHER
OTHER
OTHER
OTHER
OTHER
151552
256000
75776
73728
77824
77824
5347328
Jun
Jun
Jun
Jun
Jun
Jun
Nov
12
13
13
13
14
14
14
12:32
13:17
16:26
17:25
12:36
14:42
11:12
sys.conslog.2007-06-11.007.001
sys.conslog.2007-06-12.007.001
sys.conslog.2007-06-13.007.001
sys.conslog.2007-06-13.007.002
sys.conslog.2007-06-13.007.003
sys.conslog.2007-06-14.007.001
sys.conslog.2007-06-14.007.002
#
Ê
Convert the lowercase notation in the sys.conslog.2007-06-13.007.001 file to uppercase
notation and write the result to the conslog.out file. (The name of the input file is specified
unambiguously in abbreviated form using wildcards.)
# tr '[:lower:]' '[:upper:]' </home/bs2.conslog/*13.007.001 >conslog.out
Ê
Check the result by displaying the last records of the input and output files:
# tail
/home/bs2.conslog/*13.007.001
<C %0CYT-000.162451 % POS1020 Message of the POSIX kernel:WARNING: vfs.c: generic mount 1
<C %0CYT-000.162451 % POS1020 Message of the POSIX kernel:WARNING: vfs.c: generic mount 2
<C %0CYT-000.162451 % POS1020 Message of the POSIX kernel:WARNING: vfs.c: generic mount 3
<C %0CYT-000.162451 % POS1020 Message of the POSIX kernel:WARNING: bs2_vfsops.c: entering
bs2fs_mount
<C %0CYT-000.162451 % POS1020 Message of the POSIX kernel:WARNING: bs2_subr.c:
couldbebs2fsname
<0 %0CYT-000.162451 % EXC0420 /LOGOFF PROCESSED. CPU TIME USED: 2.8544 SEC, USER ID: TSOS,
TASK ID: 1DDF00D6
<0 %0CYU-000.162528 % JMS0154 'TSOS' LOGGED ON FOR 'MCP0212C/STATION9'. JOB NAME
'BACHMANN'. CALLER 'TSN 0BEP'. TID 1C7E0082
<0 %0CYU-000.162529 % EXC0420 /LOGOFF PROCESSED. CPU TIME USED: 0.3710 SEC, USER ID: TSOS,
TASK ID: 1C7E0082
/0B1Q-000.162609 CHANGE-CONSLOG PROCESSOR NAME: MCP0212C STATION NAME: STATIO10 AUDIT-ID:
00000000000000000000000000000000
TCLOG
.162609 ***2007-06-13*** 000004 **** UTC+02:00
****************************************************************
# tail conslog.out
<C %0CYT-000.162451 % POS1020 MESSAGE OF THE POSIX KERNEL:WARNING: VFS.C: GENERIC MOUNT 1
<C %0CYT-000.162451 % POS1020 MESSAGE OF THE POSIX KERNEL:WARNING: VFS.C: GENERIC MOUNT 2
<C %0CYT-000.162451 % POS1020 MESSAGE OF THE POSIX KERNEL:WARNING: VFS.C: GENERIC MOUNT 3
<C %0CYT-000.162451 % POS1020 MESSAGE OF THE POSIX KERNEL:WARNING: BS2_VFSOPS.C: ENTERING
BS2FS_MOUNT
<C %0CYT-000.162451 % POS1020 MESSAGE OF THE POSIX KERNEL:WARNING: BS2_SUBR.C:
COULDBEBS2FSNAME
<0 %0CYT-000.162451 % EXC0420 /LOGOFF PROCESSED. CPU TIME USED: 2.8544 SEC, USER ID: TSOS,
TASK ID: 1DDF00D6
<0 %0CYU-000.162528 % JMS0154 'TSOS' LOGGED ON FOR 'MCP0212C/STATION9'. JOB NAME
'BACHMANN'. CALLER 'TSN 0BEP'. TID 1C7E0082
<0 %0CYU-000.162529 % EXC0420 /LOGOFF PROCESSED. CPU TIME USED: 0.3710 SEC, USER ID: TSOS,
TASK ID: 1C7E0082
/0B1Q-000.162609 CHANGE-CONSLOG PROCESSOR NAME: MCP0212C STATION NAME: STATIO10 AUDIT-ID:
00000000000000000000000000000000
TCLOG
.162609 ***2007-06-13*** 000004 **** UTC+02:00
****************************************************************
#
26
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
12. March 2014 Stand 08:57.50
3 Use of bs2fs file systems
This chapter shows you how to manage bs2fs file systems and what you need to take into
consideration when working with bs2fs file systems.
3.1 Managing bs2fs file systems
Managing bs2fs file systems involves the following tasks:
●
Creating the bs2fs container with the POSIX installation program
●
Mounting and unmounting the bs2fs container
●
Mounting and unmounting bs2fs file systems
●
Modifying administration files if the lists of the resources which need to be provided or
mounted need to be updated
●
Locating and correcting problems when bs2fs file systems are used (see chapter
“Diagnosis and enhancing performance” on page 87)
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
3.1.1 Creating and mounting the bs2fs container
To permit the use of bs2fs file systems, a special area is required in POSIX which temporarily stores the files of bs2fs file systems: the bs2fs container. This is a ufs file system
which is explicitly identified as the bs2fs container. Only one single bs2fs container can
exist in a POSIX system. This bs2fs container accommodates all the files of all mounted
bs2fs file systems.
You can create a file system with the property bs2fs container using the POSIX installation
program when installation takes place.
Its size should be defined in accordance with the size of the files actually used. As it can be
assumed that not all BS2000 files will be opened at the same time and therefore copied into
the container, you could, for example, estimate the space required in the container as
follows: multiply the sum of the sizes of all BS2000 files which are mounted with bs2fs by
0.5.
U41802-J-Z125-2-76
27
Managing bs2fs file systems
Use of bs2fs file systems
Creating the bs2fs container using the POSIX installation program
When append is specified for a new ufs file system or an existing file system (with
overwrite=yes) together with the bs2fscontainer option, this file system is internally
assigned the property bs2fs container which is required to enable this file system to be
used as the bs2fs container.
Mounting the bs2fs container using the POSIX installation program
Like any other ufs file system, the bs2fs container (= ufs file system with the property
bs2fscontainer) can be mounted in the file system hierarchy when Automount=Y is specified.
The Automount=Y entry which is then created in the file system table /etc/vfstab also causes
the file system to be mounted automatically when the POSIX subsystem is started.
Mounting the bs2fs container during ongoing operation using the mount command
To mount the bs2fs container, enter a corresponding mount command with the
-o bs2fscontainer option, e.g.
mount -F ufs -o bs2fscontainer /dev/dsk/23 /home/bach/mount2
A prerequisite for mounting a file system as the bs2fs container is that it has been created
as described under “Creating the bs2fs container using the POSIX installation program” on
page 28, otherwise a mount command with the -o bs2fscontainer option is rejected.
A bs2fs container is by default empty when it is mounted. If it is not empty, its contents are
deleted when it is mounted.
3.1.2 Mounting bs2fs file systems
When you mount bs2fs file systems, you define which BS2000 files are made available at
which position in the POSIX file system hierarchy.
You can only mount a bs2fs file system after the bs2fs container has been mounted.
To mount a bs2fs file system, enter a corresponding mount command with the -F bs2fs
option, e.g.
mount -F bs2fs ’:v70a:$bach.ass.*.s’ /home/bach/bs2.1
28
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Managing bs2fs file systems
3.1.3 General aspects relating to mounting and unmounting operations
Both bs2fs file systems and the bs2fs container can be mounted and unmounted during
ongoing operation using the mount and umount commands respectively. However, the order
specified below must be observed here. Command entries which deviate from this order
result in errors and cause the command concerned to abort:
Order for mounting and unmounting operations
●
Mounting the bs2fs container
The ufs file system with the property bs2fs container must be mounted before the first
bs2fs file system can be mounted.
●
Mounting or unmounting bs2fs file systems
bs2fs file systems can be mounted and unmounted again while the bs2fs container is
mounted.
●
Unmounting the bs2fs container
The bs2fs file system with the property bs2fs container can be unmounted only when
no bs2fs file system is mounted.
3.1.4 Automation
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
bs2fs container
The mount/umount operations are managed and thus also automated in the same way as
other POSIX file systems of the type "ufs". When the file system is created using the POSIX
installation program (append function), an entry is automatically generated in the /etc/vfstab
table. If Automount=yes was specified, the file system is mounted immediately. The
Automount=yes entry in the /etc/vfstab table also causes the file system to be mounted
automatically the next time the POSIX subsystem starts. Modifications in the /etc/vfstab
table can also be made using the POSIX installation program (modify function).
bs2fs file systems
Similarly to nfs file systems, bs2fs file systems cannot be managed using the POSIX installation program. To automate mount operations, the administrator must edit the system
tables or create scripts.
U41802-J-Z125-2-76
29
Managing bs2fs file systems
Use of bs2fs file systems
Overview of the various automation options:
1. Enter mount commands in shell scripts and execute these at any time.
2. Enter mount commands in shell scripts of the type "rc". These will then be executed
automatically when the POSIX subsystem is started.
3. Enter the required bs2fs file systems in vfstab format (with Automount=yes) in a
separate file system table. This table can then be used as an argument in the mountall
command.
4. Enter the required bs2fs file systems (with Automount=yes) in the /etc/vfstab table. This
table is processed by the mountall command if no file system table is specified. When
the POSIX subsystem starts, the mount operations will be executed automatically.
Recommendation: entries suitable for vfstab can be created with mount –p for a bs2fs file
system which has already been successfully mounted.
An example of this is provided on page 23.
Starting and terminating POSIX
If the mount/umount operations take place automatically when POSIX is started and terminated, this has the advantage that the necessary order is observed.
Startup: ufs file systems, bs2fs file systems, nfs file systems
Termination: nfs file systems, bs2fs file systems, ufs file systems
If the administrator wants to mount and unmount the file systems explicitly, they must
observe this order.
30
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Working with the bs2fs file system
3.2 Working with the bs2fs file system
This section describes what types of files and library elements you can edit in a bs2fs file
system, which access rights apply in a bs2fs file system, and what special aspects you must
bear in mind when working with bs2fs file systems.
3.2.1 BS2000 files and library elements supported
In bs2fs file systems you can edit “simple” DMS files and PLAM libraries and their elements.
3.2.1.1
BS2000 DMS files
When BS2000 DMS files are supported in a bs2fs file system, a distinction must be made
between whether a new file is created or whether it already exists in BS2000.
Newly created DMS files
●
FCBTYPE
New DMS files created in a bs2fs file system are always of the type FCBTYPE SAM
with a variable record length.
●
BACL (see also section “Access attributes” on page 37)
The BACL attributes are automatically enabled in BS2000 in accordance with the mode
specified in open() and the umask setting (mode & ~umask).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
For example, a umask setting of 022 (default in the shell) and a mode pattern of 777
result in the protection bit rights rwx r-x r-x.
When BACL is enabled, modifications are no longer evaluated with /MODIFY-FILEATTRIBUTES in USER-ACCESS or ACCESS.
●
Transfer mode ftyp
Whether a SAM file is transferred to the bs2fs container and back in text, textbin or binary
mode depends on the ftyp option of the mount command for the bs2fs file system. The
POSIX command ftyp is not evaluated and is reserved for bs2cp. See also mount
command, -o option ftyp.
U41802-J-Z125-2-76
31
Working with the bs2fs file system
Use of bs2fs file systems
Existing DMS files
●
FCBTYPE
Existing DMS files can be of the FCBTYPE SAM, ISAM or PAM. The FCBTYPE is also
retained if such a file is overwritten.
The following is meant by overwriting: creat(), open (O_CREAT |O_WRONLY |O_TRUNC),
cp command, but not the rename() or mv command.
No FCBTYPE (no open) is yet defined for existing DMS files which were created only
by means of /CREATE-FILE or something similar. After being opened as bs2fs files,
these files are assigned the FCBTYPE SAM.
●
BACL (see also section “Access attributes” on page 37)
When write or overwrite accesses take place, all the original protection attributes are
retained. If the BACL attributes had not been enabled beforehand, they are also not
enabled later. The mode specifications in open() or creat() are, as is customary in POSIX,
ignored in the case of existing files.
In the bs2fs file system the BACL properties of existing files without BACL can only be
enabled by explicitly modifying the protection attributes with chmod() or the chmod
command.
●
Transfer mode ftyp
ISAM files are always transferred in text mode, PAM files always in binary mode. In the
case of SAM files - as with newly created SAM files - the ftyp option of the mount
command is relevant for the bs2fs file system.
i
Take care when copying within the bs2fs world
If the file attributes of the source and the target are different with respect to the
transfer mode (text, binary), this can lead to unexpected (data) results.
There is no analogy to the /COPY-FILE command, where the target file is
assigned the same FCBTYPE as the source file, when copying using POSIX
bs2fs means.
32
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Working with the bs2fs file system
BS2000 file types which are not supported
The following files/libraries are not displayed in a bs2fs file system under POSIX, in other
words, for example, they are not output by the ls command:
–
–
–
–
–
–
Exported (migrated) files/libraries (please also observe the notes on handling migrated
files in section “Various special features of bs2fs” on page 42)
Alias catalog entries
File generation groups and file generations
Files/libraries which are accessed via RFA
Tape files
Files on private disks
Files protected by a password are visible in the bs2fs file system under POSIX, but their
contents can only be accessed by operations which are not protected by a password, e.g.
read access to a file protected by a write password.
Files protected by guards are visible in the bs2fs file system under POSIX. Their contents
can only be accessed if the protecting guard permits corresponding access for the caller’s
user ID.
Overview of the open modes of DMS files
DMS files in POSIX are interpreted differently according to their FCBTYPE, the record
length and the ftyp option when the bs2fs file system is mounted (see page 52). An overview
of how these properties are mapped to the file type and the corresponding open modes in
the bs2fs file system is provided by the table below:
FCBTYPE ftyp
Open mode
for reading
Open mode for writing
text (default) Text file
r
w,tabexp=yes
binary
Binary file
rb,type=record wb,type=record,forg=seq
textbin
Binary text file
r
w,tabexp=no
ISAM
-
Text file
r
w,tabexp=yes
PAM
-
Binary file
r
wb,type=record
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
SAM
File type
The record keys of ISAM files are ignored when reading; when writing, numeric record keys
(in character representation) are generated.
U41802-J-Z125-2-76
33
Working with the bs2fs file system
3.2.1.2
Use of bs2fs file systems
PLAM libraries and elements
Element types supported
The bs2fs file system supports the same element types as the bs2cp command (and /COPYPOSIX-FILE), namely the predefined text types (D, J, M, P, S, X) and LLMs (L). Furthermore,
types defined by the user (e.g. with openFT) and types derived from the standard types
mentioned above are also supported in bs2fs file systems.
Like DMS files of the type SAM, the text, textbin or binary mode in which the text elements
(all types other than LLM) are copied to and from the bs2fs container depends on the ftyp
option of the mount command for the bs2fs file system.
Representation of PLAM libraries and library elements
A PLAM library is represented as a directory with the name of the PLAM library. By default
this library directory contains the subdirectories of the standard types D, J, L, M, P, S and X
in which the corresponding element types are stored.
in addition, the library directory contains further directories with the names of derived
element types, but only if at least one element of the relevant type exists and the derived
type has one of the standard types D, J, L, M, P, S or X as its basic type.
The standard types C, F, H, R, U and SYSJ and types derived from these are not supported.
This also applies for the reserved types whose names begin with '$' or 'SYS'.
If the PLAM library contains user-defined types which do not have a standard type as their
basic type, these are represented as corresponding subdirectories. Such types can, for
instance, be created with openFT. New PLAM libraries cannot be created with bs2fs means
(e.g. with mkdir). New PLAM libraries which were created with other means (e.g. with LMS)
after a bs2fs mount can, however, be edited at any time in the bs2fs file system. To permit
this, it is only necessary for their name to match the wildcard pattern of the bs2fs mount
command.
Library elements are stored as files in the directory whose name corresponds to their type.
Their names have the following format: elementname+version. This name permits targeted
access to a particular version of an element. The name elementname permits access to the
highest version of an element. A hard link to the highest version of an element is involved
here.
34
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Working with the bs2fs file system
The following special features apply for this hard link:
–
If a new element version with a higher version identifier is generated in the bs2fs file
system, elementname automatically refers to the new highest version.
–
Deleting elementname (e.g. with rm elementname) has the following effect:
First the file containing the highest version of the element is deleted. If other versions
of the element exist, elementname refers to the new highest version. Otherwise the hard
link elementname is also deleted.
–
If a particular version of an element is deleted (e.g. with rm elementname+ver), the hard
link elementname is also deleted if no other versions of the element exist.
–
All versions of an element can be deleted with rm elementname+*.
Example of the structure of a PLAM library
input
input+@
project-library
x
d
testfile
testfile+001
j
s
l
compile
p
m
testfile+002
testfile+003
compile+001
compile+002
© cognitas GmbH 2001-2007
compile+003
read
write+002
write+001
read+001
read+002 write
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
run
run+@
U41802-J-Z125-2-76
35
Working with the bs2fs file system
Use of bs2fs file systems
3.2.2 bs2fs naming conventions
When a new BS2000 DMS file or a library element is created with POSIX means (open
O_CREAT, creat, rename), the name is specified as a POSIX file name, but it is subject to the
rules for creating DMS file names, element names (name length, valid characters) and to
further bs2fs-specific rules.
Uppercase/lowercase notation
The names can be specified in uppercase or lowercase or a mixture of both. However, the
file names are always output in lowercase notation. When wildcards are used in the shell
or in the find command, though, lowercase notation must be used, otherwise no corresponding file names are found. If uppercase notation were used in the file name pattern, file
name replacement in the shell would fail as the pattern is transferred with wildcard syntax.
Uppercase/lowercase notation is possible for all objects below the bs2fs mount point, i.e in
the case of PLAM libraries also for the library directories and the type directories
Wildcard patterns in the mount command for the bs2fs file system
When a new DMS file is created, the file name is checked to see whether it is contained in
the range of the BS2000 wildcard pattern which was specified for the bs2fs file system in
the mount command. If it is not, the open is rejected with errno ENAME. This check also takes
place when a file (e.g. for a reading open, access, stat, etc.) or PLAM library is searched for;
a non-match is rejected with errno ENAME.
36
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Working with the bs2fs file system
3.2.3 Access attributes
Mapping the BS2000 standard attributes
The table below shows how the BS2000 standard attributes are mapped to the access
rights in POSIX:
BS2000 attributes
POSIX standard ACL
user
group
others
ACCESS
USER-ACCESS
r
w
x
r
w
x
r
w
x
*READ
*OWNER-ONLY
x
-
x
-
-
-
-
-
-
*READ
*ALL-USERS/*SPECIAL
x
-
x
x
-
x
x
-
x
*WRITE
*OWNER-ONLY
x
x
x
-
-
-
-
-
-
*WRITE
*ALL-USERS/*SPECIAL
x
x
x
x
x
x
x
x
x
The rights defined with ACCESS and USER-ACCESS are mapped to POSIX access rights
for a file only if no Basic Access Control List (BACL) is defined for it.
The following attributes of files and library elements are not mapped to the bs2fs file system:
●
EXEC-PASSWORD, READ-PASSWORD, WRITE-PASSWORD (file or library element)
EXPIRATION-DATE (file)
GUARD (file or library element)
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
If one of these protection attributes is set for a file or an element, an access attempt can
be rejected with EACCES even if the status information supplied seems to permit
access.
●
CODED-CHARACTER-SET (file or library element)
If this attribute is not *NONE, *STD or EDF03IRV, the result of processing in a bs2fs file
system can differ from the result of the same processing in BS2000 (for example with
the same C program): e.g. different collating sequence, different representation on the
screen, etc.
●
STATE=*IN-HOLD (library element)
If this attribute is set, an access attempt is rejected with EACCES, even though the
status information supplied does not enable you to recognize this.
U41802-J-Z125-2-76
37
Working with the bs2fs file system
●
Use of bs2fs file systems
ACCESS-METHOD (library element)
In the case of elements with SOURCE-ATTRIBUTE=*KEEP, this attribute can be
adopted by a corresponding file. However, it is ignored in the bs2fs file system. If such
an element is copied or moved to a file, processing this file can yield different results
from processing the original element.
●
ISAM key
ISAM keys from BS2000 files are not visible when access takes place via a bs2fs file
system. If an ISAM file in a bs2fs file system is copied to another existing ISAM file
which is overwritten in the process, the records of the two files then generally have
different ISAM keys.
●
ISAM key (library element)
ISAM keys from an ISAM file which was transferred to a library element using
SOURCE-ATTRIBUTE=*KEEP are a component part of the records concerned.
●
STORAGE-FORM=*DELTA (library element)
Library elements with this attribute (delta elements) can only be read in a bs2fs file
system.
●
Record format B (library element)
Library elements with records with format B can only be edited if they are of type L or
of a type derived from this type. Attempts to access elements of a different type are
rejected with EMODE.
●
Record type other than 1 (library element)
Record types with format A and a record type other than 1 are not visible when reading
takes place.
In the case of an update (O_RDWR or O_APPEND) and when renaming within the
same library and the same type or between types with the same basic type, they are
retained provided the source is not open in write mode when renaming takes place.
They are not transferred when a library element is copied or moved into a file or another
library or when it is renamed as a type with a different basic type or when a source which
is open in write mode is renamed.
38
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
Working with the bs2fs file system
BACL attributes
When new bs2fs files and library elements are created and when attributes are modified
with chmod, the BACL attributes in BS2000 are enabled. When the BACL is enabled,
modifications made using /MODIFY-FILE-ATTRIBUTES are no longer evaluated for
USER-ACCESS or ACCESS.
●
Systems without SECOS:
The BACL group rights of the BS2000 file and the group rights mapped in the bs2fs file
system can be modified on systems on which SECOS is not active, but they are not
evaluated when files are accessed. Only the user rights and the rights for OTHER
count. This behavior is identical in the case of both BS2000 and bs2fs accesses.
●
Systems with SECOS, *UNIVERSAL group:
On systems on which SECOS is active but no SECOS groups have been configured
and assigned, all IDs are members of the *UNIVERSAL group. This has the following
consequences:
–
In the case of file accesses in native BS2000, only the user rights and the rights for
GROUP are evaluated, but not the rights for OTHER.
–
In the case of accesses in the bs2fs file system, the POSIX group and the rights for
OTHER play a role:
If the protection attribute -rw-r----- (0640) is set, read access by another ID is
rejected if the ID does not also belong to the same POSIX group. If this is not the
case, read permission for OTHER is required to permit access in read mode (0644).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
If the ID belongs to the same POSIX group, only read permission for GROUP is
required.
3.2.4 Security
bs2fs file systems permit access to files in BS2000. You can prevent unauthorized
accesses using the various protection mechanisms of BS2000 and POSIX. The following
protection mechanisms are available:
– User administration (BS2000)
– File access protection (BS2000 and POSIX)
Access rights in the bs2fs file system
Users always have the same access rights in the bs2fs file system as in BS2000. They can
therefore only perform the same types of access (read, write, execute) to a file or library
element which they would also be permitted to perform in native BS2000.
U41802-J-Z125-2-76
39
Working with the bs2fs file system
Use of bs2fs file systems
Access restrictions can occur in individual cases. If, for instance, a file is protected by a
password or guard, it is possible that a POSIX user is forbidden from using an access which
they would be permitted to use as a BS2000 user.
The table below provides an overview of the access rights required for the various
accesses:
Access
Object
Rights required for access
Create
File
Owner (userid) of the file system
(if the file already exists, write permission for the file is sufficient)
Library
element
Write and read permission for the library and also administration
permission for the library or the type if this permission is defined
(if the library element already exists, write permission for the library
element is sufficient)
Open for
reading
File
Read permission for the file
Library
element
Read permission for the library and read permission for the element
Open for
writing
File
Write and read permission for the file
Library
element
Write and read permission for the library and write permission for the
element
If the element is not opened for overwriting (O_TRUNC), read
permission for the element is also required.
Delete
Rename
File
Owner (userid) of the file system and write permission for the file
Library
element
Write and read permission for the library and also administration
permission for the library or the type if this permission is defined and
write permission for the element
File/
Library
element
Deletion rights are required for the source and creation rights for the
target
The rights defined with Basic Access Control Lists (BACLs) or ACCESS and USERACCESS apply for all other accesses.
i
By default the TSOS ID has the same rights as the owner of a file or a PLAM library.
No special rights are assigned to the SYSROOT ID for accessing objects in bs2fs
file systems.
40
U41802-J-Z125-2-76
Working with the bs2fs file system
Application recommendations
The following recommendations apply to ensure that the files and library elements which
are mounted in a bs2fs file system can be accessed by a POSIX user:
●
If a user is to be permitted to execute a file or library element, at least one read
permission must be set for them. The user must also have read permission in order to
open a file in write mode (e.g. open with O_WRONLY) or a library element in write
mode, but not in overwrite mode (e.g. open with O_WRONLY without O_TRUNC).
Read permission may not be withdrawn from the user if the object is protected by a
guard.
●
If a user is to be permitted to read, write or execute a file or a library element, the
relevant access may not be protected by a password.
●
To permit write access to a file or library element or to permit it to be deleted, this object
may not be protected by an expiration date (EXPIRATION-DATE) which is later than
the current dates.
●
For each access to library elements, to their properties or to the directory of a library a
user requires at least read permission for the library.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
U41802-J-Z125-2-76
41
Working with the bs2fs file system
Use of bs2fs file systems
3.2.5 Various special features of bs2fs
When you work with files in a bs2fs file system, you must bear in mind a few special features
(e.g. compared to working with local POSIX files), and these are summarized again below:
42
●
New files created in a bs2fs file system
New files or library elements which are created in a bs2fs file system are assigned
BACL attributes.
The file attributes ACCESS and USER-ACCESS are consequently irrelevant with
regard to protection.
●
DMS wildcard in the case of the mount command for the bs2fs file system
New DMS files (always of the type SAM) can be created only if their names comply with
the pattern specified in the mount command and the syntax rules of BS2000. In
particular, it is not possible to create files whose names begin with a period (.).
●
PLAM library: BS2000 file lock of the type read
When a PLAM library in the bs2fs file system is accessed (e.g. with ls), it is assigned a
read lock ("Lock Type INPUT") in BS2000. You can then not rename, delete or modify
the PLAM library’s file attributes. Read locks are automatically released again by the
bs2fs daemon if no further bs2fs accesses to the PLAM library have taken place within
approx. 20 seconds.
The bs2fsd daemon immediately releases locks of the type write ("Lock Type INOUT")
for a PLAM library as soon as there is no longer any element open for the job submitting
user in the library.
●
Write permission is required to delete files in a bs2fs file system
Deletion of files in a bs2fs file system is always rejected if no write permission exists. In
POSIX files can be deleted even without write permission (cf. rm with query or with the
-f option or mv with query).
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
●
Working with the bs2fs file system
Access rights for the privileged user
As is the case with the access rights in BS2000, only the TSOS ID has the same rights
as the owner of a file or of a PLAM library by default when accessing files in the bs2fs
file system. The SYSROOT ID, which is also assigned the privileged uid=0 in POSIX,
is treated like any other nonprivileged ID when bs2fs file accesses occur.
In the POSIX/UNIX world privileged IDs (uid=0) are allowed to perform accesses which
go beyond the file owner’s access options, e.g. reading a file, although the protection
bits do not grant the file owner read access.
However, in bs2fs file systems TSOS also only has the permissions which the owner of
an object has in BS2000. When SECOS is used (e.g. GUARDS), the owner can also
restrict TSOS rights.
●
Updating files in text mode
When existing records in SAM or ISAM files which are mounted with ftyp=text are
updated, the record length may not be modified. Shortening or lengthening a record
results in an EIO error. The write() is not executed, and after close() only those parts of
the BS2000 file are modified which were written successfully with write().
●
Time stamp
The time stamps of a file or library element determined with FSTAT are mapped in the
bs2fs file system to the fields of the stat data structure in POSIX as follows:
BS2000
POSIX
CRE-DATE/TIME
st_ctime
ACC-DATE/TIME
st_atime
CHANG-DATE/TIME
st_mtime
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The following must be borne in mind here:
–
While a file is open, the st_atime and st_mtime fields of the ufs shadow file are
accepted as the st_atime and st_mtime of the bs2fs in the bs2fs file system; after the
file has been closed, the values which FSTAT supplies in BS2000 are output again.
This can result in the modification times for a file displayed with the /SHOW-FILEATTRIBUTES command differing from the times ascertained with ls -l in the bs2fs
file system as long as the file is open.
–
The CRE-DATE/TIME and st_ctime fields are interpreted differently in BS2000 and
POSIX/UNIX. In BS2000, CRE-DATE/TIME specifies the time when a file was
created; in POSIX/UNIX, st_ctime logs the time when the management structures
were last changed (e.g. because of chmod)
U41802-J-Z125-2-76
43
Working with the bs2fs file system
44
Use of bs2fs file systems
●
File sizes
As long as a file or library element has not not been opened in bs2fs after mounting
(mount), the size of a file specified in the output of the ls -l command is calculated from
the number of PAM pages supplied by the BS2000-FSTAT macro multiplied by 2048.
Only after a file has been opened in the bs2fs file system is the display precise to the
byte. This value can be incorrect if the size of the file was modified outside the bs2fs.
●
Handling migrated files
Processing of migrated files is not supported. This has the following consequences:
–
ls, stat(), fstat() do not display the names of migrated files.
–
If you attempt to open a migrated file, errno 8 (ENOENT, "No such file or directory")
is returned.
–
If you attempt create a new file with the name of a migrated file, e.g. with creat(),
rename() or with the shell command cp, mv, errno 206 (ENXIO, "No such device or
address") is returned. ENXIO is unambiguous for bs2fs accesses, i.e. an attempt to
create a new file with the name of a migrated file is always involved.
●
X bit for directories
The x bit for a directory is interpreted differently in the bs2fs file system than is
customary in the UNIX world. In the bs2fs file system you require read permission to
search for an entry in the table of contents of a PLAM library, not execution permission.
●
The directory structure of a bs2fs file system is defined by the mount operation and
cannot be modified as it maps the structure of PLAM libraries. Directories in a bs2fs file
system can therefore not be deleted or renamed, nor is it possible to create new directories. Consequently no new PLAM libraries can be generated using bs2fs means (e.g.
mkdir), either. Nor can the access rights of directories be modified.
●
Library elements can be created in the defined directories (types) with any names which
comply with the syntax rules for PLAM libraries, except in directories which map a type
which does not have a standard type as its basic type.
●
Library elements which are assigned to a type which does not have a standard type as
its basic type can only be read.
●
Library elements of a type with the basic type L can only be renamed as or copied to
library elements of the same type or of another type with the basic type L. Likewise, files
or library elements of a type with a basic type other than L cannot be renamed as or
copied to an element of a type with the basic type L.
U41802-J-Z125-2-76
●
Working with the bs2fs file system
Library elements are stored in the bs2fs file system under a name which is composed
of the element name and the version identifier:
elementname+ver
In addition, a hard link elementname exists to the highest current version of the element.
This has the following consequences:
–
If a new element version with a version identifier which is higher than the existing
highest version identifier is generated in the bs2fs file system, elementname automatically refers to the new highest version.
–
Deleting elementname (e.g. with rm elementname) has the following consequences:
First the file containing the highest version of the element is deleted. If other
versions of the element exist, elementname subsequently refers to the new highest
version. Otherwise the hard link elementname is also deleted.
–
If a particular version of an element is deleted (e.g. with rm elementname+ver), the
hard link elementname is also deleted if no other versions of the element exist.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k03
Use of bs2fs file systems
U41802-J-Z125-2-76
45
Working with the bs2fs file system
46
Use of bs2fs file systems
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
12. March 2014 Stand 08:57.50
4 Interfaces for supporting bs2fs file systems
This chapter contains descriptions of the interfaces for managing bs2fs file systems, as well
as information on shell commands, POSIX tools and C program interfaces.
4.1 Administration interfaces
POSIX supports the management of bs2fs file systems with:
–
commands
–
daemons
–
administration files
You require POSIX root authorization (BS2000 ID TSOS or SYSROOT) to manage bs2fs
file systems.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
4.1.1 Commands
This section describes commands which are relevant for the use of bs2fs file systems in
alphabetical order. The notational conventions which are used in the command syntax are
explained in section “Notational conventions” on page 10.
Command
Function
mount
mount a file system
mountall
mount file systems
show_pubset_export
show file systems affected by EXPORT-PUBSET
start_bs2fsd
start copy daemons
umount
unmount a file system
umountall
unmount file systems
Table 1: bs2fs commands
U41802-J-Z125-2-76
47
mount
Commands
mount
mount a file system
mount (Format 2 and Format 3, see page 48) mounts a ufs file system into the file system
hierarchy at the path name position mountpoint. This position must already exist. If mountpoint possesses any contents prior to the mount operation these remain hidden until the file
sysem is dismounted again.
mount (Format 4 and Format 5, see page 51) mounts a bs2fs file system at a particular
position in the POSIX file system. A bs2fs file system is understood to be a selectable set
of files in BS2000 which are made available transparently in POSIX so that they can be
accessed using POSIX means (commands, program interfaces). The files are selected via
the user and catalog ID and wildcard symbols.
In addition, mount (Format 1) can be used to output a list of all the mounted file systems.
Syntax
Format 1: mount[Ë-vË|Ë-p]
Format 2: mount[Ë-FËufs][Ë-V][Ë-r][Ë-oËspec_options]{ËresourceË|Ëmountpoint}
Format 3: mount[Ë-FËufs][Ë-V][Ë-r][Ë-oËspec_options]ËresourceËmountpoint
Format 4: mount[Ë-FËbs2fs][Ë-V][Ë-r][Ë-oËspec_options]
{ËresourceË|ËmountpointË}
Format 5: mount[Ë-FËbs2fs][Ë-V][Ë-r][Ë-oËspec_options]ËresourceËmountpoint
Display a list of mounted file systems
Format 1
mount[Ë-vË|Ë-p]
No option specified
mount displays a list of all mounted file systems (see Example).
option
-v Displays a new presentation of the output. The new output contains the file system type
and options in addition to the information in the old output. The fields mountpoint and
resource change places (see Example).
-p Displays a list of the mounted file systems in the /etc/vfstab format (see Example).
mount ufs file systems
Format 2
mount[Ë-FËufs][Ë-V][Ë-r][Ë-oËspec_options]{ËresourceË|Ëmountpoint}
Format 3
mount[Ë-FËufs][Ë-V][Ë-r][Ë-oËspec_options]ËresourceËmountpoint
The descriptions of Format 2 and Format 3 have been combined, since they differ only in
terms of the (optional) specifications resource and mountpoint.
48
U41802-J-Z125-2-76
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
mount
Format 2 can be used only if an entry for the relevant file system already exists in the
/etc/vfstab file. The missing specification for resource or mountpoint is then added from this.
Formats 2 and 3 can be entered only by the POSIX administrator.
No option specified
mount displays a list of all mounted file systems.
-FËufs
Specifies ufs as the file system type.
-V Displays the entire command line on screen but does not execute the command. The
command line is displayed together with the options and arguments entered by the user
and with the values derived from /etc/vfstab. This option allows you to check the general
validity of the command line.
-r Mounts the file system with read access.
-o Specifies ufs file system-specific options. Multiple options should be comma-separated.
If invalid options are specified a warning is issued and the invalid options are ignored.
The following options may be selected:
f
Imitates an /etc/mntab entry but does not mount a file system. The parameters are not checked.
n
Mounts the file system without making an entry in /etc/mnttab.
journal
If there is no journal, one is created when the file system is mounted. This
journal is used for a quick restart after a system crash. When the ro option
is specified, the journal specification has no effect, i.e. there is no journaling
in the event of read-only accesses.
rw | ro
Read/write or read only access. The default value is rw.
nosuid
By default, the file system is mounted in such a way that the s-bit is set for
users. If you specify nosuid then the default value is deactivated and the file
system is mounted without the s-bit being set for users.
remount
Is used together with rw. A file system which has been mounted with read
access only can be remounted with read/write access. This option fails if the
file system is not currently mounted or has been mounted with rw.
bs2fscontainer
Specifies the file system which is to be mounted as the bs2fs container, i.e.
as a file system which temporarily accommodates files from bs2fs file
systems.
This option may only be specified for a single ufs file system. Any other
mount commando with this option is rejected.
U41802-J-Z125-2-76
49
mount
Commands
The -r, -o ro, -o journal and -o remount options may not be specified together
with the bs2fscontainer option.
When the POSIX installation program is used, this option can be entered via
the option line.
i
This option can only be specified for a file system which was flagged as the
bs2fs container when it was created with the POSIX installation program.
When the append function is applied to a ufs file system which is to be
created or overwritten, the bs2fscontainer option must be specified in the
option line for this purpose.
If this is not done, the mount is aborted with an error and the file system,
together with its content, is retained.
The ufs file system that serves as the bs2fs container is expected to be an
empty file system.
If it is not empty, its content is deleted using the –o bs2fscontainer option
when the mount command is executed.
After a successful mount, two bs2fsd copy daemons are started automatically.
resource
Specifies the file system which is to be mounted.
mountpoint
Specifies the local position for mounting resource. You must specify an absolute path
name. If mountpoint is a symbolic reference then the file system is mounted in the
directory to which the symbolic reference points rather than being mounted alongside
the symbolic reference.
50
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
mount
mount bs2fs file sytems
Format 4
mount[Ë-FËbs2fs][Ë-V][Ë-r][Ë-oËspec_options]{ËresourceË|ËmountpointË}
Format 5
mount[Ë-FËbs2fs][Ë-V][Ë-r][Ë-oËspec_options]ËresourceËmountpoint
The descriptions of Format 4 and Format 5 have been combined, since they differ only in
terms of the (optional) specifications resource and mountpoint.
Format 4 can be used only if an entry for the relevant file system already exists in the
/etc/vfstab file. The missing specification for resource or mountpoint is then added from this
(see also the note on page 54).
Formats 4 and 5 can be entered only by the POSIX administrator.
A prerequisite for entering a mount command of Format 4 or 5 is that a bs2fs container is
already mounted.
No option specified
mount displays a list of all mounted file systems.
-FËbs2fs
Specifies bs2fs as the file system type.
-V Displays the entire command line on screen but does not execute the command. The
command line is displayed together with the options and arguments entered by the user
and with the values derived from /etc/vfstab. This option allows you to check the general
validity of the command line.
-r Mounts the file system with read access.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
-o Specifies bs2fs file system-specific options. Multiple options should be commaseparated. If invalid options are specified a warning is issued and the invalid options are
ignored. The following options may be selected:
rw | ro
Read/write or read only access. The default value is rw.
nosuid
The file system is mounted without setting the s bit for users. This option is
enabled by default for bs2fs file systems and cannot be disabled.
remount
Is used together with rw. A file system which has been mounted with read
access only can be remounted with read/write access. This option fails if the
file system is not currently mounted or has been mounted with rw.
U41802-J-Z125-2-76
51
mount
Commands
ftyp={text|binary|textbin}
This option has the same effect as the ftyp command when copying files
using the bs2cp command. It defines whether BS2000 SAM files and texttype PLAM library element (element type other than L) are interpreted in
POSIX as text or binary files. PAM files are always interpreted as binary
files, ISAM files always as text files.
This option should only be specified once. If it is specified more than once,
the specification with the highest priority applies, where ftyp=textbin has the
highest priority, ftyp=text the next highest priority and ftyp=binary the lowest
priority.
The default is ftyp=text.
ftyp=text
SAM files and text-type library elements are interpreted as text files. When
writing to a bs2fs file, end-of-line characters (X'15') are converted to a
record change and tabulator characters (X'05') to the corresponding
number of blanks.
ftyp=binary
SAM files and text-type library elements are interpreted as binary files. A 1:1
transfer takes place without interpreting and converting data (record
change/end-of-line characters, tabulator/blanks, etc.).
ftyp=textbin
SAM files and text-type library elements are interpreted as binary text files.
When writing to a bs2fs file, only end-of-line characters (X'15') are
converted to a record change. Tabulator characters (X'05') are not
converted to blanks.
resource
Defines which BS2000 files are to be mounted. The following syntax applies for the
option:
:cat:$user.filename-with-wild
The option can be entered in upper- or lowercase or in a mixture of both. Special
characters of the POSIX shell such as ’$’ or ’*’ must be escaped explicitly.
52
cat
Catalog ID
user
BS2000 user ID
U41802-J-Z125-2-76
mount
filename-with-wild
BS2000 file name with wildcard symbols
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
Wildcard
Meaning
*
Replaces an arbitrary (even empty) string.
/
Replaces precisely one arbitrary character.
Terminating
period
Partially -qualified entry of a name.
Corresponds implicitly to the string “./*”, i.e. at least one
arbitrary character follows the period.
<sx:sy>
Replaces a string that meets the following conditions:
– It is at least as long as the shortest string (sx or sy)
– It is not longer than the longest string (sx or sy)
– It lies between sx and sy in the alphabetic collating
sequence; numbers are sorted after letters (A…Z, 0…9)
– sx can also be an empty string which is in the first position
in the alphabetic collating sequence
– sy can also be an empty string which in this position
stands for the string with the highest possible code
(contains only the characters X’FF’)
<s1,…>
–
-s
Replaces all strings that do not match the specified s. The
minus sign may only appear at the beginning of the string.
Replaces all strings that match any of the character
combinations specified by s. s can also be an empty
string. Any such string can also be a range specification
“sx:sy”.
The file set defined with resource can consist of both existing files and files which are to
be created. When a new file is to be created, the required file name must match the
wildcard pattern of the corresponding mount command.
mountpoint
Specifies the local position for mounting resource. You must specify an absolute path
name. If mountpoint is a symbolic reference then the file system is mounted in the
directory to which the symbolic reference points rather than being mounted alongside
the symbolic reference.
U41802-J-Z125-2-76
53
mount
Hint
Commands
If an entry for the file system concerned exists in the /etc/vfstab file, one of the options
resource or mountpoint can be omitted (Format 2 or 4). When a bs2fs file system is used, the
following must be observed in this case:
–
If the mountpoint option is specified, an entry in /etc/vfstab can then be identified
unambiguously and the corresponding file system is mounted.
–
If only the resource option is specified, multiple suitable entries for a bs2fs file system
can be contained in the /etc/vfstab file as this file system can be mounted in parallel at
multiple locations. In this case only the first bs2fs file system entered in the /etc/vfstab
file is mounted.
Only entries with an identical wildcard string are recognized as suitable entries. Entries
with a different wildcard string are also not taken into account even if they define the
same set of files.
File
/etc/mnttab
Table of mounted file systems.
/etc/vfstab
Table of automatically mounted file systems.
Example
Mounting the bs2fs container and a bs2fs file system. The example executes under the
POSIX administrator ID.
# mount –F ufs –o bs2fscontainer /bs2fscont
# mount -F bs2fs ':V70a:$sysaudit.sys.conslog.2007-06*'
# mount
/home/bs2.conslog
/ on /dev/root read/write/setuid on Tue Nov 27 11:31:04 2007
.
.
.
/bs2fscont on /dev/dsk/23 bs2fscontainer/setuid/read/write/noquota on Tue Nov 27
11:35:23 2007
/home/bs2.conslog on :V70A:$SYSAUDIT.SYS.CONSLOG.2007-06* ftyp=text/nosuid on Tue Nov
27 13:52:23 2007
#
54
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
mountall
mountall
mount file systems
mountall is used to mount file systems on the basis of a file_system_table (/etc/vfstab is the
default file system table). The special file name "-" reads from the standard input. If you
specify the hyphen then the standard input must possess the same format as /etc/vfstab.
12. March 2014 Stand 08:57.50
Before the individual file systems are mounted, fsck performs a plausibility test to determine
whether the system appears to be viable for mounting (not in the case of file systems of the
type bs2fs or nfs). If the system is not viable for mounting, fsck corrects it before an attempt
is made to mount it.
If only file_system_type is specified then mountall applies only to file systems of the specified
type.
The file systems are mounted in the order ufs - bs2fs - nfs. This ensures that when the bs2fs
file systems are mounted, the bs2fscontainer file system required for this purpose is already
mounted in ufs.
Syntax
mountall[Ë-FËfile_system_type] [Ë-l | Ë-r][file_system_table]
No option specified
mountall mounts all file systems for which the field automnt in the file_system_table is set
to yes.
options
-FËfile_system_type
Specifies the type of file system to be mounted.
-l
Limits the process to local file systems (ufs and bs2fs).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
-r Limits the process to remote file system types (nfs).
file_system_table
If file_system_table is not specified, mountall refers to /etc/vfstab.
Hint
If the -F option is specified together with one or more of the options -l, -r and -b and the
options are mutually compatible, the -l, -r and -b options have priority. For example, mountall
-F bs2fs -l and mountall -F ufs -l have the same effect as mountall -l: all local file systems (i.e.
all ufs and bs2fs file systems) are mounted. The entries mountall -F bs2fs and mountall -b
also lead to the same result: all bs2fs file systems are mounted.
Error
If the file systems are viable for mounting and error-free, no message is output. Error and
warning messages are issued by fsck and mount or by mountall in the case of incorrect
syntax.
File
/etc/vfstab
Default file system table.
See also
fsck, mount, umountall
U41802-J-Z125-2-76
55
show_pubset_export
Commands
show_pubset_export
show file system affected by pubset export
This command supplies the system administrator with information on which file systems in
POSIX are affected by the export of a pubset and must therefore be unmounted before the
EXPORT-PUBSET command is executed.
This information is particularly helpful when bs2fs file systems are used. Unlike with the ufs
and NFS file systems, it is not sufficient in this case to check whether the mount point of a
file system is located on the pubset concerned.
With bs2fs file systems the location of the BS2000 files mounted by means of bs2fs is also
relevant. The mount point of the bs2fs container also plays a role. If it is located on the
pubset to be exported, all mounted bs2fs file systems are affected by the export,
irrespective of their location.
The table below shows when a file system is affected by EXPORT-PUBSET:
Syntax
File system type
Affected if ... located on the pubset to be exported
ufs
Mount point or container file
nfs
Mount point
bs2fs
Mount point or mounted BS2000 files or mount point/container file of the bs2fs
container (cf. ufs)
show_pubset_export cat-id
cat-id
Catalog ID of the pubset which is to be checked (without enclosing colons “:”). The entry
can be made in upper- or lowercase notation or a mixture of both; the check is
performed with the cat-id converted to uppercase notation.
Files
The following files are searched for the specified catalog ID in order to determine the file
systems affected:
/etc/mnttab
Table of all mounted file systems
/etc/partitions
Table of all possible partitions
If the catalog ID entry is missing in this file, the default ID is determined via BS200 and used
for the check.
SYSSSI.POSIX-BC.<version>
SYSSSI file of POSIX from BS2000
The BS2000 file name of the container file of the root file system is determined from this file
(ROOTFSNAME parameter) as this name is not entered in the /etc/partitions file.
56
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
Example
show_pubset_export
The file systems affected by the export of the DATA pubset are determined. The container
file of the ufs file system mounted under the home directory /home/bach resides on the DATA
pubset.
# show_export DaTa
the nfs filesystems on pubset DATA
nfs filesystem PGTR0157:/home4 mounted on /home/bach/nfs4
nfs filesystem PGTR0157:/home5 mounted on /home/bach/nfs5
the bs2fs filesystems on pubset DATA
bs2fs filesystem :DATA:$BACH.ASS.* mounted on /home/bach/bs2/mount.ass
bs2fs filesystem :V70A:$BACH.CCC.* mounted on /home/bach/bs2/mount.ccc
bs2fs filesystem :DATA:$BACH.PLAM* mounted on /home/bach/bs2/mount.plam
the ufs filesystems on pubset DATA
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
ufs filesystem /dev/dsk/4 mounted on /home/froede
ufs filesystem /dev/dsk/5 mounted on /home/bach
#
U41802-J-Z125-2-76
57
start_bs2fsd
Commands
start_bs2fsd start copy daemons
start_bs2fsd enables the system administrator to start additional copy daemons bs2fsd.
Syntax
start_bs2fsd [number]
No option specified
start_bs2fsd provides information about the number of copy daemons currently running.
number
Specifies how many copy daemons are to be started in addition to the copy daemons
already started. Up to 8 copy daemons can execute.
Example
Check how many daemons are currently running
# start_bs2fsd
/sbin/start_bs2fsd: 2 bs2fs daemons are running
Start an additional daemon and repeat the check
# start_bs2fsd 1
/sbin/start_bs2fsd: start additional daemon 1 of 1
# start_bs2fsd
/sbin/start_bs2fsd: 3 bs2fs daemons are running
58
U41802-J-Z125-2-76
umount
umount
unmount a file system
The command umount unmounts a file system which was mounted using mount. The file
system entry is deleted from the table /etc/mnttab.
Syntax
umount[Ë-V]Ë{resource | mountpoint}
option
-V Outputs the entire command line on screen but does not execute the command. The
command line is generated with all the options and arguments specified by the user as
well as the values taken from /etc/vfstab. You should select this option in order to subject
a command line to a general check and validity check.
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Commands
resource
Specifies the resource which is to be unmounted. Format as for mount.
For bs2fs file systems the option must be specified in UPPERCASE in accordance with
the entry in the internal tables. Special characters of the POSIX shell such as ’$’ or ’*’
must be escaped explicitly.
mountpoint
Specifies the local position at which resource must be unmounted. You must specify an
absolute path name.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Hint
Specification of the mountpoint option is always recommended for unmounting bs2fs file
systems. If a bs2fs file system is mounted in multiple positions in the POSIX file system (i.e.
identical resource specification in mount) and only resource is specified for umount, only the
file system mounted last is unmounted. However, if mountpoint is specified, the corresponding file system is always unmounted.
The umount command is rejected if it refers to a file system mounted using the bs2fscontainer
option and at least one bs2fs file system is still mounted. In the case of a successful umount
for the bs2f container, the bs2fsd copy daemons are terminated automatically.
File
/etc/mnttab
Table of mounted file systems.
/etc/vfstab
Table of automatically mounted file systems.
See also
mount, umountall
mount, umount [3]
U41802-J-Z125-2-76
59
umountall
Commands
umountall unmount file systems)
umountall unmounts all mounted file systems with the exception of root, /proc, /var and /usr.
If only file_system_type is specified, umountall relates only to file systems of the specified
type. The file systems are unmounted in the order nfs - bs2fs - ufs. This ensures that the
bs2fscontainer file system required for bs2fs file systems is only unmounted in the ufs when
it is no longer needed, i.e. when no further bs2fs file systems are mounted.
Syntax
umountall[Ë-FËfile_system_type][Ë-k][Ë-l |Ë-r]
-F Specifies the file system type to be unmounted.
-k Sends the SIGKILL signal to processes which have opened files in the file system.
-l
Limits the operation to local file systems (ufs and bs2fs).
-r Limits the operation to remote file systems (nfs).
-b Limits the operation to bs2fs file systems.
Error
If the file systems can be unmounted no message is output. Error and warning messages
are issued by fsck and mount.
Hint
If the -F option is specified together with one or more of the options -l, -r and -b and the
options are mutually compatible, the -l, -r and -b options have priority. For example, mountall
-F bs2fs -l and mountall -F ufs -l have the same effect as mountall -l: all local file systems (i.e.
all ufs and bs2fs file systems) are unmounted. The entries mountall -F bs2fs and mountall -b
also lead to the same result: all bs2fs file systems are unmounted.
File
/etc/mnttab
Table of mounted file systems.
/etc/vfstab
Table of automatically mounted file systems.
See also
60
fsck, mount, mountall, umount
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Interfaces for supporting BS2000 file systems
Copy daemon bs2fsd
4.1.2 Copy daemon bs2fsd
This section describes the copy daemon bs2fsd.
The notational conventions which are used in the command syntax are explained in section
“Notational conventions” on page 10.
The copy daemon bs2fsd transfers BS2000 files and elements of PLAM libraries from
BS2000 to the corresponding area of the bs2fs container and vice versa.
A file or library element is transferred from BS2000 to the bs2fs container automatically
when the first access to the file takes place in the bs2fs file system (first open); in the case
of write access, transfer back to BS2000 takes place after editing in POSIX has been
completed (last close).
If the file or library element has been opened for writing, it remains open in BS2000 from
the start of transfer from BS2000 until its transfer back to BS2000 has been completed. It
can therefore not be modified by other users while it is being edited in POSIX. This
guarantees that the file has a consistent status.
By default 2 bs2fsd daemons are started when the bs2fs container is mounted. However, the
system administrator has the option of starting further copy daemons when required (e.g. a
large number of bs2fs file systems, a large number of files, large files) using the start_bs2fsd
command (page 58).
The daemons run as POSIX background processes mainly in the TU under the TSOS user
ID. After they have been started, they “sleep” until they are “woken up” by a request.
In addition, you can use the start_bs2fsd command to determine how many copy daemons
are running at the moment.
When the bs2fs container is unmounted, all copy daemons are terminated automatically.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
You can monitor the daemons by, for instance, entering the following POSIX command:
ps -ef | grep bs2fsd
U41802-J-Z125-2-76
61
Administration files
Interfaces for supporting BS2000 file systems
4.1.3 Administration files
This section describes the administration files.
The administration files /etc/mnttab and /etc/vfstab are employed for the application of bs2fs
file systems, as is customary for file system management in POSIX.
The /etc/mnttab file is automatically created by POSIX at startup and updated with each
mount/umount operation.
The entries for ufs file systems and native file systems in the /etc/vfstab file are created by
the POSIX installation program. Entries for nfs and bs2fs file systems can be processed by
the administrator (root authorization) using an editor.
The administration files are listed in the table below:
Path
File
Function
/etc
mnttab
Table of the mounted files systems
/etc
vfstab
Table of the defined files systems
Table 2: Administration files
62
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Administration files
/etc/mnttab
/etc/mnttab
table of the mounted file systems
The /etc/mnttab file contains details of all the file systems mounted on the local computer.
This file contains information which is generated by the mount command.
Each line contains the following information, which is separated by an arbitrary number of
blanks and/or tabulators:
Format
resource
mountp
fstyp
spec-options
time
resource
Absolute path name of the mounted file system or, in the case of bs2fs file
systems, mounted BS2000 files in wildcard syntax.
For bs2fs file systems the entry differs from the specification in the mount
command as follows:
– It is converted completely into uppercase notation
– Characters to be escaped are displayed without the associated escape
character
Example:
mount –F bs2fs –o ftyp=text :v70a:\$bach.sys\* /home/bach/bs2.1
generates the following entry in /etc/mnttab:
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
:V70A:$BACH.SYS* /home/bach/bs2.1 bs2fs ftyp=text ...
mountp
Absolute path name of the mount point.
fstype
File system type.
spec-options
Options as specified in the mount command.
time
Mount time, specified in seconds since 1/1/1970
Entries in the /etc/mnttab file are deleted again if the umount or umountall command is
executed for corresponding file systems or file system types.
U41802-J-Z125-2-76
63
/etc/mnttab
Administration files
Example
Enter the following in the POSIX shell: cat /etc/mnttab
/dev/root
/proc
/dev/fd
/dev/dsk/3
/dev/dsk/4
/dev/dsk/10
/dev/dsk/5
/dev/dsk/2
/dev/dsk/23
:V70A:$BACH.ASS.*.S
:V70A:$BACH.CCC.*.C
:V70A:$BACH.PLAMLIB*
:V70A:$BACH.SEM*.C
64
/
/proc
/dev/fd
/var
/home/froede
/home/gast
/home/bach
/home/bach/mount99
/bs2fscont
/home/bach/bs2.1
/home/bs2.2
/home/bach/bs2.2
/home/bs2000
ufs
proc
fdfs
ufs
ufs
ufs
ufs
ufs
ufs
bs2fs
bs2fs
bs2fs
bs2fs
rw,suid
rw
rw
suid,rw,noquota
suid,rw,noquota
suid,rw,noquota
suid,rw,noquota
suid,rw,noquota
bs2fscontainer,suid,rw,noquota
ftyp=binary
ftyp=text
ftyp=textbin
ftyp=text
1196069614
1196069614
1196069614
1196069614
1196069615
1196069615
1196069621
1196069621
1196070061
1196084245
1196084250
1196084255
1196084261
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Administration files
/etc/vfstab
/etc/vfstab
table of the defined file systems
The /etc/vfstab file describes every file system defined on the local computer. You can
process the file using an editor.
