Storing Database Files on ACFS on ODA
Frequently Asked Questions
Storing Database Files in ACFS on Oracle Database Appliance
Comprehensive internal benchmarks using OLTP and DSS
Oracle ASM Cluster File System (ACFS) provides
workloads have proven the performance of Oracle ACFS and
unprecedented simplicity, automation and functionality
ASM to be similar. The charts below reflect ACFS vs. ASM
performance for some OLTP workloads. Refer to the “Benefits
in storage and file management for all file types
of Oracle ACFS” white paper for more details on how ACFS
including Oracle database and general-purpose files.
performs relative to ASM.
This FAQ provides more information about storing
The following chart shows performance based on OLTP
database files in ACFS on Oracle Database Appliance.
workload. As you can see, the throughput performance and
transaction response time are almost identical regardless of
whether the database use ACFS or ASM directly.
Oracle ACFS was chosen as the primary file system for
Oracle Database Appliance because it provides:
Increased functionality without requiring additional
Equivalent performance to Oracle ASM
Industry standard and simple user interface
Database snapshots to quickly and easily provision
test and development environments at a fraction of
the disk storage required with full database copies
Advanced functionality for general purpose files such
as replication, tagging, encryption, security, and
Oracle ACFS is designed to deliver best possible performance
with Oracle databases and general purpose files. Oracle ACFS
obtains an extent map list and disk list information from the
Oracle ASM instance whenever it opens an ASM file. This
means, neither Oracle ASM nor Oracle ACFS are in the I/O
path when accessing files on disk. Oracle ACFS caches ASM
disk mapping information and uses the mapping on
subsequent ACFS I/O operations to directly read and write to
ASM disk group disks.
In a virtualized Oracle Database Appliance, the storage is
directly attached to the VM hosting the databases
(ODA_Base). This provides I/O performance on par with a
bare metal deployment of Oracle Database Appliance for
databases in ODA_Base. The shared repository is created in
ODA_Base using ACFS, but is then exported to Dom0 and
then made available as virtual storage to individual VMs. It is
the additional overhead of these exports that impacts
Q: Doesn’t ACFS add additional management burden by
introducing a file system that requires maintenance?
Question and Answer
A: No. Appliance Manager (OAK) automatically manages the
file systems. When the first non-CDB database of version or greater is created, three default file systems
are created automatically in the Oracle Database
Appliance ASM disk groups (DATA, RECO, and REDO).
The associated database files for each non-CDB database
are created in these three ACFS file systems instead of in
the ASM disk groups directly. For each CDB database that
is created, three additional file systems in DATA, REDO,
RECO are created for the database files.
Q: Why did Oracle switch from ASM to ACFS for storing
database data files on Oracle Database Appliance?
Q: Does the DBA now have to manage database files stored
in ACFS?
A: ACFS is built on top of ASM, and provides all the benefits
of ASM plus the benefits of ACFS. These include spaceefficient snapshots of database files for test and
development. The latest releases of ACFS support direct
I/O, so there is no performance penalty. In other words,
ACFS provides all the benefits and more, without any cost.
A: ACFS in the Oracle Database Appliance requires no
administration. The Appliance Manager automatically
manages the storage; including ACFS file systems and the
underlying ASM diskgroups.
performance, not ACFS. This overhead is acceptable for
hosting VM Images and templates, as these operations do not
require high I/O performance. However, you should not
attempt running an I/O intensive workload (for example, a
database) against virtual storage based on the shared
repository. All databases should reside in ODA_Base.
Q: Why not give customers a choice of ASM or ACFS for
database files?
A: Oracle Database Appliance is an appliance, which means
we’ve put together a configuration that works best for most
customers, without tuning and tweaking. We test that
configuration extensively, to ensure users get the best and
most reliable experience. That requires choosing a single
configuration, so we can focus our development and
testing resources on that configuration. Offering a choice
of storage dilutes those development and testing efforts,
leading to a less functional and less reliable product for all.
Q: In the past, Oracle said ASM was better for database files
than a file system. Why now use a file system?
A: ASM was introduced with Oracle 10g to solve the storage
management problem for database files. It added features
that ensured there would be no I/O bottlenecks, that
storage was highly available and resilient, and could be
expanded or migrated without downtime. Existing file
systems could not provide these benefits. However, in
Oracle 11g Release 2, Oracle introduced the ASM Cluster
File System (ACFS), built on top of ASM. It inherits all the
benefits of ASM, and adds additional features and
functionality. With the recent addition of Direct I/O, it
provides features and performance comparable to ASM.
Q: Why are existing databases not moved to ACFS
automatically after an upgrade to Appliance
A: Existing databases after an upgrade to Appliance Manager
12.1.2 are left in ASM because it would take additional
downtime to migrate them to ACFS. Customers may
choose to migrate them to ACFS at a later time.
Q: How do I migrate a database from ASM or from another
server to ACFS on ODA?
A: You must first create the proper directories to receive the
database files before you can migrate them to ACFS on
Oracle Database Appliance. This will be automated in the
next release of the Database Appliance, but in the mean
time, the steps are documented in the white paper “Steps
to Migrate Non-CDB Databases to ACFS on Oracle
Database Appliance 12.1.2.”
Q: Can I leave my existing databases in ASM while creating
new database in ACFS?
A: Yes, both types of databases can coexist on the same
appliance. There is no need to move an existing database
to ACFS. However, if a database remains in ASM, you
cannot clone it using snapshot functionality.
Oracle Corporation, World Headquarters
Worldwide Inquiries
500 Oracle Parkway
Phone: +1.650.506.7000
Redwood Shores, CA 94065, USA
Fax: +1.650.506.7200
Copyright © 2014, Oracle and/or its affiliates. All rights reserved. This document is provided for information purposes only, and the
contents hereof are subject to change without notice. This document is not warranted to be error-free, nor subject to any other
warranties or conditions, whether expressed orally or implied in law, including implied warranties and conditions of merchantability or
fitness for a particular purpose. We specifically disclaim any liability with respect to this document, and no contractual obligations are
formed either directly or indirectly by this document. This document may not be reproduced or transmitted in any form or by any
means, electronic or mechanical, for any purpose, without our prior written permission.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and
are trademarks or registered trademarks of SPARC International, Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are
trademarks or registered trademarks of Advanced Micro Devices. UNIX is a registered trademark of The Open Group. 1014
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF