TIBCO ActiveSpaces® Transactions 2.4.1 Release Notes Welcome

TIBCO ActiveSpaces® Transactions 2.4.1 Release Notes
Welcome to TIBCO ActiveSpaces® Transactions
ActiveSpaces® Transactions is an in-memory transactional application server that provides scalable highperformance transaction processing with durable object management and replication. ActiveSpaces®
Transactions allows organizations to develop highly available, distributed, transactional applications
using a standard Java programming model.
TIBCO ActiveSpaces® Transactions provides these capabilities:
Transactions - high performance, distributed All-or-None ACID work.
In-Memory Durable Object Store - ultra low-latency transactional persistence.
Transactional High Availability - transparent memory-to-memory replication with instant fail-over and
fail-back.
Geographic Redundancy - Replicate transactions across a WAN with split brain avoidance and
reconciliation
Distributed Computing - location transparent objects and method invocation allowing transparent
horizontal scaling.
Data Grid - secondary object storage, adhoc queries and notification of transactional data changes.
Where to Find Information
See the Quick Start Guide in the TIBCO ActiveSpaces® Transactions SDK for quickly getting TIBCO
ActiveSpaces® Transactions up and running and configured with your IDE of choice.
The following information is also shipped in the SDK:
Installation Guide - Installation instructions
Architect's Guide - Architectural details
Java Developer's Guide - Programmer documentation
Administrators's Guide - Administration documentation
Performance Tuning Guide - Application performance tuning documentation
System Sizing Guide - Sizing machine resources for TIBCO ActiveSpaces® Transactions applications
API reference - javadoc for all APIs
Code snippets - source code for the programming snippets in the Developer's Guide
Examples - source code examples of integrating TIBCO ActiveSpaces® Transactions with common Java
technologies, application templates, and channels.
We encourage you to join the TIBCO ActiveSpaces® Transactions development community at DevZone
We look forward to hearing your feedback to help us improve TIBCO ActiveSpaces® Transactions to
meet your application needs.
Product Compatibility
The TIBCO ActiveSpaces® Transactions Data Grid feature is compatible with TIBCO ActiveSpaces®
Enterprise Edition 2.1.2 Hot Fix 14.
Supported Platforms
TIBCO ActiveSpaces® Transactions is supported on
Red Hat Enterprise Linux Server versions 5.8 to 6.5 on 64-bit x86 (x86_64) processors.
Oracle Solaris 10 1/13 in 64 bit mode on SPARC V9 architecture on both Sun and Fujitsu hardware.
Red Hat Platform Requirements
The compat-expat1 (Red Hat 6 only), pstack, gdb, and sysstat packages must be installed when running
on Red Hat. The rpm package manager can be used to query and verify that these packages are
installed.
rpm -q compat-expat1 - compat-expat1-1.95.8-8.el6.x86_64
rpm -q pstack - pstack-1.2-7.2.2
rpm -q gdb - gdb-7.0.1-42.el5
rpm -q sysstat - sysstat-7.0.2-11.el5
Red Hat 6.2, and later releases, ship with a new implementation of malloc that consumes more heap
memory by default. Details can be found here - 13.4. GNU C Library (glibc). Tuning the
MALLOC_ARENA_MAX environment variable can reduce memory allocation.
Solaris Platform Requirements
The sudo package must be installed when running on Solaris.
Supported Browsers
TIBCO ActiveSpaces® Administrator supports these browsers. The version number is the minmum
version supported.
Chrome 16
Firefox 24.6.0
Internet Explorer 9
Safari 5
Embedded JRE Version
TIBCO ActiveSpaces® Transactions ships with version 1.7.0_67 of the JRE.
Major Improvements
Security upgrades to address the POODLE security vulnerability.
OpenSSL, OpenLDAP, Apache, and PHP 3rd-party version upgrades.
String handling performance improvements.
On-going stability improvements.
Upgrading
The following changes must be made to upgrade from previous releases.
SSL3 support was disabled for all administration access - both command line and Web, to eliminate the
exposure to the POODLE exploit.
Deprecations
The following features were deprecated in this release and will be removed in a future release. Please
update your applications and development environment.
None
2.4.1 Resolved Issues
Bug
[FLUENCY-7008] - DIST: Deadlocks seen when define/enable of many partitions is done.
[FLUENCY-7014] - JMX: Improve getFreePort() allocation.
[FLUENCY-7019] - DIST: Failover can activate disabled partitions.
[FLUENCY-7021] - DIST: Crash in Partition cardinality() method.
[FLUENCY-7032] - RT: deadlock crash
[FLUENCY-7052] - ADMINISTRATOR: comand - Memory fault(coredump)
[FLUENCY-7082] - RT: Lock ordering issues with object services and distribution.
[FLUENCY-7137] - DIST: getPartitionStats() can fail if partition is not active.
[FLUENCY-7139] - ASTA: not updated for kabira/ast application name change
[FLUENCY-7160] - DIST: nonResponseTimeout processing is broke.
[FLUENCY-7163] - RT: SWSocket::completeConnect() can return SocketTimedOut.
[FLUENCY-7164] - DIST: Errors returned by getRemoteHostAddr() are asserted.
[FLUENCY-7172] - DIST: mutex deadlock in runtime
[FLUENCY-7221] - MONITOR: string.cpp:3389: Invalid pointer 'cString', address 0x0 is SW_HOME not set
[FLUENCY-7224] - RUNTIME: cannot start node with UTF-8 multibyte character as node name
[FLUENCY-7225] - DEPLOY: cannot deploy JVM with UTF-8 multi-byte characters in deploy options
[FLUENCY-7228] - ADMINISTRATOR: typo in error message - Not able to remove node diretory
[FLUENCY-7240] - RT: Leak of fd in SWSocket::networkSupported() on Linux
[FLUENCY-7249] - DIST: Race condition in reportLocalTransactions().
[FLUENCY-7275] - DIST: Failover check for method dispatches on partitioned instances is broke.
Documentation
[FLUENCY-7051] - DEVGUIDE: describe error handling for enable partition types
Improvement
[FLUENCY-6970] - RUNTIME: Improve Java String creation performance by directly encoding the char
array.
[FLUENCY-6980] - ASTA: disable Apache Track/Trace request
[FLUENCY-7040] - RT: improve detection of OSProxyObject lock ordering issues
[FLUENCY-7276] - PERFORMANCE: improve scaling of string copyin/copy-out
Task
[FLUENCY-6909] - 3RDPARTY: upgrade OpenSSL to at least 1.0.1j
[FLUENCY-7033] - DOCTOOLS: update Two-Second Advantage tag line on title page
[FLUENCY-7078] - 3RDPARTY: upgrade Apache to at least 2.2.29
[FLUENCY-7079] - 3RDPARTY: upgrade PHP to at least 5.5.18
[FLUENCY-7080] - 3RDPARTY: upgrade OpenLDAP to at least 2.4.40
[FLUENCY-7084] - 3RDPARTY: disable SSL3 protocol for Apache httpd (POODLE)
[FLUENCY-7085] - 3RDPARTY: disable SSL3 socket support (POODLE)
[FLUENCY-7088] - Doc Impact of FLUENCY-7085 3RDPARTY: disable SSL3 socket support (POODLE)
2.4.0 Resolved Issues
Bug
[FLUENCY-426] - DIST: Arrays of ManagedObject not supported as distirbuted parameter/return type
[FLUENCY-1977] - DIST: distribution switchadmin plugin displaystatistics
[FLUENCY-1978] - DIST: distribution switchadmin plugin enablestatistics and displaystatistics
[FLUENCY-1981] - PERF: visualvm profiling tools do not work
[FLUENCY-1982] - FLUENCY: visualvm cpu monitor does not work
[FLUENCY-2660] - ASTA: kabman test modifying distrib
[FLUENCY-3153] - KDM: intermittent failure in component notifier test
[FLUENCY-3608] - STATISTICS: partition statistics filter support breaks snapshot
[FLUENCY-3888] - STATISTICS: businessstatemachine statistics filter support breaks snapshot
[FLUENCY-3912] - CONFIGURATION: accept() doesn't throw IOException when run within configuration
notifier
[FLUENCY-4045] - HA: Determine how to handle concurrent define/enable of partitions.
[FLUENCY-4312] - DATAGRID: kabira-init stop does not stop as-agent
[FLUENCY-4322] - ADMINISTRATION: bogus error message for unknown command/target
[FLUENCY-4459] - SECURITY: switch core security unit test has dependency upon switchadmin unit test
[FLUENCY-4518] - DIST: failed C++ assertion 'm_psmenv->distribData[DistribTranMgr] ==
SWSMOffset::sNULL'
[FLUENCY-4685] - DIST: Type processing fails for discovered nodes.
[FLUENCY-4852] - JAVASE: running with debug generates warnings about signal handlers
[FLUENCY-4930] - ASTA: print statistics not working
[FLUENCY-4957] - CACHE: Missing unit tests for cache plugin.
[FLUENCY-5098] - DEPLOY: incorrect operator for string comparison in checkDistributedResult()
[FLUENCY-5113] - MONITOR: cannot display sequence using system decoding level
[FLUENCY-5328] - SNAPSHOT: event logs not captured if the Node Agent filepathTemplate was
reconfigured
[FLUENCY-5329] - Doc Impact of FLUENCY-4528 SECURITY: no mechanism to hide credentials when using
ASTA
[FLUENCY-5382] - SWDIAG: Crash when using displaying type descriptors.
[FLUENCY-5384] - RT: null values in auto-box arrays causes oscdr.cpp:498: failed C++ assertion
'SW_BOMB'
[FLUENCY-5399] - MONITOR: Cannot start monitor if JRE is downgraded.
[FLUENCY-5406] - DEPLOY: hangs when deploying through domain manager
[FLUENCY-5407] - Doc Impact of FLUENCY-5384 RT: null values in auto-box arrays causes oscdr.cpp:498:
failed C++ assertion 'SW_BOMB'
[FLUENCY-5410] - DOMAIN: display domain error message missing "application"
[FLUENCY-5411] - RUNTIME: loaderutil.cpp:3486: failed C++ assertion 'td->getImplArchType() ==
DSE_IA_JAVA'
[FLUENCY-5414] - CLASSLOADER: Marshalling of arrays of objects doesn't handle null entries.
[FLUENCY-5415] - CLASSLOADER: Marshalling of autoboxed object doesn't handle null.