The file systems for which yes is entered in the automnt column in the /etc/vfstab file are
mounted automatically when POSIX starts up or when the mountall command is executed.
The entries in the file are also used to add any missing details for resource or mountpoint and
mount options when a mount command is executed.
Corresponding entries are generated automatically in the /etc/vfstab file for ufs file systems
which are defined using the POSIX installation program. The entries must be created
manually for all other file systems (e.g. bs2fs or nfs file systems) when required.
In contrast to the /etc/mnttab file, execution of the mount and umount commands for the
/etc/vfstab file has no effect. Corresponding entries are retained.
The fields in the table are separated by tabulators and/or blanks. A hyphen (–) indicates an
empty entry in the field. The table contains the following fields:
Format
specialËfsckdevËmountpËfstypeËfsckpassËautomntËmntopts
special
Defines the resource to be mounted.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The following must be borne in mind in the case of (manual) entries for bs2fs
file systems:
– Letters may only be specified in uppercase.
– Special characters may not be escaped, nor is it permissible to enclose the
string in quotes.
fsckdev
Name of the block-oriented device or of the resource of the character-oriented
device.
mountp
Mount point: absolute path name of the directory in which the resource is to be
mounted.
fstype
File system type.
fsckpass
The sequence number to be used for multiple fsck commands.
automnt
Specifies whether the resource is to be mounted automatically when POSIX
starts up or by means of the mountall command (yes) or not (no).
mntopt
List of options for mounting the file system, separated by commas. The options
correspond to the spec_options of the mount command.
U41802-J-Z125-2-76
65
/etc/vfstab
Administration files
Example
Enter the following in the POSIX shell: cat /etc/vfstab
/dev/root
/dev/rroot
/proc
/dev/fd
/dev/dsk/3
/dev/rdsk/3
172.25.86.64:/home2/froede/SHARE PGOB0004:/home2/froede/SHARE
/dev/dsk/4
/dev/rdsk/4
/dev/dsk/10
/dev/rdsk/10
/dev/dsk/13
/dev/rdsk/13
/dev/dsk/8
/dev/rdsk/8
/dev/dsk/23
/dev/rdsk/23
/dev/dsk/24
/dev/rdsk/24
/dev/dsk/25
/dev/rdsk/25
/dev/dsk/26
/dev/rdsk/26
/dev/dsk/5
/dev/rdsk/5
/dev/dsk/2
/dev/rdsk/2
/dev/dsk/6
/dev/rdsk/6
:V70A:$BACH.ASS.*.S
:V70A:$BACH.CCC.*.C
:V70A:$BACH.PLAMLIB*
:V70A:$BACH.SEM*.C
-
66
/
/proc
/dev/fd
/var
/home/froede/RETSINA
/home/froede/PGOB0004
/home/froede
/home/gast
/mnt/ascii
/mnt/dat1
/bs2fscont
/home/bach/mount3
/home/bach/mountxxx
/home/bach/mountyyy
/home/bach
/home/bach/mount99
/suderlan
/home/bach/bs2.1
/home/bs2.2
/home/bach/bs2.2
/home/bs2000
ufs
proc
fdfs
ufs
nfs
nfs
ufs
ufs
ufs
ufs
ufs
ufs
ufs
ufs
ufs
ufs
ufs
bs2fs
bs2fs
bs2fs
bs2fs
1
1
1
1
1
1
1
1
1
1
1
1
1
-
yes
no
no
yes
no
no
yes
yes
no
no
no
no
no
no
yes
yes
no
yes
yes
yes
yes
soft
soft
-o
ftyp=binary
ftyp=text
ftyp=textbin
-
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Information on shell commands and POSIX tools
df
4.2 Information on shell commands and POSIX tools
This section describes special aspects when accessing files of bs2fs file systems using
shell commands and POSIX tools.
df
report free disk space)
The statvfs() interface (64-bit variant) is used implicitly in the df command. In the case of a
bs2fs file system, the data output must then be interpreted accordingly (see also statvfs()
page 74). All 3 formats of the df command are supported for the file system type bs2fs.
Syntax
Format 1: df[Ë-FËFSType]Ë -P[Ë-klV][Ëfile]
Format 2: df[Ë-FËFSType][Ë-begklntVv][Ë-oËufs_options]...[Ëfile]
Format 3: dfË-c[Ëfile]...
bs2fs-specific options and rules of the df command:
-F bs2fs
Identifies a bs2fs file system.
file
A path name or the resource can be specified.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Path name:
bs2fs mount point, simple bs2fs file, PLAM library index, type directory, library
element
Resource:
Resource of the mount command for the bs2fs file system in accordance with the
mnttab entry in uppercase notation (e.g. 'V70A:$BACH.ASS.*.S').
dumpfs
dump file system
The dumpfs command is not supported for bs2fs file systems.
fsck file system check
The fsck command is not supported for bs2fs file systems.
U41802-J-Z125-2-76
67
fsexpand
Information on shell commands and POSIX tools
fsexpand
expand existing file systems
The fsexpand command is not supported for bs2fs file systems.
pathchk
check pathnames
In the case of files or directories in a bs2fs file system, an error message is output if the
name length does not comply with the rules of a bs2fs file system.
pdbl/posdbl
set up and manage user-specific/global
program cache
Preloading of executables (LMS element, type L or simple "binary" DMS file) residing in the
bs2fs file system with posdbl or pdbl (-b option) is supported.
Syntax
pdbl{Ë-s[Ësid]|Ë-u}Ë-bËpath
(relevant
options only)
posdblË-bËpath
Example for an LLM element
# posdbl -s
POSIX-DBL:
POSIX-DBL:
Cache POSIX@DBL
68
linker ON
loader ON
holder of global cache: TSN=031F PID=19
CREATED: 04/10/08 12:24:46
SIZE: 32 MB
ENTRIES: 16
FREE PAGES: 7112
U41802-J-Z125-2-76
pdbl/posdbl
# posdbl -l
RM
39 Apr 10 12:25:30 $TSOS.SINLIB.POSIX-BC.070.SHELL
+IN@RLOGIND
113 Apr 11 11:16:47 $TSOS.SINLIB.POSIX-BC.070.ROOT
+HD
35 Apr 10 12:25:08 $TSOS.SINLIB.POSIX-BC.070.ROOT
PS
49 Apr 11 11:33:22 $TSOS.SINLIB.POSIX-BC.070.SHELL
SH
243 Apr 11 12:20:29 $TSOS.SINLIB.POSIX-BC.070.SHELL
MAN
37 Apr 11 12:14:39 $TSOS.SINLIB.POSIX-SH.070
WHO
46 Apr 10 15:25:43 $TSOS.SINLIB.POSIX-SH.070
MORE
107 Apr 11 12:14:39 $TSOS.SINLIB.POSIX-SH.070
SORT
53 Apr 11 12:20:32 $TSOS.SINLIB.POSIX-BC.070.SHELL
CRON
59 Apr 10 12:25:31 $TSOS.SINLIB.POSIX-SH.070
GREP
41 Apr 11 12:02:40 $TSOS.SINLIB.POSIX-BC.070.SHELL
+STTY
39 Apr 11 11:17:06 $TSOS.SINLIB.POSIX-BC.070.ROOT
EXPR
40 Apr 10 12:24:57 $TSOS.SINLIB.POSIX-BC.070.SHELL
UNAME
35 Apr 11 11:17:00 $TSOS.SINLIB.POSIX-BC.070.SHELL
+POSDBL
44 Apr 11 12:38:53 $TSOS.SINLIB.POSIX-BC.070.ROOT
+IN@RSHD
97 Apr 10 12:25:11 $TSOS.SINLIB.POSIX-BC.070.INET
# posdbl -b /home/bach/bs2/plamlib.4/l/usp
# posdbl -l
RM
39 Apr 10 12:25:30 $TSOS.SINLIB.POSIX-BC.070.SHELL
+IN@RLOGIND
113 Apr 11 11:16:47 $TSOS.SINLIB.POSIX-BC.070.ROOT
+HD
35 Apr 10 12:25:08 $TSOS.SINLIB.POSIX-BC.070.ROOT
PS
49 Apr 11 11:33:22 $TSOS.SINLIB.POSIX-BC.070.SHELL
SH
243 Apr 11 12:20:29 $TSOS.SINLIB.POSIX-BC.070.SHELL
+usp
59 Apr 11 12:41:02 /home/bach/bs2/plamlib.4/l/usp
MAN
37 Apr 11 12:14:39 $TSOS.SINLIB.POSIX-SH.070
WHO
46 Apr 10 15:25:43 $TSOS.SINLIB.POSIX-SH.070
MORE
107 Apr 11 12:14:39 $TSOS.SINLIB.POSIX-SH.070
SORT
53 Apr 11 12:20:32 $TSOS.SINLIB.POSIX-BC.070.SHELL
CRON
59 Apr 10 12:25:31 $TSOS.SINLIB.POSIX-SH.070
GREP
41 Apr 11 12:02:40 $TSOS.SINLIB.POSIX-BC.070.SHELL
+STTY
39 Apr 11 11:17:06 $TSOS.SINLIB.POSIX-BC.070.ROOT
EXPR
40 Apr 10 12:24:57 $TSOS.SINLIB.POSIX-BC.070.SHELL
UNAME
35 Apr 11 11:17:00 $TSOS.SINLIB.POSIX-BC.070.SHELL
+POSDBL
44 Apr 11 12:41:18 $TSOS.SINLIB.POSIX-BC.070.ROOT
+IN@RSHD
97 Apr 10 12:25:11 $TSOS.SINLIB.POSIX-BC.070.INET
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Information on shell commands and POSIX tools
U41802-J-Z125-2-76
69
Programming notes
Interfaces for supporting BS2000 file systems
4.3 Information for C program interfaces
This section describes special aspects when accessing files of bs2fs file systems via
program interfaces.
You can use the sysfs interface (see the “C Library Functions for POSIX Applications“
manual and “sysfs - query information about the file system type” on page 75) to determine
whether bs2fs file systems are supported in an existing POSIX installation. POSIX
V7.0/V8.0 and higher under BS2000/OSD-BC V7.0 and higher supports bs2fs file systems
by default.
You can check whether a bs2fs file system is mounted using the mount command (see
page 48) or by searching the mount table /etc/mnttab (see page 63). For each mounted file
system this table contains an entry with the data of this file system.
70
U41802-J-Z125-2-76
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Interfaces for supporting BS2000 file systems
Programming notes
4.3.1 errnos
You can access files of bs2fs file systems using the same program interfaces as for when
you access files of ufs file systems. However, in the event of an error the errno variable can
contain additional values or values with a different meaning. The meaning of the various
error numbers when you access bs2fs file systems is provided in the table below:
Error number
Meaning
EAGAIN
A BS2000 file can temporarily not be accessed, e.g. because the file was opened
by another process or because the connection to the pubset has been interrupted.
EDOM
Internal interface error
EFAULT
Internal processing error or invalid address (e.g. buffer address)
EIO
DMS or LMSUP or PLAM reported I/O error
EMACRO
DMS or LMSUP or PLAM reported processing error
ENAME
Part of the path name is not a valid BS2000 file name, PLAM type name or element
name (the latter possibly including a valid version identifier)
ENODEV
The volume or the device cannot be accessed
ENOEXEC
An LLM element which is to be edited does not have an LLM structure.
ENONET
Pubset is not available, possibly because of a problem in the MSCF network
ENOSTR
An attempt is being made to access a BS2000 file which does not have a file type
which is supported
ENOSUBSYS
No connection could be established to LMSUP or PLAM
ENOSYS
Operation not supported by the bs2fs file system
ENXIO
File was displaced (migrated)
EOPR
Operation is not supported for the specified file
ESRCH
No bs2fs daemon exists which could perform the accesses to the BS2000 object.
Response: notify the system administrator
Error values such as EIO, ENOEXEC or ENOSPC can also be set for close if the file was
opened in write mode and could not be written back properly to the BS2000 file or the library
element. In these cases the copy is retained in the bs2fs container.
U41802-J-Z125-2-76
71
Programming notes
Interfaces for supporting BS2000 file systems
4.3.2 fstat, stat – bs2fs file structure
The stat structure of the system calls stat() and fstat() supplies all the important information
about a POSIX file. This information is, among other things, also evaluated and edited by
the POSIX command ls.
In the case of bs2fs files there are a few special features, and these are listed in the table
below.
In some cases this information differs depending on whether the file is currently open or
closed. When it is open, the file resides in the bs2fs container as a “shadow file”.
stat() enables the information to be called for both closed and open objects, fstat() only for
closed objects.
Information for closed bs2fs object
(stat)
st_ino
Unique number between a bs2fs mount and umount
st_dev
Unique number of the bs2fs file system in which the file identified with st_ino resides.
st_nlink
st_size
In the case of DMS files: "normally" 1
In the case of library elements with highest version: 2
In the case of other library elements:1
DMS file:
last written PAM page in bytes
Library element:
size of PAM pages in bytes
Size of the shadow file in the container
st_uid
bs2-POSIX-user
st_gid
bs2-POSIX-group
st_atime
Attributes of the BS2000 file or library
st_mtime
element
st_ctime In the case of library elements st_atime has
the same value as in the library or an
element-specific value if LMS is activated
st_mode
72
Information for opened bs2fs object
(stat or fstat)
Attributes of the shadow file in the
container
Attributes of the BS2000 file or library element
U41802-J-Z125-2-76
Programming notes
Explanation of the time stamps
st_atime
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Interfaces for supporting BS2000 file systems
When an object is opened in read mode, the BS2000 file’s time stamp is
updated in the context of the BS2000 close operation, which takes place
immediately after the shadow file has been copied into the container, which may
be necessary. During read accesses in the container, the time stamp is updated
following every read access to the shadow file. The time stamps of the BS2000
and shadow files then differ, i.e. the BS2000 file always has an “older2” atime
status than the shadow file. (While the file is open, the application only ever
sees the shadow file’s time stamp. After the close(), however, it sees the “older”
time stamp of the BS2000 file again.)