[FLUENCY-5421] - RUNTIME: OSTEnv::lockOffset have wrong slot description.
[FLUENCY-5422] - JAVASERVICE: jstub::JNIError exception not handled
[FLUENCY-5425] - HA: Updated JOIN_CLUSTER_PURGE and JOIN_CLUSTER_RESTORE to do work in
chunks.
[FLUENCY-5426] - JAVASERVICE: Transaction.execute() passes in unused stack to native abort()
[FLUENCY-5434] - DIST: Write failures are ignored on stale cached objects.
[FLUENCY-5436] - SECURITY: audit messages not logged for java admin targets
[FLUENCY-5438] - RT: Race condition with newly created objects and ReadCommittedSnapshot.
[FLUENCY-5442] - RUNTIME: asynchronous method with array parameters type mismatch exception
[FLUENCY-5449] - Doc Impact of FLUENCY-5442 RUNTIME: asynchronous method with array parameters
type mismatch exception
[FLUENCY-5460] - MC: Codegen of dates is broke.
[FLUENCY-5464] - SECURITY: access violation message does not contain user name
[FLUENCY-5468] - DIST: Need to synchronize transactions when keepalive timeouts cause failover.
[FLUENCY-5476] - ASTA: displays unknown version and build
[FLUENCY-5479] - COORD: sceng.cpp:2143: failed C++ assertion 'm_engineState != S_Killing'.
[FLUENCY-5486] - DEPLOY: slow detecting node stop in attached mode
[FLUENCY-5487] - DEPLOY: restarting node using domain deployment hangs JVM startup
[FLUENCY-5489] - RUNTIME: OSReferenceCache shared memory data corruption
[FLUENCY-5493] - SECURITY: administrator add security, password parameter help confusion
[FLUENCY-5503] - DIST: Default slot decoder crashes with string slots.
[FLUENCY-5504] - DIST: Default slot mismatch processing leaks shared memory.
[FLUENCY-5521] - DIST: restoreFromNode() skips restore if executed after enable() throws exception.
[FLUENCY-5524] - Doc Impact of FLUENCY-1978 DIST: distribution switchadmin plugin enablestatistics
and displaystatistics
[FLUENCY-5539] - SWDIAG: core file generated if -C option fails
[FLUENCY-5542] - ADMINISTRATOR: Improve error handling and logging.
[FLUENCY-5551] - COMPONENT: ComponentNotifier class not found when packaged in the ear file
[FLUENCY-5568] - CLASSLOADER: audit fails when inheriting static fields
[FLUENCY-5569] - DIST: Cannot discover cluster after quorum takes all nodes offline.
[FLUENCY-5571] - CLASSLOADER: not auditing all the way up the inheritance chain for POJO fields
[FLUENCY-5572] - Doc Impact of FLUENCY-5568 CLASSLOADER: audit fails when inheriting static fields
[FLUENCY-5576] - DIST: auto-flush of non-local partitioned objects doesn't work.
[FLUENCY-5586] - HA: object not found error using restoreFromNode property
[FLUENCY-5591] - RUNTIME: dispatch to method on deleted handle
[FLUENCY-5595] - DIST: Timer stress test failure on solaris.
[FLUENCY-5602] - DIST: Crash when bad discovery packet is recieved.
[FLUENCY-5609] - EVENTSERVICE: Leaking com.kabira.platform.logging.EventService instance when jvm
stopped
[FLUENCY-5613] - PERFORMANCE: poor allocator scaling until warmed up
[FLUENCY-5639] - CLASSLOADER: FATAL ERROR in native method: Non-array passed to JNI array
operations
[FLUENCY-5646] - JAVASERVICE: native_object.cpp:1807: failed C++ assertion '_ju.ExceptionCheck() ==
JNI_TRUE'
[FLUENCY-5655] - TIMER: Update to provide best effort resolution
[FLUENCY-5660] - Doc Impact of FLUENCY-5613 PERFORMANCE: poor allocator scaling until warmed up
[FLUENCY-5707] - RT: dirty reads don't work correctly with readcommitted transactions.
[FLUENCY-5715] - DEPLOY: initial load of application slower in BE-X 1.1.0 versus BE-X 1.0.3
[FLUENCY-5717] - JAVASERVICE: classpath with whitespace in directory names not handled correctly
[FLUENCY-5735] - DIST: javatest failed with disteng_skl0.cpp:2980: failed C++ assertion 'SW_BOMB'
[FLUENCY-5738] - RT: incorrect error message in SWProcessManager::setScheduler
[FLUENCY-5759] - RUNTIME: smarthash resize table skips from 500K to 10M
[FLUENCY-5772] - COORD: Update coordinator to handle gdb attaches.
[FLUENCY-5787] - DIST: configuration does not support Infiniband
[FLUENCY-5790] - SNIPPETS: AtomicCreate comments incorrect
[FLUENCY-5793] - CLUSTER: QuorumState node status not displayed
[FLUENCY-5795] - SECURITY: export configuration missing from usage description
[FLUENCY-5797] - Doc Impact of FLUENCY-5415 CLASSLOADER: Marshalling of autoboxed object doesn't
handle null.
[FLUENCY-5799] - SECURITY: display returns multiple rows for same principal if principal has multiple
roles
[FLUENCY-5801] - SECURITY: user not required to use different password after reset
[FLUENCY-5811] - CLASSLOADER: assertion failure in TypeDescriptor.putType()
[FLUENCY-5822] - RUNTIME: javaservice.h:1058: failed C++ assertion 'ordinal < tagCt'
[FLUENCY-5837] - Doc Impact of FLUENCY-5793 CLUSTER: QuorumState node status not displayed
[FLUENCY-5846] - NODE: invalid service name value returned from display node
[FLUENCY-5860] - DEPLOY: unable to deploy ear or jar if name includes dash
[FLUENCY-5870] - DIST: Object create error handling can cause SEGV crashes.
[FLUENCY-5871] - DIST: State conflict on duplicate key'd object creates can cause SEGV crashes.
[FLUENCY-5880] - DIST: State conflict propagated to the caller when searching for objects.
[FLUENCY-5882] - CLASSLOADER: assertion failure with array types
[FLUENCY-5889] - CLASSLOADER: unexpected java.lang.IncompatibleClassChangeError
[FLUENCY-5896] - RT: KeyQuery getMinimumResult() doesn't return null on out of range queries.
[FLUENCY-5916] - HA: partition state not updated on nodes that are not in node list
[FLUENCY-5917] - HA: 'loclhash.cpp:321: failed C++ assertion 'keyLen > 0'' if null returned from partition
mapper
[FLUENCY-5932] - COORD: can't shut node down cleanly when waiting for a notifier to be created
[FLUENCY-5935] - KDM: deployment failing to domain nodes from separate client machine
[FLUENCY-5936] - DIST: Attempted to set a prepare trigger from within a commit or abort trigger
[FLUENCY-5937] - DEPLOY: node failures generate continuous error messages
[FLUENCY-5944] - Doc Impact of FLUENCY-5917 HA: 'loclhash.cpp:321: failed C++ assertion 'keyLen > 0''
if null returned from partition mapper
[FLUENCY-5950] - DEPLOY: mishandling of .. in the deploy tool classpath
[FLUENCY-5958] - DEPLOY: continuous java.net.SocketException: Connection reset errors
[FLUENCY-5960] - DEPLOY: null pointer exception in distributed deployment
[FLUENCY-5961] - DEPLOY: ArrayIndexOutOfBoundsException during distributed deployment
[FLUENCY-5964] - DATAGRID: creating/deleting archived object in same transaction fails
[FLUENCY-5975] - RT: SWSocket error handling for SSL is broke.
[FLUENCY-5977] - DIST: Failover of object instances doesn't happen on orphan partitions.
[FLUENCY-5978] - DIST: Deadlock retries forever on objects in orphan partitions.
[FLUENCY-5987] - DIST: Processing global transactions at startup can lead to concurrent commits.
[FLUENCY-5989] - DIST: Deadlock handling when finding remote instances is broke.
[FLUENCY-5990] - DIST: Runtime using blocking NoLock when attaching objects.
[FLUENCY-5996] - DIST: Shutting down VM for a node causes problems at other nodes
[FLUENCY-5997] - DIST: Deferred writes lost when remote methods updated distributed objects.
[FLUENCY-6025] - DEPLOY: java.lang.NullPointerException on node connection failure
[FLUENCY-6028] - HA: simultaneous define/enable of same partition in cluster causes distributed
deadlock
[FLUENCY-6029] - DEPLOY: error string 'File exists' coordinator log after deployment
[FLUENCY-6032] - RT: Exceptions in dispatched Java methods results in a FATAL trace.
[FLUENCY-6033] - DIST: "Scan for uncommitted partition" warning generated when node restarts.
[FLUENCY-6034] - Doc Impact of FLUENCY-5997 DIST: Deferred writes lost when remote methods
updated distributed objects.
[FLUENCY-6035] - JVM: jvm target missing usage description.
[FLUENCY-6042] - DEPLOY: increase initial timeout for deploy tool waiting for a JVM to start.
[FLUENCY-6043] - DEPLOY: better handling of remote request thread detected failures against domain
targets
[FLUENCY-6076] - DIST: SWHash::DuplicateKey crash seen in the global transaction manager.
[FLUENCY-6085] - STATISTICS: businessstatemachine statistics empty
[FLUENCY-6090] - HA: Partition not migrated if re-defined with a new node list.
[FLUENCY-6092] - Doc Impact of FLUENCY-1797 DIST: Node crash if one node's location is wrong
[FLUENCY-6099] - DIST: Partition notifier node lists are inconsistent when restoring a node.
[FLUENCY-6110] - HA: migrate() allowed on disabled partitions.
[FLUENCY-6113] - DIST: Partitioned instances remain in shared memory if partition is migrated off of
node.
[FLUENCY-6114] - DEPLOY: deploying from Windows adds bogus classpath entries
[FLUENCY-6118] - Doc Impact of FLUENCY-6113 DIST: Partitioned instances remain in shared memory if
partition is migrated off of node.
[FLUENCY-6120] - RT: Oneway dispatches to instance deleted on remote node starts spinning.
[FLUENCY-6122] - DIST: Crash when processing oneways from multiple nodes.
[FLUENCY-6150] - DEPLOY: java.lang.NullPointerException on node shutdown failure
[FLUENCY-6164] - ASTA: web server does not start on Solaris
[FLUENCY-6166] - DIST: forcedReplication is ignored when adding self as replica.