No access time stamp is normally maintained for library elements; the st_atime
field then contains the library’s access time stamp. Only if the maintenance of
access time stamps was defined for the library using the LMS statement
//MODIFY-LIBRARY-ATTRIBUTES with the ACCES-DATE=*KEEP operand or
with the corresponding LMSUP function will the access time stamps for its
elements be updated from this point, and they can then be transferred to the
st_atime field.
The st_ctime field for the file status is not changed after chmod(); as for BS2000
files, st_ctime represents the creation time (CREATION TIME in FSTAT).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
st_ctime
U41802-J-Z125-2-76
73
Programming notes
Interfaces for supporting BS2000 file systems
4.3.3 fstatvfs, statvfs - file system information
The program interfaces fstatvfs and statvfs supply information about the file system which
contains a file. This file is specified by a file descriptor in the case of fstatvfs and by its name
in the case of statvfs.
In both cases the information is output in the format which is defined in the sys/statvfs.h
header in the statvfs_t structure.
typedef struct statvfs {
unsigned long f_bsize;
unsigned long f_frsize;
unsigned long f_blocks;
unsigned long f_bfree;
unsigned long f_bavail;
unsigned long f_files;
unsigned long f_ffree;
unsigned long f_favail;
unsigned long f_fsid;
char f_basetype[FSTYPSZ];
unsigned long f_flag;
unsigned long f_namemax;
char f_fstr[32);
unsigned long f_filler[16];
} statvfs_t;
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
/*
Block size of the file system */
Fragment size */
# Blocks on file system with size f_frsize */
# Free blocks with size f_frsize */
# Available free blocks for non-superusers */
# File nodes (inodes) */
# Free file nodes (inodes) */
# Available free inodes for non-superusers */
File system ID (number) */
Name of the target file system, Null-terminated */
Bit mask for the options of f_flag */
Maximum length of the file names */
File-system-specific string */
Reserved for future versions */
This format is, however, tailored to ufs file systems. For a bs2fs file system the user must
interpret this information as shown in the table below. The data is generated using the
BS2000 interface $SJINFO.
unsigned long f_bsize
Set to 2048 (PAMBLOCKSIZE)
unsigned long f_frsize
Set to 2048 (PAMBLOCKSIZE)
unsigned long f_blocks (*) This data is obtained from the corresponding values of the
BS2000 interface $SJINFO. However, not all BS2000 values can
be mapped to an exact equivalent which is used for ufs file
systems.
unsigned long f_bfree
(*)
unsigned long f_bavail (*)
unsigned long f_files
(*)
unsigned long f_ffree
(*)
unsigned long f_favail (*)
74
unsigned long f_fsid
Contains the index which is used by the sysfs interface
char f_basetype[FSTYPSZ]
Contains the string “bs2fs” for identification
unsigned long f_flag
Only the ST_RDONLY flag is relevant
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Interfaces for supporting BS2000 file systems
unsigned long f_namemax
Programming notes
Is set to 38 - 41 in accordance with the catalog and user IDs. This
value applies for DMS files and PLAM libraries only. In the case of
library elements or types, the variable is assigned the value for the
higher-ranking PLAM library. The restrictions for names of
elements and element types can only be ascertained with
pathconf. Within the restriction of "Elementname+Version" to 89
characters it must be borne in mind that the Elementname
component is limited to 64 characters and the Version component
to 24 characters.
char f_fstr[32)
(*) unsigned long long in the case of fstatvfs64() or statvfs64()
4.3.4 sysfs - query information about the file system type
The sysfs program interface supplies information about the file system types configured in
a system.
Depending on the function argument, sysfs supplies
–
the index which corresponds to the file system type with the name specified in the
system
–
the name which corresponds to the file system type with the specified index
–
the maximum number of file system types configured in the system
Example
if (sysfs(GETFSIND, “bs2fs“) < 0) {
/* bs2fs NOT SUPPORTED */
}
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
This interface enables you to check whether a system is capable of processing bs2fs files.
The argument for the file system type bs2fs is “bs2fs”.
U41802-J-Z125-2-76
75
Programming notes
Interfaces for supporting BS2000 file systems
4.3.5 Further special aspects and restrictions
Restrictions for accessing files in bs2fs file systems exist for the following program interfaces:
chmod()
This function is accepted only at file and element level. In all other cases it is
rejected with EISDIR. The S_ISUID ("set user id on execution") and S_ISGID ("set
group id on execution") bits are not defined for BS2000 files and library elements
and are ignored (if specified in chmod). The st_ctime field for the file status is not
changed; as for BS2000 files, it represents the creation time (cf. fstat).
chown(), fchown()
The functions are rejected with ENOSYS if an attempt is made to modify the user
and/or group number.
close()
After errors in the case of close() (e.g. with the errno values EIO, ENOEXEC, ENOSPC),
the shadow file is retained. With the support of the POSIX administrator it can be
used to save the data.
creat(), open()
These functions are accepted only at file and element level. In all other cases they
are rejected with EISDIR. A file or library element cannot be opened in write mode
simultaneously in BS2000 and in a bs2fs file system. If a file or a library element in
a bs2fs file system is opened in write mode, it cannot be opened in another bs2fs
file system. If a file is not generated by the owner (user ID of the file system) but by
a process with TSOS privilege, it is nevertheless assigned the uid and gid of the file
system’s owner. If a file generation group (which is not visible in the bs2fs file
system) with the specified file name exists, the call is rejected with ENOSTR.
A library element may be generated not only by the owner, but also by any user who
has write permission and, if necessary, administration permission for the library or
type. However, if the user who generates a library element is not the owner, the
protection attributes specified in the mode argument are ignored and standard
protection attributes are set.
creat(), open() mit O_EXCL:
If a file which is not visible in the bs2fs file system exists, in other words one with
properties which are not supported (STORAGE-LEVEL î S0, SUPPORT î
PUBLIC), the call is rejected with ENXIO. Nevertheless the file remains invisible (e.g.
for the ls command), but "bs2cmd fstat ..." can be used to check whether it exists.
link()
This function is not supported. The call is rejected with ENOSYS.
mkdir(), rmdir()
These functions are always rejected with ENOSYS.
76
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k04
Interfaces for supporting BS2000 file systems
Programming notes
open() with O_RDWR under ftyp=text or ftyp=textbin
In the case of text files (ftyp=text or ftyp=textbin) of the type SAM, CRTE in
conjunction with SAM forces the existing record structure to be retained because of
fopen(...,"r+") in the bs2fs daemon, i.e. in the file area which exists after open() each
new line must remain unchanged in the position it occupies and no new line may be
inserted. New records can be appended and modified at the end of the existing file.
readlink()
This function is always rejected with ENOSYS.
remove(), unlink()
These functions are accepted only at file and element level. In all other cases they
are rejected with EISDIR.
rename()
This function is accepted only at file and element level. In all other cases it is
rejected with EISDIR. However, it is possible to convert files to library elements and
vice versa, e.g. using the cp or mv command. Library elements of the basic type L
cannot, however, be renamed as files or text elements, and files and library
elements of the basic types D, J, M, P, S, X cannot be renamed as elements of the
basic type L. Renaming an element of the basic type X as a text element (basic type
D, J, M, P or S) can corrupt the contents. If the target is simultaneously opened in
write mode in BS2000 or another bs2fs file system, rename() is rejected with EAGAIN.
rmdir()
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
This function is always rejected with ENOSYS.
fstat(), stat()
If stat() or fstat() refers to a file in a bs2fs file system, the information supplied
depends on whether or not the file is open. If the file is open, the file size precise to
the byte is supplied, and the time stamps are supplied precise to the microsecond,
as is customary for POSIX files. If the file is not open, stat() supplies the size of the
file as specified in BS2000 (i.e. in multiples of 2K blocks), and the time stamps are
only precise to the second; the fields with the microsecond details then always
contain 0.
symlink()
This function is always rejected with ENOSYS.
U41802-J-Z125-2-76
77
Programming notes
Interfaces for supporting BS2000 file systems
utime(), utimes()
These functions are accepted only at file and element level. In all other cases they
are rejected with EISDIR. The st_ctime field for the file status is not changed (cf.
chmod). Maintaining the access time stamps (access time st_atime) of elements is
a special library property which can be set with LMS or LMSUP. By default no
access time stamps are recorded for elements; stat() then outputs the library’s
access time stamp. The time at which the time stamps (access time st_atime,
modification time st_mtime) for directories (root directory of a bs2fs file system,
directories at library and type level) are updated is undefined. Only the current time
stamp can be set. If the time stamp specified differs from the current time stamp by
more than +/-3600 seconds, the call is ignored.
78
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k05
12. March 2014 Stand 08:57.50
5 Access to bs2fs file systems via NFS
This chapter describes the measures required to enable bs2fs file systems to be accessed
via NFS and what special features need to be taken into account for this access.
5.1 Releasing bs2fs files
bs2fs files are released with the NFS command share:
Syntax
share -F nfs [-o spec_options] [-d description] path name
The following special features compared to the release of ufs files apply here:
–
The path name specified must be the root directory of the bs2fs file system. It may not
point to a subdirectory in this file system (PLAM library or member type in a PLAM
library).
–
All client computers which are to be assigned access to the released bs2fs file system
must be specified explicitly with one of the options rw=client[:client]... (for read/write
access) or ro=client[:client]... (for read access).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The following spec_options have also been introduced for releasing bs2fs files:
bs2anon=bs2000_uid
Specifies the BS2000 user ID under which “anonymous” accesses are to take
place. Client processes with a UID other than the POSIX UID of the bs2fs owner
then have access to the bs2fs file system only if the bs2anon option is used. This
option enables the POSIX administrator to define precisely one BS2000 ID with
whose rights these client processes access the files of the bs2fs file system
concerned.
U41802-J-Z125-2-76
79
Mounting released bs2fs files
bs2conv
Access to bs2fs file systems via NFS
Specifies that character set conversion is to take place. This option has no
arguments. If the option is specified, file contents are converted from
EBCDIC.DF.04-1 to ISO 8859-1 when reading and in the other direction when
writing.
i
This option should always be specified when bs2fs file systems which
are mounted with ftyp=text or ftyp=textbin are released except for
clients which process EBCDIC files. If a client is to process ASCII files
on a bs2fs file system without converting them, the file system must be
mounted with ftyp=bin, otherwise conflicts will arise between ASCII new
ilnes (0x0A) and EBCDIC new lines (0x15).
bs2nameconv
Specifies that file name conversion is to take place. This option has no
arguments.
When this option is specified, specific file names are converted (see section
“Converting file names” on page 82).
5.2 Mounting released bs2fs files
Released bs2fs file systems are mounted on the NFS client with the means made available
by the client operating system. As a rule the mount command is used. Details can be found
in the NFS documentation for the relevant client operating system. For information on
BS2000 clients, see the “NFS (BS2000/OSD)” manual [4].
Here there are no special features compared with the procedure for mounting ufs files.
However, please note the information in section “Recommendations for mounting bs2fs file
systems on NFS clients” on page 84.
80
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k05
Access to bs2fs file systems via NFS
Security aspects and access rights
5.3 Security aspects and access rights
To ensure that the NFS clients are not inadvertently granted access rights which undermine
the BS2000 security strategies, the following restrictions for the release of bs2fs files apply
in comparison with the release of ufs files:
–
Releasing bs2fs files using the share command is permissible only under the BS2000
user ID TSOS.
–
The administration must define all client computers which are to be granted access to
the released bs2fs file system explicitly with one of the options rw=... (for read/write
access) or ro=... (for read access). Only these clients will be granted access. If a client
name is contained in both lists, it will only be granted read access. If it is contained in
neither, it is granted no access to the released bs2fs file system.
–
Client processes which run under the UID which matches the POSIX UID of the owner
(BS2000 user ID) of the bs2fs files are granted access to the released bs2fs files with
the rights of their owner.
For client processes which run under a different UID, the bs2anon option of the share
command can be used to define a BS2000 user ID under which these processes can
access the released bs2fs file system. If this option is not specified, these processes
are not granted access to the bs2fs file system.
The administrator of the NFS server (BS2000) must verify the trustworthiness of the
admitted clients – as is also the case with other NFS releases. In this case the client
administrator (root) must be trusted, because this administrator assigns a user his/her
UID. Client users with a UID which corresponds to the POSIX UID of the owner of the
bs2fs files must also be trusted in the context of the access rights granted to them.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
NFS clients can thus either perform access under the BS2000 user ID of the owner of
the released bs2fs files or – if specified in the share command – under the BS2000 user
ID assigned for “anonymous” accesses.
U41802-J-Z125-2-76
81
Converting file names
Access to bs2fs file systems via NFS
5.4 Converting file names
The namespace of BS2000 files and PLAM library members is as a rule smaller than that
of the client system. On the client system applications can therefore exist which cannot or
can only with difficulty cope with the restrictions to the namespace which apply on the
server. For example, editors create backup copies or buffer information under file names
which are not permitted on the bs2fs file system on the server.
When files which have not yet been opened are deleted, NFS clients temporarily create files
whose names consist of “.nfs” followed by varying numbers of hexadecimal digits. Consequently these file names are always converted as follows:
File name on the NFS client
File name in the bs2fs file system
.nfs[:xdigit:][:xdigit:]*
@nfs[:xdigit:][:xdigit:]*
When a bs2fs file system is released, the bs2nameconv option of the share command can be
used to specify that additional conversions should take place:
82
File name on
the NFS client
File name in the
bs2fs file system
Description of the file name on the NFS client
and of the conversion rule
\.[^.].*
@[^.].*
File names containing at least two characters, the first
character being a dot and the second character not a dot:
the leading dot (.) is converted to an at character (@).
..*~.*
..*#.*
File names which contain a tilde in addition to the first
character:
each tilde (~) is converted to a hash (#).
U41802-J-Z125-2-76
Further special features
5.5 Further special features when working with bs2fs files on
the NFS client
The following special features must be borne in mind when processing bs2fs files systems
which are mounted on NFS clients:
●
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k05
Access to bs2fs file systems via NFS
When files or library members are created, copied or renamed, the relevant naming
regulations of BS2000 must be observed, otherwise the error number EINVAL (“invalid
argument”) will be returned to the application when open() takes place. This can, for
instance, lead to problems in the case of:
–
editors which automatically generate copies of the original file or save intermediate
results (.filename.swp, filename~, etc.)
–
file browsers which generate thumbnails of files and store these temporarily in the
relevant directory
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
As a rule this can be configured in the editors or browsers either by disabling the
function concerned or by selecting a storage location outside the bs2fs file system.
●
In order to optimize internal performance, the closing of a bs2fs file on the NFS server
is slightly delayed (by 2 seconds) if there is no shortage of space in the bs2fs container.
As a result, a file on the NFS server remains open for 2 seconds and from the viewpoint
of BS2000 remains locked for this time although, from the viewpoint of the NFS client,
it already appears to have been closed.
●
The first open() of the bs2fs file as a rule takes place when the client application calls
the access() function. This can result in the response time of the NFS function ACCESS
being unexpectedly long. The response time in particular also depends on the size of
the bs2fs file (see also section “Recommendations for mounting bs2fs file systems on
NFS clients” on page 84).
●
On UNIX file systems, the hard link count of directories is the same as the number of
subdirectories plus two (because of the pseudo directories "." and ".."). In bs2fs file
systems the mountpoint and also PLAM libraries can contain subdirectories. For them,
the hard link count is set to this value only when the directory contents are read (readdir
function; used, for example, by the ls command). Before this the count is 2 or the value
which the last call of the readdir function supplied.
In some UNIX applications this leads to problems. The find command on LINUX clients
(GNU variant) can, for example, under some circumstances issue a warning that the
hard link count of a directory is incorrect and therefore internal optimization algorithms
do not function correctly and suggests that this optimization should be disabled.
U41802-J-Z125-2-76
83
Recommendations
●
Access to bs2fs file systems via NFS
The file type detection on UNIX systems as a rule takes place by means of a heuristic
examination of a small part of the file contents. For this purpose the file is opened, a
specific quantity of data is read and analyzed, and the file is closed again. In the case
of bs2fs files the entire BS2000 file must first be copied into the bs2fs container on the
NFS server to permit this, and the larger the file is the longer this process takes.
This can result in very long wait times on the NFS client which depend on the total size
of the files in this directory.
5.6 Recommendations for mounting bs2fs file systems on NFS
clients
The following measures may be required on the NFS clients on which bs2fs file systems
are mounted:
84
●
Disable the attribute cache
As the size of a bs2fs file can change when the first open() takes place, the file attributes
on the NHS client should not be stored in the attribute cache. On NFS clients the mount
option noac, for example, is available to prevent this.
●
Permit increased response times
Because of the unexpectedly long response times, especially with the NFS function
ACCESS, the timeout values specified on the NFS client must be sufficiently high. On
NFS clients the mount option timeo=n, for example, is provided for this purpose.
Depending on the size of the bs2fs files, ten times the default value or even more must
be set here.
If relatively large bs2fs files are to be processed, it may be necessary to perform a socalled "hard mount" (as a rule by omitting the mount option soft). As a result of this file
accesses are always repeated after timeouts.
●
Disable the READDIRPLUS function
Using the NFS function READDIRPLUS instead of READDIR is designed to reduce the
number of network I/Os and consequently to enhance the performance. However, a
loss of performance may under some circumstances be recognized on some NFS
clients. These offer, for example, a mount option nordirplus in order to disable the use of
READDIRPLUS.
●
Disable input/output buffering
Some NFS clients use local buffering of the input/output data to reduce the number of
network I/Os. This can, however, lead to possible input/output errors in bs2fs files being
concealed, as a result of which the application on the NFS client incorrectly assumes
that an input/output was successful. Input/output buffering on the NFS client should
therefore be disabled. On NFS clients the mount option forcedirectio, for example, is
available for this purpose.
U41802-J-Z125-2-76
12. March 2014 Stand 08:57.50
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k05
Access to bs2fs file systems via NFS
Examples
5.7 Examples
The following prerequisite applies for the examples in this section: The BS2000 IDs
MORITZ (POSIX UID=110), GAST (POSIX UID=100) and TSOS (POSIX UID=0)exist on
the NFS server.
Example 1
The :DATA:$MORITZ.*.LIB files are mounted as a bs2fs file system. This file system is
released for reading and writing for the clients client01 and client02 and only for reading for
the client client03. On the clients a user with the UID 110 is granted access with the rights
of the BS2000 ID MORITZ. Every other user is denied access because the bs2anon option
is not specified:
mount -F bs2fs -o rw,ftyp=binary ':DATA:$MORITZ.*.LIB'
share -F nfs
-o rw=client01:client02,ro=client03
/bs2mnt/moritz_lib
/bs2mnt/moritz_lib
Example 2
The :DATA:$ MORITZ.*.SRC files are mounted as a BS2000 file system. This file system is
released for reading and writing for the client client04 and only for reading for the client
client05. On the clients the user with UID=110 is granted access with the rights of the
BS2000 ID MORITZ and all others are granted access with the rights of the BS2000 ID
GAST:
mount -F bs2fs -o rw,ftyp=text ':DATA:$MORITZ.*.SRC'
/bs2mnt/moritz_src
share -F nfs
-o rw=client04,ro=client05,bs2anon=GAST /bs2mnt/moritz_src
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Example 3
The :HOME:$TSOS.SYSDAT.BCAM.* files are released for reading and writing for the super
user (UID=0) on the client client06; the other users connected to this client are not granted
access:
mount -F bs2fs -o rw,ftyp=text ':HOME:$TSOS.SYSDAT.BCAM.*' /bs2mnt/bcam.dat
share -F nfs
-o rw=client06
/bs2mnt/bcam.dat
Example 4
The :HOME:$SYSAUDIT.*CONSLOG* files are released for reading for any user connected to
client client07, the file contents being converted from EBCDIC.DF.04-1 to ISO 8859-1. In
other words the files contain EBCDIC texts and are output on the NFS clients as ASCII
texts:
mount -F bs2fs -o ro,ftyp=text ':HOME:$SYSAUDIT.*CONSLOG*'
share -F nfs
-o ro=client07,bs2anon=SYSAUDIT,bs2conv
U41802-J-Z125-2-76
/bs2mnt/conslog
/bs2mnt/conslog
85
Examples
Access to bs2fs file systems via NFS
Example 5
The :HOME:$MAX.* files are released for reading for every user connected to client client08,
provided the access rights permit this, the file contents being converted from
EBCDIC.DF.04-1 to ISO 8859-1. In other words the files contain EBCDIC texts and are
output on the NFS clients as ASCII texts. In addition, file names with specific special
characters are converted:
mount -F bs2fs -o ro,ftyp=text ':HOME:$MAX.*'
share -F nfs
-o ro=client08,bs2conv,bs2nameconv
86
/bs2mnt/max
/bs2mnt/max
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
14. November 2012 Stand 10:43.36
6 Diagnosis and enhancing performance
This chapter provides an insight into the internal procedures and describes a diagnostic aid
measures for enhancing performance.
The technical details described in this chapter are designed to facilitate diagnosis for an
experience system administrator.
6.1 Overview of the mount and unmount operations
This section describes the internal procedures for mounting and unmounting the bs2fs
container and bs2fs file systems.
Mounting the bs2fs container
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
Mounting the ufs file system with the property bs2fs container (mount -o bs2fscontainer...) is
the prerequisite for mounting bs2fs file systems. The bs2fs container is the physical
medium in which the files of bs2fs file systems are stored temporarily during the time they
are edited. When the bs2fs container is mounted, it is expected to be empty. If this is not
the case because of an error in a previous session, the content is deleted when the bs2fs
container is mounted.
A ufs file system can be mounted as these bs2fs container only if it was flagged appropriately when it was created (POSIX installation program). This prevents a ufs file system
which was not intended for use as the bs2fs container from being mounted inadvertently as
the bs2fs container, which would delete its contents. As additional protection you are
recommended as far as possible to mount the bs2fs container automatically.
The bs2fs container should be invisible for the users. Only indirect access to the contents
of the bs2fs container should be possible for them, i.e. by accessing the files in the mounted
bs2fs file systems. These are redirected internally to the files in the bs2fs container.
Direct access to the bs2fs container should only be possible for the system administrator
and solely for diagnostic purposes.
U41802-J-Z125-2-76
87
Overview of the mount and unmount operations
Diagnosis and enhancing performance
Unmounting the bs2fs container
The bs2fs container can only be unmounted if no bs2fs file systems are still mounted.
Mounting a bs2fs file system
When a bs2fs file system is mounted, a directory with the name catid.userid.nummer is
created in the bs2fs container, where:
catid
catalog ID from the resource operand of the mount command
userid
user ID from the resource operand of the mount command
nummer
number with which the mount operations for bs2fs file systems are numbered in
chronological order within a session
This directory is used to contain maps of objects (file, library, element type or library
element) of a bs2fs file system if these have been opened for editing with open() or opendir().
These maps in the bs2fs container are always single files, even if the corresponding object
in the bs2fs file system is presented as a directory. The mounted BS2000 files are therefore
not stored physically at the mount point of the bs2fs file system but are located (“invisibly”
for the user) in a directory of the bs2fs container reserved for this bs2fs file system.
In addition, directories also exist on the uppermost level of a bs2fs file system. Their map
files which are created with open or opendir are views of these directories which are local to
the process and are called %userid%pid, where userid is the BS2000 user ID (8 characters
with blanks) of the caller of open or opendir and pid is their process ID, specifically, for
example, "%BACH
%00123".
88
U41802-J-Z125-2-76
14. November 2012 Stand 10:43.36
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
Diagnosis and enhancing performance
Overview of the mount and unmount operations
The table below shows how the names of the objects in the bs2fs file system are mapped
to file names in the bs2fs container:
Type in BS2000
Name in the bs2fs
file system
Type in the
bs2fs file
system
Name of the file in the
bs2fs container
DVS file
<file>
File
<file>
PLAM library
<lib>
Directory
<lib>%
Element type
<lib>/<type>
Directory
<lib>%<type>%
Library element
<lib>/<type>/<memb>
File
<lib>%<type>%<memb>+<ver>
where
<file>
<lib>
<type>
<memb>
<ver>
File name
Library name
Element type
Element name
Element version
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
The table below shows this mapping again for specific examples:
Object that is opened with open() or
opendir()
Name of the
bs2fs file system
File name in the
bs2fs-Container
DVS file SOURCEFILE
sourcefile
sourcefile
PLAM library MYPLAMLIB
myplamlib
myplamlib%
Type S subdirectory of the PLAM library
MYPLAMLIB
myplamlib/s
myplamlib%s%
Type S element SOURCE1 of the PLAM myplamlib/s/source1
library MYPLAMLIB (where the highest
element version is Version 001)
myplamlib%s%source1+001
The map files in the bs2fs container are created with first open() or opendir() for the relevant
object and are deleted again with the last close() or closedir(). No map files exist in the bs2fs
container for files or directories which are accessed in a different manner.
Unmounting a bs2fs file system
When a bs2fs file system is unmounted, the corresponding directory in the bs2fs container
is deleted.
U41802-J-Z125-2-76
89
Diagnosis
Diagnosis and enhancing performance
6.2 Diagnosis
In the event of an error the information regarding which files stored in the bs2fs container
correspond to which file system may be required. The get_container_index command is
provided to ascertain this assignment.
get_container_index
assign bs2fs container index to bs2fs file system
This command provides the administrator with the assignment of the index of a directory in
the bs2fs container to the corresponding bs2fs file system.
Syntax
get_container_index [ mountpoint]
No option specified
Output takes place for all mounted bs2fs file systems.
mountpoint
Output takes place only for the file system to which the specified file belongs.
Example 1 Listing the information about all the mounted bs2fs file systems
# get_container_index
index = 3 for /home/bach/bs2.2 (:V70A:$BACH.PLAMLIB*)
index = 2 for /home/bs2.2 (:V70A:$BACH.CCC.*.C)
index = 1 for /home/bach/bs2.1 (:V70A:$BACH.ASS.*.S)
Example 2 Listing the information about a particular bs2fs file system
# get_container_index /home/bs2.2
index = 2 for /home/bs2.2
90
U41802-J-Z125-2-76
14. November 2012 Stand 10:43.36
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
Diagnosis and enhancing performance
Correcting errors
6.3 Correcting errors
When a file in a bs2fs file system has been modified and cannot be copied back to BS2000
again after processing has been completed, the ufs file concerned is moved to a special
bs2fs_lost+found directory in the bs2fs container. If a file which has the same name and
stems from an earlier failed attempt to write it back exists there, it is overwritten without an
inquiry.
The user can issue the bs2fs_recover command to obtain a list of files which have been relocated in this way and then (for example with a modified name) copy these to BS2000
and/or delete them in the bs2fs container.
Alternatively this task can also be performed by the system administrator in the role of a
deputy.
To prevent files which have been relocated to the bs2fs_lost+found area from remaining unnoticed, when the bs2fs container is mounted a message is output on the console which
provides information on the user IDs for which files exist in this area. The bs2fs_lost+found
area is of course not deleted like the rest of the bs2fs container when the latter is mounted.
The bs2fs_recover command provides a simple way to check whether files of a specific ID
exist in the bs2fs_lost+found area. When intensive use is made of the bs2fs file system, it
is recommendable to configure an automatic check with this command, e.g. in the logon
script ($HOME/.profile or /etc/profile).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
In order to get to know how the command works and the mechanisms associated with it,
you can use the POSX file /etc/.bs2fsdSimulateCloseError to simulate errors which occur
when the bs2fs files are written back. If this file (which requires no content) exists, the ufs
file is relocated to the bs2fs_lost+found area irrespective of the success of its being written
back to BS2000. If the /etc/.bs2fsdSimulateCloseError file is removed or renamed, the simulation will be canceled.
U41802-J-Z125-2-76
91
Correcting errors
Diagnosis and enhancing performance
bs2fs_recover
Performing recovery for the bs2fs_lost+found area
This command offers the following functions:
Syntax
–
Listing the files in the bs2fs_lost+found area
–
Outputting the number of these files
–
Generating a shell script which copies the files from the bs2fs_lost+found area to
BS2000 and/or deletes them from this area
–
Direct copying and/or deletion of files in the bs2fs_lost+found area, either in interactive
mode with an inquiry or in an automatic mode with default settings.
Format 1: bs2fs_recover [-l] [ -m level]
[ -a after] [ -b before] [ -u user|*ALL] [ -v] [ file selection]
Format 2: bs2fs_recover {-g | -x} [ -d] [ -w] [ -c] [ -f n|y] [ -p prefix] [ -s suffix]
[ -a after] [ -b before] [ -u user|*ALL] [ -v] [ file selection]
Main options
The main options define which function is to be executed. Only one main option may be
specified.