[FLUENCY-6167] - DIST: objectChunks is hardcoded to 1 when adding self as replica.
[FLUENCY-6193] - JAVASERVICE: Cannot find reference warnings
[FLUENCY-6220] - DIST: setMapper() audit can crash on empty instances.
[FLUENCY-6231] - ADMINISTRATOR: Async commands left running when client is killed.
[FLUENCY-6232] - STATISTICS: crash on exit with JNI statistics enabled
[FLUENCY-6241] - KDM: configuration cache promotion deadlock with queued command timer
[FLUENCY-6243] - KDM: configuration cache state out of sync after deadlock
[FLUENCY-6251] - RUNTIME:KeyQuery.getOrCreateSingleResult shows ObjectNotUnique exceptions /
deadlocks in the log
[FLUENCY-6267] - NODE: throttle value not related to low shared memory warning
[FLUENCY-6272] - DIST: canceling HA timer in notifier cause invalid instance handle for object type
[FLUENCY-6298] - UPGRADE: upgrade tool mismatch handling report
[FLUENCY-6307] - UPGRADE: upgrade tool crashes with serialVersionUID values large than 2^32
[FLUENCY-6311] - DIST: NullPointerException can be thrown when a delete() is done while instances are
migrated.
[FLUENCY-6314] - ASTA: javatransaction statistics not integrated
[FLUENCY-6317] - STATISTICS: eventlog status not included in status statistics
[FLUENCY-6321] - NODE: getadminport nodename parameter not used
[FLUENCY-6323] - TIMER: restoring highly available timers causes duplicate key exceptions
[FLUENCY-6336] - CONFIGURATION: configuration load failing with ResourceUnavailable
[FLUENCY-6338] - DIST: enablePartitions() can throw duplicate key exceptions when concurrent deletes
are executing.
[FLUENCY-6342] - Doc Impact of FLUENCY-6321 NODE: getadminport nodename parameter not used
[FLUENCY-6343] - DIST: handle distribution configuration with different name than the active
distribution configuration
[FLUENCY-6345] - DIST: create can be executed multiple times with async replication.
[FLUENCY-6352] - DIST: Empty type name in trace when remote types are published.
[FLUENCY-6356] - DIST: Node down causes multiple "Pushing type" INFO traces.
[FLUENCY-6365] - SWDIAG: allocator report doesn't respect the verbose option
[FLUENCY-6371] - ASTA: audit security drop down does not include target names
[FLUENCY-6374] - RUNTIME: Java dispatcher does not process system exceptions correctly.
[FLUENCY-6397] - CLUSTER: join cluster admin command does not handle NodeMismatch exception
[FLUENCY-6410] - RT: unknown op in distributed callstacks.
[FLUENCY-6418] - DIST: Missing stats updates when replicas are offline or runtime throws an exception.
[FLUENCY-6419] - ARCHGUIDE: Failover removes undiscovered replicas.
[FLUENCY-6421] - RUNTIME: timeout deadlock reporting can examine invalid memory
[FLUENCY-6423] - UPGRADE: upgrade tool doesn't handle mutable key annotation
[FLUENCY-6438] - DIST: Enabling "sparse" partition performs incorrect state transitions.
[FLUENCY-6439] - UPGRADE: changing serial identifier causes type mismatch
[FLUENCY-6441] - RT: Crash when getOrCreateSingleResult() with missing ordered key fields is executed.
[FLUENCY-6444] - ASTA: state enabled buttons not working
[FLUENCY-6446] - Doc Impact of FLUENCY-6418 DIST: Missing stats updates when replicas are offline or
runtime throws an exception.
[FLUENCY-6449] - Doc Impact of FLUENCY-6231 ADMINISTRATOR: Async commands left running when
client is killed.
[FLUENCY-6465] - Doc Impact of FLUENCY-6267 NODE: throttle value not related to low shared memory
warning
[FLUENCY-6470] - SWDIAG: Active transaction dump duplicates locked objects.
[FLUENCY-6471] - DIST: split brain dispatches can cause recursion that kills node.
[FLUENCY-6472] - KDM: deadlocks when executing queued configuration cache commands
[FLUENCY-6485] - DIST: Partition cache can get out of sync with partition in shared memory.
[FLUENCY-6487] - DIST: Sparse instances can have wrong flushable state after a migrate()
[FLUENCY-6495] - CLASSLOADER: VerifyError Bad type on operand stack
[FLUENCY-6496] - COMPONENT: Notifier. getTargetArchive only supports .ear and .jar archives
[FLUENCY-6497] - CLASSLOADER: java.lang.VerifyError - Inconsistent stack height 1 != 0
[FLUENCY-6503] - DIST: Move partition to Unavailable state if enable of partition fails.
[FLUENCY-6506] - FLUSHER: flusher thread disabled by default
[FLUENCY-6507] - FLUSHER: CacheManager.Cache.flush() behavior
[FLUENCY-6508] - UPGRADE: upgrade tool doesn't detect all managed classes
[FLUENCY-6510] - CACHE: CacheManager.Cache.getClasses() ClassNotFoundException
[FLUENCY-6511] - CACHE: flushes per second statistic incorrect if there has been a clear
[FLUENCY-6520] - TIMER: replicated timers prevent clean node shutdown
[FLUENCY-6523] - ASTA: partition display not working
[FLUENCY-6529] - DIST: Enabling sparse partitions generated misleading traces.
[FLUENCY-6530] - DIST: Enabling sparse partitions with force replication can result in state conflicts.
[FLUENCY-6532] - DIST: Distributed queries return unpartitioned objects.
[FLUENCY-6539] - DIST: definePartition() can accidentally enable remote partitions.
[FLUENCY-6542] - Doc Impact of FLUENCY-6438 DIST: Enabling "sparse" partition performs incorrect
state transitions.
[FLUENCY-6543] - RT: Extent iteration can throw null pointer exceptions.
[FLUENCY-6544] - DIST: updated engine termination to stop keepalive thread first.
[FLUENCY-6545] - KDM: queued command timer interval set to incorrect value
[FLUENCY-6548] - Doc Impact of FLUENCY-6503 DIST: Move partition to Unavailable state if enable of
partition fails.
[FLUENCY-6550] - Doc Impact of FLUENCY-6343 DIST: handle distribution configuration with different
name than the active distribution configuration
[FLUENCY-6553] - KDM: segmentation fault when inactivity timeout occurs
[FLUENCY-6554] - NODE: extra new-line in help node usage string
[FLUENCY-6558] - DIST: failover exceptions can cause nodes to crash.
[FLUENCY-6560] - DIST: Remote updates overwrite state, but not status.
[FLUENCY-6561] - DIST: Update keepalive error signaling.
[FLUENCY-6565] - RT: Remove SWSmartReference
[FLUENCY-6567] - DIST: Uncaught application exceptions in notifiers can kill node.
[FLUENCY-6568] - DIST: node list mismatch for sparse partitions
[FLUENCY-6569] - DIST: Incorrect state transitions happen with new versioned partitions.
[FLUENCY-6570] - Doc Impact of FLUENCY-6560 DIST: Remote updates overwrite state, but not status.
[FLUENCY-6576] - CONFIG: remove configuration throws invalid instance handle if nested configuration
object already deleted
[FLUENCY-6577] - JAVASERVICE: java configurations removed during reset=true cleanup even if another
JVM is running
[FLUENCY-6580] - Doc Impact of FLUENCY-6568 DIST: node list mismatch for sparse partitions
[FLUENCY-6584] - ADMIN: upgrade node/create snapshot failure
tibcojre/1.7.0_45_x86_64/lib/amd64/server does not exist
[FLUENCY-6602] - DIST: Aborted enablePartition() doesn't restore state.
[FLUENCY-6603] - DIST: Crash when a engine restart commits a partition.
[FLUENCY-6604] - ASTA: named cache tab not refreshed after creating a cache
[FLUENCY-6609] - DIST: getPartitions() needs to screen out invalid partitions.
[FLUENCY-6613] - ASTA: apache server fails to start on RedHat 6.3
[FLUENCY-6617] - SNAPSHOT: corrupt shared memory message problems
[FLUENCY-6619] - CLASSLOADER: null pointer setting initialized array element
[FLUENCY-6621] - DIST: Async partition broadcasts may be done on partially defined partitions.
[FLUENCY-6624] - DIST: Detection of type mismatches can be missed, causing allocation crashes.
[FLUENCY-6645] - RT: Deadlock truncation is broke.
[FLUENCY-6655] - ADMINGUIDE: Incorrect multi-master restore command line.
[FLUENCY-6656] - DIST: ManagedObject.extent(QUERY_CLUSTER, READLOCK) returns invalid object
[FLUENCY-6667] - CACHE: cache migration failure
[FLUENCY-6668] - CACHE: confusing flush exception
[FLUENCY-6669] - CACHE: stress test crash, flusher calling notifier
[FLUENCY-6670] - BUILD: compilation fails on RedHat 6.4
[FLUENCY-6671] - INSTALLER: update AST installer for P.AST 2.4.0 archive
[FLUENCY-6674] - CACHE: shutdown hang in flusher stress test
[FLUENCY-6677] - Doc Impact of FLUENCY-6668 CACHE: confusing flush exception
[FLUENCY-6687] - RUNTIME: object locking transaction statistics may be inaccurate
[FLUENCY-6700] - RT: Warning trace in findScopedSlot() method.
[FLUENCY-6702] - RT: user's LD_LIBRARY_PATH directories should be at the end of LD_LIBRARY_PATH
[FLUENCY-6703] - MONITOR: sort by name option doesn't work in system allocation window
[FLUENCY-6718] - CACHE: race condition in flusher/disttest
[FLUENCY-6724] - FLUSHER: local objects not flushed
[FLUENCY-6725] - CACHE: removing a named cache doesn't clear cached state of associated type
descriptors
[FLUENCY-6727] - DIST: Type mismatches with one type having no fields can crash the runtime.
[FLUENCY-6737] - UPGRADE: audit failure with cardinality of zero and a non-partitioned object
[FLUENCY-6738] - HA: partition state on remote node set to RemoteEnabled before enabled
[FLUENCY-6739] - STATISTICS: namedcache statisitc not displaying correctly
[FLUENCY-6748] - FLUSH: delete trigger not called on flush
[FLUENCY-6750] - PARTITION: broadcastupdates not reset to true following an enable
[FLUENCY-6754] - DIST: Need to synchronize async thread with distribution shutdown.