No main option specified
All selected files are listed, which corresponds to main option -l.
92
-l
The number or a list of selected files is output to the default output. The -m option
defines the scope of the output.
-g
A shell script is generated which performs the actions specified with the -d und -w
options when it is executed. The generated shell script is output to the default output.
-x
The actions specified with the -d und -w options are performed immediately. With
these actions an output is by default only made in the event of an error. If you want
to see the progress of processing, you should specify the -v option at least once so
that the file name concerned is listed before the actions for this file are executed.
U41802-J-Z125-2-76
Correcting errors
Additional options
The additional options are used to modify the action concerned or for file selection.
-m level
This option controls the scope of the output when the -l option is specified. If the -m
option is not specified, -m 1 is assumed.
It is obligatory to specify level. The following options can be specified:
14. November 2012 Stand 10:43.36
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
Diagnosis and enhancing performance
0
The number of files found is output to the default output in a line with the format
“Number file(s)”. in this case the end status also supplies information on whether
at least one file was found (end status = 0) or not (end status = 1). This provides
a simple way of checking whether the bs2fs_lost+found area contains any of
your own files.
1
A list of files is output to the default output. This contains the date and time of
the last modification, the file size and the name of the associated BS2000 file or
the complete specification of the corresponding library member (library
name(member name,type,version)). The name is prefixed by the serial number of
the bs2fs mount, separated by a “/”.
2
Output as for -m 1 with the following additional information:
File processing mode of the bs2fs mount T, TB, B (for ftyp text, textbin or binary)
and absolute ufs file name in the bs2fs container (bs2fs_lost+found area).
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
3
Output as for -m 2 with the following additional information:
Access method (PAM, SAM, ISAM or PLAM), block format, block length, record
format and record length.
If certain information, such as record foramt, does not exist for PAM files, ’-’ is
output instead. You can use this information to write your own scripts for copying and deleting the files.
Example of an output with -m 3:
2012-03-20 09:39:06
8
3/:90GB:$TSOS.TSOSDAT T
/grossercont/bs2fs_lost+found/TSOS/:90GB:3:TD01SV00000:tsosdat SAM DATA
(STD,1) V -
-d
In conjunction with the main options -g and -x: The selected files are deleted from
the bs2fs container area bs2fs_lost+found. This option can also be used together
with the -w option.
-w
In conjunction with the main options -g and -x: The selected files are copied to
BS2000. This option can also be used together with the -d option.
U41802-J-Z125-2-76
93
Correcting errors
Diagnosis and enhancing performance
-c
Suppresses the use of the catalog ID and user ID employed to date when copying
to BS2000 (-w option). This option can be used (if necessary together with the -p
option) to perform backup under a different catalog ID/user ID from the one specified in bs2fs-mount.
-f n|y
Determines whether existing BS2000 files and LMS members are overwritten when
copying to BS2000 (-w option).
Depending on the main option, the -f option has the following effects:
–
Main option -g (Generating a shell script)
When the -w option is specified, the OV variable is always set in the script. This
controls the behavior when copying to BS2000 while the generated script is executed.
When the -f option is not specified or -f n is specified, the OV variable is assigned the value “NR”, as a result of which BS2000 files are not overwritten
when copying takes place. When -f y (OV=”Y”) is specified, they are overwritten.
When the -w and -d options are specified simultaneously, files are deleted only
if the preceding copy operation was successful either because the file did not
yet exist in BS2000 or because it was possible to overwrite it successfully.
If the script is generated only for the purpose of deletion (-d option), the OV variable is not set. The files in the Lost+Found area are then deleted without an inquiry irrespective of the -f option.
–
Main option -x (Immediate deletion or copying)
The use of the -f option decides whether an inquiry should be made before each
action takes place (interactive mode) or not (automatic mode).
When the -f option is not specified, the following inquiries are made:
1. Should the file be copied
(with the -w option)
2. Should the file be overwritten
(with the -w option if the target file already exists)
3. Should the file be deleted from the Lost+Found area
(with the -d option).
If one of these inquiries is answered with “no”, no further inquiries or actions
take place for the same file. In particular, a file is not deleted if copying or
overwriting was rejected beforehand.
When the -f y or -f n option is specified, no inquiries are made either when copying (-w) or when deleting (-d). When copying, depending on the -f option existing
files are either overwritten (-f y) or not (-f n).
94
U41802-J-Z125-2-76
© cognitas GmbH 2001-2007
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
Correcting errors
-p prefix
Specifies a string with which the BS2000 file name is prefixed when copying to
BS2000 (-w option). If prefix contains a catalog ID and/or a user ID, the -c option
must also be specified to permit a valid BS2000 file name to be created.
-s suffix
Specifies a string which is appended to the BS2000 file name when copying to
BS2000 (-w option).
-a after
The file selection is restricted to files which were last modified after the time specified with after .
14. November 2012 Stand 10:43.36
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
Diagnosis and enhancing performance
The time is specified as follows:
[[YY]yy]MMDDhhmm[.SS]
[YY]yy
Year (two or four characters)
A two-character year specification is complemented as following:
– yy > 68 means 19yy
– yy Î 68 means 20yy
If no year is specified, the current year is assumed.
MM
Month, two characters (01 through 12)
DD
Day, two characters (01 through 31)
hh
Hour, two characters (00 through 23)
mm
Minute, two characters (00 through 59)
SS
Second, two characters (00 through 61)
The values 60 and 61 are provided for leap seconds.
If second is not specified, 0 is assumed
-b before
The file selection is restricted to files which were last modified before the time specified with before. The format of the time specification is described under the -a option.
U41802-J-Z125-2-76
95
Correcting errors
Diagnosis and enhancing performance
-u user
The file selection is restricted to the files of the specified user ID. The entry for the
user ID is not case-sensitive. As a rule, because of the default attributes of the subdirectories in the Lost+Found area (drwx --- ---) owners of nonprivileged IDs can
only view and edit their own files.
-u user not specified:
The BS2000 user ID of the caller is used.
*ALL (not case-sensitive)
All BS2000 IDs are processed for which files are contained in the Lost+Found
area. As a rule this specification is only relevant for the system administration
TSOS ID) as only the system administration has the necessary rights to copy
and delete files in foreign IDs.
-v
Additional information on the actions of the tool is output. This option can also be
specified more than once (up to three times) in order to increase the degree of detail
of the information.
file selection
file selection is used to control the selection of the files located in the Lost+Found
area (in addition to the selection with the -u option).
Sample printouts can be used like those which are offered by the shell for creating
file names. These printouts are then used for the BS2000 file names and LMS
member specifications, as the tool outputs them when the -l option is used.
Example:
In the following output with -l -m1 the file name printed in bold is relevant for the
sample printout.
2012-03-20 09:39:06
8
3/:90GB:$TSOS.TSOSDAT
In this case the complete file name (without sample printout) would also have
to include the prefix 3/.
file selection not specified:
All existing files of the calling user.
96
U41802-J-Z125-2-76
Enhancing performance
6.4 Measures to enhance performance
●
14. November 2012 Stand 10:43.36
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.k06
Diagnosis and enhancing performance
–
The wildcard in the resource option should be selected in such a manner that if
possible only the BS2000 files which are actually needed are mounted as the
number of mounted files influences the performance.
–
If it is syntactically not possible or only extremely difficult to separate the required
files from those which are not required, one of the following options for bypassing
this problem should be considered before mounting files for the first time: export the
files required to another pubset (under the same or a different user ID) or rename
these BS2000 files with meaningful names.
–
You should avoid the same BS2000 files being mounted several times at different
positions. This occurs when multiple mount operations are executed with identical
or overlapping resource specifications. If such a file is open for write accesses in one
of the bs2fs file systems involved, it cannot be opened simultaneously in another
bs2fs file system.
If required, start additional copy daemons with the start_bs2fsd command.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
●
Check whether the following points were taken into account for the selection of BS2000
files which are to be made available in a bs2fs file system:
U41802-J-Z125-2-76
97
Enhancing performance
98
Diagnosis and enhancing performance
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.mix
9. November 2012 Stand 13:23.31
Glossary
bs2fs container
A file system of the type ufs which is used solely to accommodate files of bs2fs file
systems temporarily. These files are then stored in directories (one directory for each
mounted bs2fs file system) in the bs2fs container.
bs2fs file system
Selectable set of files in BS2000 which are made available in POSIX as a file system,
thus enabling them to be accessed using POSIX means (commands, program interfaces). The files are selected using the user and catalog IDs and wildcard symbols.
copy daemon
System process which copies from BS2000 to the bs2fs container and back again.
daemon
Daemons are system processes which run permanently and normally in background
mode, and which perform general tasks.
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
© cognitas GmbH 2001-2007
directory
A directory is used to group and organize files and subordinate directories of a hierarchical file systems.
file system
A file system is a hierarchical group of directories and files which are located physically
on the same storage medium, e.g. in a partition or in a container file.
The term is used for organizational structures of files, such as UNIX file system, POSIX
file system, hierarchical file system, other BS2000 file systems (DMS and LMS), for
example.
U41802-J-Z125-2-76
99
Glossary
file system type
Type of a file system in the file tree of POSIX or of a UNIX computer. The most familiar
types are as follows:
Type bs2fs: BS2000 files which are made available under POSIX.
Type ufs: local file systems containing user data.
Type nfs: file systems which are located physically on remote computers.
Examples of further types: fdfs, proc, bfs, rfs and s5.
All types of file systems used in a system must be known when the system is
configured; no new file system type can be added later. Information about the file
system types which can be used in a system is provided by the sysfs() interface.
file tree
Overall hierarchy of the files on a UNIX computer or in POSIX. The UNIX or POSIX file
hierarchy is organized on the basis of a tree structure. The root of the file tree is the root
directory (/). All other directories are branches which emanate from the root. The files
are the leaves of the tree. Each file is accessible via precisely one path of the file tree.
mounting
Local resources are made accessible in a local file system using the mount command;
this is also referred to as mounting.
POSIX file system
File system on a BS2000 computer running POSIX. The POSIX file system
corresponds to a UNIX file system.
root directory
The file system at which the file tree begins. The root directory is represented by the
slash (/).
transparent file access
The user can access bs2fs files like ufs files (with the few restrictions described) using
the same access functions or commands.
ufs (UNIX file system)
The ufs file system is the local standard file system in POSIX. It is implemented by a
BS2000 file (of the type PLAM) in which the files of the file system are stored as on a
UNIX disk.
100
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.lit
9. November 2012 Stand 13:23.32
© cognitas GmbH 2001-2007
Dokuschablonen 19x24 Version 7.32de für FrameMaker V7.x vom 28.03.2007
Related publications
You will find the manuals on the internet at http://manuals.ts.fujitsu.com. You can order
manuals which are also available in printed form at http://manualshop.ts.fujitsu.com.
[1]
POSIX (BS2000/OSD)
Basics for Users and System Administrators
User Guide
[2]
POSIX (BS2000/OSD)
Commands
User Guide
[3]
C Library Functions (BS2000/OSD)
for POSIX Applications
Reference Manual
[4]
NFS (BS2000/OSD)
Network File System
User Guide
[5]
BS2000/OSD-BC
Introductory Guide to Systems Support
User Guide
[6]
BS2000/OSD-BC
Commands
User Guide
[7]
EDT (BS2000/OSD)
Statements
User Guide
[8]
EDT (BS2000/OSD)
Unicode Mode Statements
User Guide
U41802-J-Z125-2-76
101
Related publications
102
U41802-J-Z125-2-76
Pfad: P:\FTS-BS\BS2-GA\OSD-V90\9999999_Einleitungen\POSIX\bs2fs_e\bs2fs_e.six
12. March 2014 Stand 09:04.09
© cognitas GmbH 2001-2010
Dokuschablonen 19x24 Version 7.4us für FrameMaker V7.x vom 09.02.2010
Index
D
daemons 61
copy daemon bs2fsd 61
monitoring 61
DMS files 33
dumpfs (POSIX command) 67
/etc/mnttab 63
/etc/vfstab 65
A
access rights 39
administration files
/etc/mnttab 63
/etc/vfstab 65
editing 62
overview 62
application scenarios 24
automating
mount operations 23
automation
mount operations 29
B
bs2fs
managing 27
bs2fs container
mounting 87
mounting, example 17
unmounting 88
bs2fs file system 88
application scenarios 24
mounting, example 18
bs2fs_lost+found directory 91
bs2fs_recover (POSIX command)
bs2fsd (daemon) 61
C
check file system (fsck) 67
commands, overview 47
copy daemon bs2fsd 61
U41802-J-Z125-2-76
E
element types
elements 34
34
F
file names, conversion of 82
file system
display information (dumpfs) 67
mount (mount) 48
mount (mountall) 55
mount (NFS) 80
unmount (umount) 59
unmount (umountall) 60
file system check (fsck) 67
fsck (POSIX command) 67
fsexpand (POSIX command) 68
G
get_container_index (POSIX command)
92
L
library elements
90, 92
34
M
mount 88
automating 23, 29
bs2fs container 87
bs2fs container, example
bs2fs file system 88
17
103
Index
U
umount (POSIX command) 59
umountall (POSIX command) 60
unmount
bs2fs container 88
bs2fs file system 89
file system (umount) 59
remote resource (umount) 59
several file systems (umountall)
user-specific program caches
setting up/managing 68
mount (continued)
bs2fs file system, example 18
bs2fs file system, NFS 80
file system (mount) 48
file system (mountall) 55
remote resource (mount) 48
mount (POSIX command) 48
mount operation 87
mountall (POSIX command) 55
N
naming conventions 36
NFS command share 79
NFS, access via 79
notational conventions 10
60
P
path name
checking (pathchk) 68
pathchk (POSIX command) 68
pdbl 68
PLAM libraries 34
POSIX commands
pdbl 68
R
Readme file 8
representation
files in the bs2fs file system
resource
mount 48
unmount (umount) 59
13
S
share (NFS command) 79
show_pubset_export (POSIX command)
start_bs2fsd (POSIX command) 58
56
T
table of all defined fils systems 65
table of mounted file systems 63
104
U41802-J-Z125-2-76
Download PDF
Similar pages