[FLUENCY-6755] - Doc Impact of FLUENCY-6748 FLUSH: delete trigger not called on flush
[FLUENCY-6756] - ACTIVESPACES: data coherency, leaked transactions
[FLUENCY-6757] - ACTIVESPACES: memory leak
[FLUENCY-6763] - CACHE: size 0 cache - object never flushed
[FLUENCY-6769] - SNAPSHOT: create snapshot fails when use destination parameter and relative path
[FLUENCY-6774] - DIST: processTypes() crashes if dependent type list has duplicates.
[FLUENCY-6778] - FLUSHER: flusher thread not handling ResourceUnavailable
[FLUENCY-6791] - DIST: Keepalive processing ignores nonResponseTimeoutSeconds in sending response.
[FLUENCY-6798] - RT: Update illegal call to exit message
[FLUENCY-6800] - NODE: node name of E fails to start
[FLUENCY-6801] - RT: misleading deadlock backoff warning trace during shutdown
[FLUENCY-6803] - STATISTICS: bad start date in statistics output
[FLUENCY-6809] - DIST: Duplicate reference handling can result in uncaught deadlocks.
[FLUENCY-6817] - DIST: duplicate key processing is broke for secondary keys.
[FLUENCY-6830] - DIST: partition cache gets out of sync with shared memory.
[FLUENCY-6831] - DIST: Remote reference lookup can fail with uncaught operation retry when
marshaling response.
[FLUENCY-6844] - RT: engine.cpp(2856)|transaction [88:0, tid 7582]: <empty message>
[FLUENCY-6845] - TOOLS: sw_mkall -c in core leaves the shared design center running
[FLUENCY-6848] - RT: SIGSEGV using READ_COMMITTED_SNAPSHOT asynchronous method
[FLUENCY-6849] - ADMINISTRATION: object destroyed trace when checking for abandoned commands
[FLUENCY-6850] - MONITOR: display decimal number with leading 0x
[FLUENCY-6852] - DIST: Transaction can be abandoned when doing distributed find/query.
[FLUENCY-6878] - RT: runObjectNotifier() crashes if non-Java types are present in transaction.
[FLUENCY-6879] - DIST: Remote enable of partitions allowed on node that has left the cluster.
[FLUENCY-6892] - CONFIG: Crash in switchconfig.act if no inactive version exists.
[FLUENCY-6896] - RT: Dispatching twoway that returns a null array crashes the runtime.
[FLUENCY-6904] - ASTPERF: PLATFORM-HIGHAVAILABILITY-0080 - Migrate done on wrong node
[FLUENCY-6905] - ASTPERF: Performance test80 and test90 missing from nightly regressions
[FLUENCY-6917] - ASTPERF: javadoc for Base.cleanMemory() incorrect
[FLUENCY-6920] - DIST: Remove all notion of "implicit" commit and abort.
[FLUENCY-6923] - MONITOR: domain and type identifiers displayed as negative number
[FLUENCY-6934] - Doc Impact of FLUENCY-6738 HA: partition state on remote node set to
RemoteEnabled before enabled
[FLUENCY-6945] - DIST: Engine gets stuck in failover due to active channels not getting released.
[FLUENCY-6946] - DIST: Runtime retries deadlocks forever when processing types.
[FLUENCY-6960] - ASTPERF: Cannot find property "nodeList" for PLATFORM-MANAGEDOBJECT-0045
[FLUENCY-6962] - RT: implicit character set conversions cause oscdr.cpp(497)|Invalid increment of
540238 bytes
[FLUENCY-6967] - SWDIAG: assertion failure / crash with -c option (DEVELOPMENT mode)
[FLUENCY-6968] - SWDIAG: Displays negative numbers for type descriptor values
[FLUENCY-6972] - MONITOR: navigating to type spec broken
[FLUENCY-6973] - MONITOR: object references displaying with negative numbers
[FLUENCY-6975] - JAVASERVICE: Updated all Java exposed interfaces that pass binary strings to use byte
arrays.
[FLUENCY-6989] - RUNTIME: Memory leak in OSTypeReferences implementation.
[FLUENCY-6995] - DIST: sendObjectRequest() can access deleted write channels.
Documentation
[FLUENCY-3342] - HA: Determine how to handle "orphan" objects and partitions.
[FLUENCY-4774] - ADMINGUIDE: document command line
[FLUENCY-4787] - DEVGUIDE: update distributed caching to describe refresh on write
[FLUENCY-5191] - ARCHGUIDE: update transaction outcome voting
[FLUENCY-5386] - DEVGUIDE: add section on named caches
[FLUENCY-5390] - DEVGUIDE: mismatched keys not supported on upgrade
[FLUENCY-5396] - ADMINGUIDE: remove kabira.metrics from sample node configuration in
Administration guide
[FLUENCY-5397] - USERDOC: Update copyright year
[FLUENCY-5492] - ADMINISTRATION: clarify deploy directory behavior
[FLUENCY-5526] - ASTA: Apache fails to start on RedHat 6.1
[FLUENCY-5578] - 3RDPARTY: Duplicate declarations fro Lookup3.x and OpenLDAP
[FLUENCY-5634] - 3RDPARTY: remove dns_sd.h and mdns_sd.sh from 3rd-party file
[FLUENCY-5671] - 3RDPARTY: Incorrect 3rd party software vendor and contact URL information
[FLUENCY-5686] - TUNING: add schedulerpolicy deploy tool parameter
[FLUENCY-5756] - DEVGUIDE: add explanation and sample for key inheritance
[FLUENCY-5761] - TUNING: vmstat sample interval not documented
[FLUENCY-5808] - TUNING: add Linux tunable parameters to tuning guide
[FLUENCY-5851] - TUNING: add a note about hardware configuration and tuning
[FLUENCY-5881] - DIST: document getOrCreateSingleResult() behavior with query scope and partitioned
objects
[FLUENCY-6070] - ADMINGUIDE: make it explicit that domain managers are not part of cluster quorum
calculation
[FLUENCY-6080] - DEVGUIDE: add information on behavior of static fields
[FLUENCY-6160] - DEVGUIDE: describe use of PartitionManager.waitForNode
[FLUENCY-6200] - ADMINGUIDE: document event log management
[FLUENCY-6247] - HA: PartitionManager.enablePartition is synchronous
[FLUENCY-6248] - HA: add details on when PartitionManager.EnableAction types should be used
[FLUENCY-6249] - DEVGUIDE: add section on restoring a node
[FLUENCY-6250] - HA: API and administration commands should not be used together
[FLUENCY-6270] - DOCUMENTATION: Developer's Guide, Distributed Computing, Guidelines
[FLUENCY-6274] - DOCUMENTATION: Transaction javadoc
[FLUENCY-6275] - TIMER: add more details to the timer javadoc on transactional guarantees
[FLUENCY-6324] - RELEASENOTES: sudo package for starting mdns on Solaris
[FLUENCY-6325] - DEVGUIDE: ordering of partitioning enabling undefined
[FLUENCY-6348] - DOCUMENTATION: cluster wide queries and implicit write locks
[FLUENCY-6357] - DEVGUIDE: add transparent retry snippet
[FLUENCY-6368] - CONFIGURATION: ConfigurationListener javadoc never defines "register"
[FLUENCY-6453] - TUNING: document life-cycle of statistics
[FLUENCY-6463] - DEVGUIDE: devguide distributed computing
[FLUENCY-6478] - DOCS: Update copyright to 2014
[FLUENCY-6492] - ADMINGUIDE: join cluster clarifications
[FLUENCY-6509] - CACHE: document flushing behavior in flush() API javadoc
[FLUENCY-6516] - DEVGUIDE: document enable node blocking behavior waiting for active node
[FLUENCY-6538] - ARCHGUIDE: add description of sparse partitions
[FLUENCY-6566] - ADMINGUIDE: describe System V Shared Memory and node removal in Administration
Guide
[FLUENCY-6583] - ARCHGUIDE: clarify that partitioned objects are removed after node leaves cluster
[FLUENCY-6597] - DOCUMENTATION: Asynchronous methods and JVM lifecycle
[FLUENCY-6633] - DEVGUIDE: High availability questions
[FLUENCY-6638] - DEVGUIDE: Keys and Queries, Developer's Guide, Managed Object with a Unique Key
[FLUENCY-6642] - SIZING: incorrect reference to Type.memorySize
[FLUENCY-6662] - Doc Impact of FLUENCY-6509 CACHE: document flushing behavior in flush() API
javadoc
[FLUENCY-6726] - ARCHGUIDE: upgrade and non-partitioned objects
[FLUENCY-6734] - TIMER: document timer behavior when seconds=0
[FLUENCY-6736] - INSTALL: updates to installation guide for AST 2.4.0
[FLUENCY-6835] - TUNING: Add garbage collection recommendations
[FLUENCY-6837] - ARCHGUIDE: update remote node states
[FLUENCY-6839] - DEVGUIDE: upgrade JVM SE 6 references to JVM SE 7
[FLUENCY-6843] - 3RDPARTY: updated supported FireFox browser version
[FLUENCY-6870] - TUNING: add information about garbage collection log file rollover
[FLUENCY-6876] - INSTALL: install guide release notes for extra package requirements
[FLUENCY-6890] - TUNING: add section about running statistics in a production system
[FLUENCY-6925] - ASTPERF: sw_mvn command
[FLUENCY-6926] - ASTPERF: How to build test plan
[FLUENCY-6928] - ARCHGUIDE: make it explicit that transaction boundaries are maintained during async
replication.
Improvement
[FLUENCY-845] - DIST: support SSL for node-to-node connectivity
[FLUENCY-1159] - NODE: support commands on node target to stop/restart coordinator
[FLUENCY-1943] - ASTA: integrate distribution statistics
[FLUENCY-1966] - COMPONENT: support non-public notifiers
[FLUENCY-1974] - STATISTICS: integrate distribution statistics with the runtimestatistics plugin
[FLUENCY-1992] - AST: duplicate release notes in fluency and fluency-maven project
[FLUENCY-2867] - DEPLOY: add option for ignoring options file
[FLUENCY-3060] - KDM: audit connectionRetryIntervalSeconds > 0
[FLUENCY-3388] - SNAPSHOT: implement sharedmemory option
[FLUENCY-3390] - ASTA: provide navigation from partition screens to enable partitions on node screen.
[FLUENCY-3451] - DEPLOY: remove JMX information on stdout
[FLUENCY-3505] - DEPLOY: detect mismatched versions between client and server.
[FLUENCY-4102] - HA: provide access to conflicting data in compensation trigger
[FLUENCY-4154] - SNAPSHOT: help not available when node down
[FLUENCY-4184] - HA: add support for setting partition properties
[FLUENCY-4274] - MAVEN: Supply maven-based license information
[FLUENCY-4383] - CLASSLOADER: Getting synthetic m_objectReference for annotated classes is
inefficient
[FLUENCY-4504] - NODE: remove system coordinator information from install node
[FLUENCY-4534] - DEPLOY: support templated deployment specifications
[FLUENCY-4671] - SNAPSHOT: capture jstack output on JVMs
[FLUENCY-5006] - STATISTICS: rename administration target to statistics
[FLUENCY-5076] - DEPLOY: deploy tool debug tracing
[FLUENCY-5166] - DEPLOY: remote class and resource loading performance
[FLUENCY-5295] - STATISTICS: time parameter to snapshot command should include units
[FLUENCY-5310] - HA: Enhance quorum support to allow more sophisticated rules.
[FLUENCY-5350] - JAVADOC: SDK javadoc generation script
[FLUENCY-5373] - CACHE: add support for remove cache command
[FLUENCY-5389] - CACHE: rename default named partitions
[FLUENCY-5429] - ADMINISTRATION: verbosenode option type not described
[FLUENCY-5430] - RUNTIME: threadId is not set in production systems.
[FLUENCY-5439] - DIST: Log before and after partition definitions when migrating partitions.
[FLUENCY-5445] - ADMINISTRATION: java management targets should validate command before access
check
[FLUENCY-5453] - BUILD: remove monitor.mak
[FLUENCY-5469] - RUNTIME: poor hashing distribution with string/binary hash function
[FLUENCY-5481] - DEPLOY: generate event when user executes a deploy command
[FLUENCY-5501] - RUNTIME: large numbers of objects and hashing
[FLUENCY-5523] - Doc Impact of FLUENCY-1974 RUNTIMESTATISTICS: integrate distribution statistics
with the runtimestatistics plugin
[FLUENCY-5525] - Doc Impact of FLUENCY-1943 ASTA: integrate distribution statistics
[FLUENCY-5537] - Doc Impact of FLUENCY-5310 HA: Enhance quorum support to allow more
sophisticated rules.
[FLUENCY-5561] - CLASSLOADER: cache isLocal() in the Java managed object
[FLUENCY-5566] - RUNTIME: improve the performance of hashCode() for Managed objects
[FLUENCY-5567] - CLASSLOADER: improve cannot inherit fields audit failure
[FLUENCY-5574] - DIST: add a nodeActiveTimeoutSeconds configuration attribute
[FLUENCY-5582] - PARTITION: add support for defining partitions on nodes not in partition definition
[FLUENCY-5583] - HA: automatically maintain cluster wide view of partitions
[FLUENCY-5643] - RUNTIME: include thread id in transaction name
[FLUENCY-5645] - RT: Add support for changing the scheduling policy and parameters.
[FLUENCY-5661] - MC: Updated codegen of virtual dispatch to not use exceptions.
[FLUENCY-5664] - RT: Improve segmented allocator reports
[FLUENCY-5681] - RT: improve concurrency of com.kabira.platform.ManagedObject.isManagedClass
[FLUENCY-5749] - DEPLOY: copy target to server in detached mode
[FLUENCY-5776] - RT: scheduler policy failure crashes the node
[FLUENCY-5794] - RT: improve ManagedClassError exception message
[FLUENCY-5796] - Doc Impact of FLUENCY-5664 RT: Improve segmented allocator reports
[FLUENCY-5800] - CLASSLOADER: add info trace for dynamic class path additions
[FLUENCY-5834] - Doc Impact of FLUENCY-5582 PARTITION: add support for defining partitions on nodes
not in partition definition
[FLUENCY-5838] - DIST: Remove "Not Initialized" quorum state.
[FLUENCY-5862] - SNAPSHOT: add Apache log files to snapshot file
[FLUENCY-5933] - DEPLOY: add engine log tracing that captures user's deploy command
[FLUENCY-5939] - COORD: cannot enable debug tracing in the swcoordadmin engine
[FLUENCY-5951] - Doc Impact of FLUENCY-5166 DEPLOY: remote class and resource loading
performance
[FLUENCY-5968] - 3RDPARTY: Update Java SE7 version
[FLUENCY-5980] - Doc Impact of FLUENCY-845 DIST: support SSL for node-to-node connectivity
[FLUENCY-5981] - DIST: Include current broadcastAddressList value in display command.
[FLUENCY-5991] - DEPLOY: move product-specific logic out of the generic deploy tool
[FLUENCY-6006] - DEPLOY: add support for setting engine abort timer value
[FLUENCY-6007] - COORD: remove kill timer
[FLUENCY-6018] - DIST: executing distributed operation from commit trigger fails
[FLUENCY-6038] - DEPLOY: improve formatting of deploy class path on server
[FLUENCY-6045] - SHUTDOWN: need better error message when pstack not installed
[FLUENCY-6055] - Doc Impact of FLUENCY-5838 DIST: Remove "Not Initialized" quorum state.
[FLUENCY-6061] - Doc Impact of FLUENCY-3505 DEPLOY: detect mismatched versions between client
and server.
[FLUENCY-6073] - DEPLOY: load failed error message repeated
[FLUENCY-6074] - Doc Impact of FLUENCY-5574 DIST: add a nodeActiveTimeoutSeconds configuration
attribute
[FLUENCY-6075] - Doc Impact of FLUENCY-5749 DEPLOY: copy target to server in detached mode
[FLUENCY-6088] - DIST: PartitionManager::isPartitioned causing EventBus dispatch
[FLUENCY-6101] - Doc Impact of FLUENCY-5968 3RDPARTY: Update Java SE7 version
[FLUENCY-6119] - Doc Impact of FLUENCY-5583 HA: automatically maintain cluster wide view of
partitions
[FLUENCY-6127] - Doc Impact of FLUENCY-6088 DIST: PartitionManager::isPartitioned causing EventBus
dispatch
[FLUENCY-6153] - DIST: Improve scaling when migrating objects.
[FLUENCY-6191] - 3RDPARTY: Update Java SE 7 to Update 45
[FLUENCY-6195] - SNAPSHOT: old events logs make snapshot large
[FLUENCY-6206] - DIST: Add support for accessing all partitions from a remote node.
[FLUENCY-6214] - DIST: audit that restoreFromNode matches the active node
[FLUENCY-6227] - DIST: Update setMapper() to return old instance if it exists.
[FLUENCY-6229] - Doc Impact of FLUENCY-6153 DIST: Improve scaling when migrating objects.
[FLUENCY-6230] - DIST: Improve scaling when re-partitioning objects.
[FLUENCY-6234] - STATISTICS: normalize JNI and Native statistics
[FLUENCY-6256] - RT: improve ObjectServices.objectToString error reporting
[FLUENCY-6259] - RT: ObjectServices.objectToString returns empty string for objects deleted in current
transaction
[FLUENCY-6260] - DIST: add undiscovered node name to unresolved location log
[FLUENCY-6262] - RT: Deprecate swbuiltin.ObjectServices.objectToString() / stringToObject(), replace
with ManagedObject methods
[FLUENCY-6264] - CLASSLOADER: improve illegal return type exception message
[FLUENCY-6268] - RT: throttle log should be a warning
[FLUENCY-6283] - Doc Impact of FLUENCY-5006 RUNTIMESTATISTICS: rename administration target to
statistics
[FLUENCY-6287] - CLUSTER: add support for join cluster purge
[FLUENCY-6289] - Doc Impact of FLUENCY-6262 RT: Deprecate swbuiltin.ObjectServices.objectToString()
/ stringToObject(), replace with ManagedObject methods
[FLUENCY-6294] - PARTITION: rename update partition to migrate partition
[FLUENCY-6306] - CACHE: duplicate create does not raise an error
[FLUENCY-6309] - CHANNEL: make serviceName & endpointName parameter names all lowercase
[FLUENCY-6316] - STATISTICS: unknown statistics type should print supported types
[FLUENCY-6322] - NODE: nodename parameter not needed for remove node
[FLUENCY-6327] - DIST: Update waitForNode() to include whether remote node can access local node.
[FLUENCY-6335] - STATISTICS: rename snapshot time parameter to seconds
[FLUENCY-6337] - HA: update display partition to display list of classes mapped into a partition
[FLUENCY-6339] - Doc Impact of FLUENCY-6227 DIST: Update setMapper() to return old instance if it
exists.
[FLUENCY-6340] - Doc Impact of FLUENCY-6214 DIST: audit that restoreFromNode matches the active
node
[FLUENCY-6341] - Doc Impact of FLUENCY-6322 NODE: nodename parameter not needed for remove
node
[FLUENCY-6344] - Doc Impact of FLUENCY-6335 STATISTICS: rename snapshot time parameter to
seconds
[FLUENCY-6358] - DIST: improve distributed deadlock reporting
[FLUENCY-6361] - Doc Impact of FLUENCY-4184 HA: add support for setting partition properties
[FLUENCY-6362] - Doc Impact of FLUENCY-6287 CLUSTER: add support for join cluster purge
[FLUENCY-6367] - SNAPSHOT: remove hard-wired list of system process log file name
[FLUENCY-6370] - PARTITION: add Restore From Node in display output
[FLUENCY-6376] - Doc Impact of FLUENCY-1159 NODE: support commands on node target to
stop/restart coordinator
[FLUENCY-6377] - Doc Impact of FLUENCY-4504 NODE: remove system coordinator information from
install node
[FLUENCY-6381] - DIST: improve tracing of partition transitions
[FLUENCY-6382] - SNAPSHOT: add host memory / swap information
[FLUENCY-6384] - Doc Impact of FLUENCY-6358 DIST: improve distributed deadlock reporting
[FLUENCY-6386] - RT: Remove "java stack" delineators generated by javaservice library.
[FLUENCY-6387] - Doc Impact of FLUENCY-6337 HA: update display partition to display list of classes
mapped into a partition
[FLUENCY-6388] - Doc Impact of FLUENCY-6370 PARTITION: add Restore From Node in display output
[FLUENCY-6391] - DIST: Update asynchronous replication to have the work always performed on the
active node.
[FLUENCY-6393] - Doc Impact of FLUENCY-4154 SNAPSHOT: help not available when node down
[FLUENCY-6402] - RT: improve com.kabira.platform.ManagedClassError exception
[FLUENCY-6405] - HA: replace the leave cluster force parameter with an equivalent enum parameter
[FLUENCY-6406] - Doc Impact of FLUENCY-6191 3RDPARTY: Update Java SE 7 to Update 45
[FLUENCY-6420] - Doc Impact of FLUENCY-6405 HA: replace the leave cluster force parameter with an
equivalent enum parameter
[FLUENCY-6429] - NODEAGENT: publish event when remote connection established
[FLUENCY-6432] - NODEAGENT: publish event on connection failure
[FLUENCY-6434] - SNAPSHOT: add load average information
[FLUENCY-6442] - RT: Support arrays for additional fields in KeyQuery.getOrCreateSingleResult()
[FLUENCY-6443] - UPGRADE: support upgrade node that does not perform an upgrade
[FLUENCY-6445] - DIST: Update migrate state checks to only be applied to current
objectsLockedPerTransaction.
[FLUENCY-6448] - Doc Impact of FLUENCY-6443 UPGRADE: support upgrade node that does not perform
an upgrade
[FLUENCY-6469] - DIST: Adding or removing a replica is single threaded.
[FLUENCY-6486] - RT: Updated deadlock backoff to shorten sleep interval for retryable operations.
[FLUENCY-6498] - STATISTICS: add global transaction identifiers to blocked transaction report
[FLUENCY-6502] - RUNTIME: update deadlock backoff traces to include transaction identifier and
description
[FLUENCY-6512] - NODE: add product version to display
[FLUENCY-6513] - Doc Impact of FLUENCY-6512 NODE: add product version to display
[FLUENCY-6533] - DIST: Support both static and dynamic discovery in a cluster.
[FLUENCY-6536] - DIST: Improve performance of multiple remote node failures.
[FLUENCY-6541] - Doc Impact of FLUENCY-6445 DIST: Update migrate state checks to only be applied to
current objectsLockedPerTransaction.
[FLUENCY-6547] - Doc Impact of FLUENCY-6533 DIST: Support both static and dynamic discovery in a
cluster.
[FLUENCY-6582] - JMX: improve failed to start message
[FLUENCY-6610] - COORD: add user environment information
[FLUENCY-6611] - SNAPSHOT: add operating system versions
[FLUENCY-6679] - Doc Impact of FLUENCY-5373 CACHE: add support for remove cache command
[FLUENCY-6682] - DIST: include global transaction identifier in deadlock messages
[FLUENCY-6691] - COORD: user PATH not inherited on node startup
[FLUENCY-6705] - DIST: Update keepalives to tolerate transient network failures.
[FLUENCY-6706] - SNAPSHOT: add process information
[FLUENCY-6707] - HA: add partition information to ResourceUnavailableException
[FLUENCY-6719] - DIST: Skip read channel wait if channels are blocked in poll().
[FLUENCY-6745] - DIST: Update keepalive processing to process multiple node failures.
[FLUENCY-6753] - Doc Impact of FLUENCY-6498 STATISTICS: add global transaction identifiers to blocked
transaction report
[FLUENCY-6761] - Doc Impact of FLUENCY-6691 COORD: user PATH not inherited on node startup
[FLUENCY-6782] - Doc Impact of FLUENCY-6006 DEPLOY: add support for setting engine abort timer
value
[FLUENCY-6785] - CONFIG: improve notifier not available error message
[FLUENCY-6788] - FLUSHER: ignore notifier and trigger failures
[FLUENCY-6790] - RT: Update engine terminate message to remove abortTimeout
[FLUENCY-6792] - QUERY: KeyIterator finalizer and garbage collection
[FLUENCY-6796] - Doc Impact of FLUENCY-6705 DIST: Update keepalives to tolerate transient network
failures.
[FLUENCY-6815] - DIST: add request type to cannot send request trace message
[FLUENCY-6826] - RT: Excessive INFO traces for dispatched exceptions.
[FLUENCY-6834] - JAVASERVICE: enable garbage collection log
[FLUENCY-6846] - Doc Impact of FLUENCY-2867 DEPLOY: add option for ignoring options file
[FLUENCY-6856] - RT: add inexpensive quick validation for shared memory offset
[FLUENCY-6868] - SNAPSHOT: add event bus report
[FLUENCY-6883] - DIST: Add EnableAction and DisableAction values to the Info traces when doing an
enable or disable.
[FLUENCY-6889] - STATISTICS: add parameter data to event bus event reporting
[FLUENCY-6913] - RUNTIME: add debug trace message for timers
[FLUENCY-6933] - ASTPERF: change PLATFORM-MANAGEDOBJECT-0050 result reporting
[FLUENCY-6938] - SNAPSHOT: add threads report
[FLUENCY-6944] - DIST: Warning on ObjectNotUnique is misleading.
[FLUENCY-6947] - SNAPSHOT: Add distribution information to snapshot.
[FLUENCY-6954] - SNAPSHOT: Add snapshot timestamp
[FLUENCY-6956] - DEPLOY: JVMDeployed event not in audit log
[FLUENCY-6963] - RT: store strings in shared memory as UTF-8
[FLUENCY-6981] - ENGINE: update process name to engine name
[FLUENCY-6993] - Doc Impact of FLUENCY-6391 DIST: Update asynchronous replication to have the work
always performed on the active node.
New Feature
[FLUENCY-3453] - ASTA: support discover cluster
[FLUENCY-5145] - STATISTICS: add visibility into what operations are being dispatched across the event
bus
[FLUENCY-5353] - DIST: add support for Infiniband
[FLUENCY-5470] - STATISTICS: add hash distribution reporting
[FLUENCY-5490] - Doc Impact of FLUENCY-5470 RUNTIMESTATISTICS: add hash distribution reporting
[FLUENCY-5616] - DATAGRID: ship RC file to start up data grid
[FLUENCY-5812] - Doc Impact of FLUENCY-5353 DIST: add support for Infiniband
[FLUENCY-5919] - HA: add support for removing a partition mapper
[FLUENCY-5931] - Doc Impact of FLUENCY-5145 RUNTIMESTATISTICS: add visibility into what operations
are being dispatched across the event bus
[FLUENCY-5945] - Doc Impact of FLUENCY-5919 HA: add support for removing a partition mapper
[FLUENCY-5947] - CACHE: add support for flushing any managed object
[FLUENCY-6053] - CACHE: add support for flush notification
[FLUENCY-6170] - STATISTICS: add JNI statistics
[FLUENCY-6181] - DIST: Add support for cardinality() method to partition
[FLUENCY-6198] - SNAPSHOT: add swdiag types/cardinality to snapshot output
[FLUENCY-6199] - SNAPSHOT: add memory utilization report
[FLUENCY-6225] - Doc Impact of FLUENCY-6170 RUNTIMESTATISTICS: add JNI statistics
[FLUENCY-6366] - SNAPSHOT: add active transactions report to snapshot
[FLUENCY-6372] - STATISTICS: add runtime data structures memory usage report
[FLUENCY-6375] - STATISTICS: add active transactions report
[FLUENCY-6396] - Doc Impact of FLUENCY-6375 STATISTICS: add active transactions report
[FLUENCY-6398] - Doc Impact of FLUENCY-3453 ASTA: support discover cluster
[FLUENCY-6399] - SNAPSHOT: add no shared memory option
[FLUENCY-6400] - Doc Impact of FLUENCY-6372 STATISTICS: add runtime data structures memory usage
report
[FLUENCY-6413] - RUNTIME: add transaction description
[FLUENCY-6422] - Doc Impact of FLUENCY-6369 PARTITION: add support for enabling/disabling single
partitions
[FLUENCY-6431] - Doc Impact of FLUENCY-6413 RUNTIME: add transaction description
[FLUENCY-6499] - CACHE: add statistics
[FLUENCY-6501] - CACHE: support inheritance
[FLUENCY-6618] - Doc Impact of FLUENCY-6053 CACHE: add support for flush notification
[FLUENCY-6648] - Doc Impact of FLUENCY-5947 CACHE: add support for flushing any managed object
[FLUENCY-6659] - Doc Impact of FLUENCY-6499 CACHE: add statistics
[FLUENCY-6663] - HA: add support for simulating split-brain
[FLUENCY-6676] - Doc Impact of FLUENCY-6501 CACHE: support inheritance
[FLUENCY-6740] - Doc Impact of FLUENCY-6663 HA: add support for simulating split-brain
[FLUENCY-6751] - PARTITION: add support for sparse audit property
[FLUENCY-6819] - RT: Add support for access key data in Managed objects.
[FLUENCY-6842] - Doc Impact of FLUENCY-6751 PARTITION: add support for sparse audit property
[FLUENCY-6916] - ASTPERF: PlatformBase not appropriate for partition change tests
Task
[FLUENCY-4376] - SCHEDULER: deprecate timed scheduler
[FLUENCY-4377] - LOGGER: deprecate logger configuration
[FLUENCY-5139] - DEPLOY: redundant nodecleanup handling
[FLUENCY-5143] - DEPLOY: ktvmadmin::JVM::deployFiles array
[FLUENCY-5168] - 3RDPARTY: Update minimum RedHat version to 5.8
[FLUENCY-5444] - COMPONENTS: package components in product archive
[FLUENCY-5633] - 3RDPARTY: certify on RedHat 6.4
[FLUENCY-5703] - 3RDPARTY: Upgrade Apache version to a 2.2.24
[FLUENCY-5766] - 3RDPARTY: upgrade PHP to 5.4.15
[FLUENCY-5767] - 3RDPARTY: upgrade openssl to 1.0.1e
[FLUENCY-5768] - 3RDPARTY: upgrade openldap to 2.4.35
[FLUENCY-5770] - TIMERS: remove obsolete timer support from core
[FLUENCY-5821] - 3RDPARTY: upgrade universal installer version
[FLUENCY-5833] - Doc Impact of FLUENCY-5444 COMPONENTS: package components in product archive
[FLUENCY-5969] - 3RDPARTY: upgrade to a minimum of MDNS 379.37.
[FLUENCY-5970] - 3RDPARTY: remove JDK 6 from the release
[FLUENCY-5986] - Doc Impact of FLUENCY-5969 3RDPARTY: upgrade to a minimum of MDNS 379.37.
[FLUENCY-5998] - 3RDPARTY: upgrade PHP to 5.4.19
[FLUENCY-5999] - 3RDPARTY: Upgrade Apache version to at least 2.2.25
[FLUENCY-6134] - Doc Impact of FLUENCY-5633 3RDPARTY: certify on RedHat 6.4
[FLUENCY-6175] - 3RDPARTY: Solaris 10 1/13 certification
[FLUENCY-6278] - EVENT: mark event target as private
[FLUENCY-6279] - HISTORY: mark history target as private
[FLUENCY-6282] - CACHE: convert to modeled plugin
[FLUENCY-6284] - DIST: mark distribution target as private
[FLUENCY-6285] - LDAP: rename ldapauth --> ldap
[FLUENCY-6291] - SECURITY: deprecate x.509 support
[FLUENCY-6299] - Doc Impact of FLUENCY-4376 SCHEDULER: deprecate timed scheduler
[FLUENCY-6300] - SCHEDULER: remove deprecated scheduled admin command support
[FLUENCY-6302] - SNAPSHOT: remove display, expand, and remove commands
[FLUENCY-6318] - STATISTICS: remove node statistic
[FLUENCY-6319] - NODE: remove principal information from display node
[FLUENCY-6320] - 3RDPARTY: upgrade to AS 2.0.2 HF9
[FLUENCY-6334] - Doc Impact of FLUENCY-6175 3RDPARTY: Solaris 10 1/13 certification
[FLUENCY-6363] - Doc Impact of FLUENCY-6302 SNAPSHOT: remove display, expand, and remove
commands
[FLUENCY-6380] - Doc Impact of FLUENCY-6319 NODE: remove principal information from display node
[FLUENCY-6437] - Doc Impact of FLUENCY-6320 3RDPARTY: upgrade to AS 2.0.2 HF9
[FLUENCY-6455] - 3RDPARTY: upgrade PHP to at least 5.4.25
[FLUENCY-6456] - 3RDPARTY: upgrade OpenSSL to at least 1.0.1f
[FLUENCY-6457] - DOCS: Update to new branding
[FLUENCY-6479] - 3RDPARTY: Upgrade TIBCO Universal Installer version to 3.2.2
[FLUENCY-6500] - CACHE: deprecate com.kabira.platform.CacheManager.CacheFlusher
[FLUENCY-6527] - 3RDPARTY: upgrade Apache to 2.2.26
[FLUENCY-6549] - Doc Impact of FLUENCY-5970 3RDPARTY: remove JDK 6 from the release
[FLUENCY-6601] - 3RDPARTY: upgrade PHP to 5.4.27
[FLUENCY-6612] - 3RDPARTY: upgrade OpenLdap to 2.4.39
[FLUENCY-6614] - 3RDPARTY: upgrade Apache to 2.2.27
[FLUENCY-6615] - Doc Impact of FLUENCY-6500 CACHE: deprecate
com.kabira.platform.CacheManager.CacheFlusher
[FLUENCY-6623] - VMWARE: Deprecate development appliance
[FLUENCY-6626] - 3RDPARTY: upgrade OpenSSL to at least 1.0.1g
[FLUENCY-6627] - 3RDPARTY: Update Java SE 7 to Update 55
[FLUENCY-6628] - 3RDPARTY: upgrade PHP to at least 5.5.11
[FLUENCY-6629] - FLUSHER: renable flush test in distsrv/partitiontest
[FLUENCY-6631] - SWDC: Update to use swsrv binary executable
[FLUENCY-6640] - AST: rename default application from fluency to ast
[FLUENCY-6641] - Doc Impact of FLUENCY-6640 AST: rename default application from fluency to ast
[FLUENCY-6643] - DIST: update removenode distribution command to only update local node
[FLUENCY-6657] - STATISTICS: rename objectcache statistic to namedcache
[FLUENCY-6658] - Doc Impact of FLUENCY-6657 STATISTICS: rename objectcache statistic to
namedcache
[FLUENCY-6713] - CACHE: remove user visibility of default caches
[FLUENCY-6716] - Doc Impact of FLUENCY-6713 CACHE: remove user visibility of default caches
[FLUENCY-6728] - Doc Impact of FLUENCY-6300 SCHEDULER: remove deprecated scheduled admin
command support
[FLUENCY-6746] - Doc Impact of FLUENCY-6627 3RDPARTY: Update Java SE 7 to Update 55
[FLUENCY-6768] - 3RDPARTY: upgrade OpenSSL to at least 1.0.1h
[FLUENCY-6772] - 3RDPARTY: upgrade PHP to at least 5.5.13
[FLUENCY-6775] - Doc Impact of FLUENCY-6772 3RDPARTY: upgrade PHP to at least 5.5.13
[FLUENCY-6789] - 3RDPARTY: upgrade ActiveSpaces to version 2.1.2 HF 14
[FLUENCY-6860] - RT: propagate deadlock reason
[FLUENCY-6873] - 3RDPARTY: Certify and add support for RedHat Linux 6.5
[FLUENCY-6874] - 3RDPARTY: upgrade to TIBCO JRE 7 update 65
[FLUENCY-6884] - Doc Impact of FLUENCY-6873 3RDPARTY: Certify and add support for RedHat Linux 6.5
[FLUENCY-6898] - Doc Impact of FLUENCY-6789 3RDPARTY: upgrade ActiveSpaces to version 2.1.2 HF 14
[FLUENCY-6906] - ASTPERF: add numberOfThreads scaling to HA migration and replica tests
[FLUENCY-6931] - Doc Impact of FLUENCY-6874 3RDPARTY: upgrade to TIBCO JRE 7 update 65
[FLUENCY-6932] - 3RDPARTY: upgrade to JRE 7 build 67
[FLUENCY-6986] - DATAGRID: deprecate the data grid
Known Issues
A summary of the known issues in this release is below.
For each issue the following is provided:
Description of issue
Workaround (if any)
Plans for resolution
Key
Summary
Public Issue Description
FLUENCY6969
MONITOR: field
name tool tips
not displayed
The tooltips on fields in the monitor are not displayed correctly.
There is no work-around for this issue.
This will be corrected in a future release.
FLUENCY6958
DEPLOY:
Attempting to deploy a JAR file in detached mode with copytarget=false
copytarget=false fails with a message like this:
does not work
with a JAR in
detached mode java -jar deploy.jar detach=true copytarget=false app.jar
...
java.io.FileNotFoundException:
/opt/kabira/run/uepp/nodes/C/classpaths/app_jar0/app.jar (No such
file or directory)
at
com.kabira.platform.MainWrapper.getMainClassNameFromJar(MainWr
apper.java:265)
at
com.kabira.platform.MainWrapper.getMainMethod(MainWrapper.java
:248)
at
com.kabira.platform.MainWrapper.invokeMain(MainWrapper.java:53)
The work-around is to specify the main class in the deployment
command, instead of the JAR file.
FLUENCY6942
DEPLOY: shared
memory
allocation failure
with large JAR
files
Attempting to deploy very large target archives in attached mode will
cause shared memory allocation failures on the node.
The work-around is to use detached mode when deploying very large
target archives.
This issue will be resolved in a future release.
FLUENCY6937
DEVGUIDE:
Incorrect hot
deployment
description
The instructions in deploying the EAR for hot deployment in Chapter 24
Hot Deployment of the Application Developer's Guide are incorrect. In
the Deploy the EAR File that was generated by Web Studio section it
states that the newly generated EAR should replace the EAR in the
<node_name>/deploy/resources/deploy<timestamp>_ directory. This is
incorrect. The EAR should be placed into the
<node_name>/classpath/<jvm_name> directory.
This documentation error will be correct in a future release.
FLUENCY6882
DIST:
JOIN_CLUSTER_P
URGE deletes
instances even if
local node is
active node in
partition.
Enabling a partition with JOIN_CLUSTER_PURGE with the local node as
the active node for both the old and new partition definitions will
delete all instances in the partition on the local node, leaving the
partition empty. This is usually not what is desired.
The work-around is to ensure that JOIN_CLUSTER_PURGE is not used
when enabling partitions on the active node.
Future releases may raise an exception when this is attempted, or raise
a log message and not remove the objects from the partition.
FLUENCY-
DIST: Enable of
Defining and enable partitions in separate transactions may cause a
6880
FLUENCY6847
partition does
not use
distributed
partition lock
protocol.
distributed deadlock.
The work-around is to define and enable partitions in the same
transaction.
This issue will be resolved in a future release.
HA: leave cluster Leaving a cluster on a node that hosts an asynchronous replica partition
with
may generate spurious warnings in the log files while the leave is
asynchronous
executing. For example:
partitions causes
warnings
DST|WARN |7518|csmarshal.cpp(2004)|Cannot access unknown
partition 'Leave Cluster (v541980012961857)' for object type
sandbox.highavailability.LeaveCluster$X on node C.
or
2014-07-03 16:34:59.445105|DST|WARN
|49121|csmarshal.cpp(1125)|Invalid object reference
'1340264:3534776:489635240020741:64' for type
sandbox.highavailability.LeaveCluster$X, operation destroy, event
discarded
These warnings can be safely ignored - they do not affect the outcome
of the leave cluster command.
This issue will be resolved in a future release.
FLUENCY6760
DEPLOY: spaces
in jvmname
causes
deployment
failure
Spaces in the jvmname deployment tool parameter cause the
deployment to fail with a message like this:
INFO: [sandbox.lifecycle.Jvm] started on JVM [Very Important].
Listening for transport dt_socket at address: 28238
Java main class sandbox.lifecycle.Jvm.main exited with an exception.
java.lang.RuntimeException: class sandbox.lifecycle.Jvm not found
at
com.kabira.platform.MainWrapper.getMainMethodFromClass(MainWr
apper.java:319)
at
com.kabira.platform.MainWrapper.getMainMethod(MainWrapper.java
:251)
at
com.kabira.platform.MainWrapper.invokeMain(MainWrapper.java:53)
INFO: JVM [sandbox.lifecycle.Jvm] exited with status [-1].
The work-around is to not use spaces when specifying a JVM name.
This issue will be resolved or audited against in a future release.
FLUENCY6704
RT: Type
verification
message can
contain bogus
fields.
Incorrect type mismatch log messages can be reported for a child of a
parent that has added or removed fields. For example:
2014-05-07 14:58:52.208925|DST|WARN
|30572|csnode.cpp(3220)|Type mismatch detected for type
com.tibco.policyframework.userdata.UserProfile from node MCPCRF1,
remote version '-568284671649893289', local version
'-568284671649893289', mismatch processing will happen on local
node, mismatch reason = 'Mismatch detected for type
com.tibco.policyframework.userdata.UserProfile, version '568284671649893289'. Changed values: Field
policy3uk.managed.concepts.userdata.UserProfile::products was
added: Field
policy3uk.managed.concepts.userdata.UserProfile::defaultProfile was
added.
These messages can be safely ignored.
This will be corrected in a future release.
FLUENCY6694
MONITOR: offset When monitor decoding is in system or primitive, offset fields are not
values not
decoded.
decoded
The work-around is to use object level decoding.
This will be resolved in a future release.
FLUENCY6685
UPGRADE: fails
attempting to
upgrade model
containing
custom catalog
functions
There is no support for hot-deployment of models that contain custom
catalog functions. When attempting to upgrade a model containing
custom catalog functions, it will generate an exception like this:
opt/kabira/users/leon/BEX_1.1.2/RUN/A/classpaths/TestDecisionTable
_ear0/TestDecisionTable.ear failed!!!
java.lang.reflect.InvocationTargetException
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown
Source)
at java.lang.reflect.Method.invoke(Unknown Source)
at
com.tibco.cep.runtime.service.loader.ClassManager.redefineClasses(Cla
ssManager.java:410)
at
com.tibco.cep.runtime.service.loader.ClassManager.commitChanges(Cl
assManager.java:322)
at
com.tibco.cep.runtime.service.loader.BEClassLoader.update(BEClassLoa
der.java:311)
at
com.tibco.cep.runtime.service.loader.BEClassLoader.lockAndUpdate(BE
ClassLoader.java:275)
at
com.tibco.cep.runtime.service.loader.BEClassLoader.lockAndUpdate(BE
ClassLoader.java:281)
at
com.tibco.cep.runtime.service.loader.BEClassLoader.lockAndUpdate(BE
ClassLoader.java:270)
at
com.tibco.cep.runtime.session.impl.RuleServiceProviderImpl.notify(Rul
eServiceProviderImpl.java:1530)
at
com.tibco.cep.repo.hotdeploy.DirectoryDeployer.poll(DirectoryDeploye
r.java:88)
at
com.tibco.cep.repo.hotdeploy.DirectoryDeployer.run(DirectoryDeploye
r.java:48)
at java.lang.Thread.run(Unknown Source)
Caused by: java.lang.UnsupportedOperationException: class
redefinition failed: attempted to delete a method
at sun.instrument.InstrumentationImpl.redefineClasses0(Native
Method)
at sun.instrument.InstrumentationImpl.redefineClasses(Unknown
Source)
... 14 more
There is no work-around for this issue.
This may be resolved in a future release.
FLUENCY6666
DEPLOY:
Attempting to deploy an application to a stopped node causes a null
NullPointerExcep pointer exception similar to this:
tion deploying to
stopped node
Exception in thread "main" java.lang.NullPointerException
at
com.kabira.platform.management.Client.sendMessage(Client.java:185)
at
com.kabira.platform.management.Client.sendPluginRequest(Client.java
:422)
at
com.kabira.platform.management.AdminClient.pollForOutput(AdminCli
ent.java:356)
at
com.kabira.platform.management.AdminClient.waitForCompletion(Ad
minClient.java:451)
at
com.kabira.platform.management.AdminClient.runApplication(AdminCl
ient.java:824)
at
com.kabira.platform.management.AdminClient.deployApplication(Admi
nClient.java:1118)
at
com.kabira.platform.management.AdminClient.doIt(AdminClient.java:9
33)
at
com.kabira.platform.management.AdminClient.main(AdminClient.java:
1256)
This exception can be safely ignored.
This will be resolved in a future release.
FLUENCY6652
FLUENCY6599
QUERY: add
support for
query scope in
generated query
catalog functions
The generated query catalog functions only support local queries. Use a
custom catalog function and Java code to perform a distributed query.
This limitation in the generated query catalog functions will be removed
in a future release.
DEPLOY:
Attempting to deploy an application archive that uses a symbolic link to
symbolic link to a different directory fails with this exception:
EAR target
causes assertion
Muir-Beach >> java -jar deploy.jar reset=false Stress.ear
descriptor=Stress.cdd
INFO: Deployment Tool (deploy.jar) version: [TIBCO BusinessEvents(R)
Extreme 1.2.0 (build 140322)] starting at [Fri Mar 28 12:36:21 PDT
2014]
INFO: JVM remote debugger agent running on [kabiraserver.local:23886] ...
INFO: Node version: [TIBCO BusinessEvents(R) Extreme 1.2.0 (build
140322)]
INFO: Starting [Stress.ear] ...
INFO: [Stress.ear] started on JVM [Stress_ear14].
Listening for transport dt_socket at address: 23886
FATAL: Remote request thread failed for JVM [Stress_ear14]:
RequestThread unexepected parseResponse: <error>
<code>49</code>
<message>A [DSEBuiltin::ExceptionResourceUnavailable:
'java.lang.AssertionError
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.$
getDirectoryImpl(RequestWaiter.java:97)
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.g
etDirectory(RequestWaiter.java)
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.$
getPathImpl(RequestWaiter.java:88)
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.g
etPath(RequestWaiter.java)
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.$
requestCompleteImpl(RequestWaiter.java:55)
at
com.kabira.platform.ktvmnotifier.RequestWaiter$CompletionNotifier.r
equestComplete(RequestWaiter.java)
The work-around is to not use symbolic links to different directories.
This will be resolved in a future release.
FLUENCY-
CONFIG: export
If a configuration class contains a field that is an array of managed
6524
configuration
objects then the configuration file exported using export configuration
does not decode contains object references for the array elements instead of the fully
array of
decoded elements.
managed objects
There is no work-around for this issue.
This will be corrected in a future release.
FLUENCY6409
CONFIG: Autoloading of
configuration
should ignore
files with
unknown suffixes
During node startup, auto-loading of configuration files occurs, which
loads all configuration files in the node auto-load configuration
directories. Currently this process attempts to load all files as
configuration files, regardless of their suffix. This can cause node
startup failures.
The work-around is to ensure that only kcs files are in the configuration
auto-load directories.
This issue will be resolved in a future release.
FLUENCY6332
QUERY: lock
mode of none
takes read lock
Specifying a lock mode of none to the query catalog functions takes a
read lock on the returned object. For example thsi query
Employee.lookupEmployee_ByEmpId("none","1234"); will read-lock the
returned *Employee*object.
There is no work-around for this issue.
This will be corrected in a future release.
FLUENCY6273
CLASSLOADER:
DataError with
anonymous class
implementation
as distributed
POJO method
param.
Using an anonymous class as a parameter to a distributed method call
will fail with an exception like this:
A] com.kabira.platform.DataError: Error packing serialized data.
[A] at com.kabira.platform.Any.serializeObject(Any.java:231)
[A] at
com.kabira.platform.AnyBuffer.writeSerializableStream(AnyBuffer.java:
381)
[A] at com.kabira.platform.AnyBuffer.writeObject(AnyBuffer.java:319)
[A] at
com.kabira.platform.AnyBuffer.writeJavaObject(AnyBuffer.java:337)
[A] at anontest.AnonTest$Target.invokeCommand(AnonTest.java)
[A] at anontest.AnonTest$2.run(AnonTest.java:69)
[A] at com.kabira.platform.Transaction.execute(Transaction.java:457)
[A] at com.kabira.platform.Transaction.execute(Transaction.java:510)
[A] at anontest.AnonTest.runTest(AnonTest.java:58)
[A] at anontest.AnonTest.main(AnonTest.java:26)
[A] at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
[A] at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown
Source)
[A] at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown
Source)
[A] at java.lang.reflect.Method.invoke(Unknown Source)
[A] at
com.kabira.platform.MainWrapper.invokeMain(MainWrapper.java:65)
[A] Caused by: java.io.NotSerializableException: anontest.AnonTest$2
[A] at java.io.ObjectOutputStream.writeObject0(Unknown Source)
[A] at java.io.ObjectOutputStream.defaultWriteFields(Unknown
Source)
[A] at java.io.ObjectOutputStream.writeSerialData(Unknown Source)
[A] at java.io.ObjectOutputStream.writeOrdinaryObject(Unknown
Source)
[A] at java.io.ObjectOutputStream.writeObject0(Unknown Source)
[A] at java.io.ObjectOutputStream.writeObject(Unknown Source)
[A] at com.kabira.platform.Any.serializeObject(Any.java:227)
The work-around for this issue is to use a concrete class instead.
This will be corrected in a future release.
FLUENCY-
QUERY: lookup
Attempting to lookup a concept that was created in the same RTC will
6218
of a concept
created in same
RTC fails
fail. For example:
rule Rules.Work {
attribute {
priority = 5;
forwardChain = true;
}
declare {
Events.DoIt doIt;
}
when {
}
then {
Concepts.Value value = Concepts.Value.Value("", "SameRtc");
value = Value.lookupByName("none", "SameRtc"); // FIX THIS -:
Will return null
System.debugOut("INFO: Work Value is " + value);
}
}
The work-around is to call a catalog function with the newly created
concept as a parameter to force the concept to be created in shared
memory.
This issue will be resolved in a future release.
FLUENCY-
DIST: Loading
StaticDiscovery
Attempting to load a static discovery distribution configuration that
includes the local node causes a node failure. In development builds
6141
entries with self this assertion failure is seen:
included crashes
node.
csnode.cpp:2104: failed C++ assertion 'm_nodeState ==
NodeDiscovered ||
m_nodeState == NodeNotifyUp || m_nodeState == NodeNotifyDown
||
m_nodeState == NodeUp || m_nodeState == NodeDown'
There is no work-around for this issue.
This will be corrected in a future release, either by reporting the
incorrect configuration at audit time, or by supporting this
configuration.
FLUENCY5701
JAVASERVICE:
shutting down at
debugger
breakpoint
causes engine to
fail with an error
Stopping a JVM with a breakpoint set in the debugger will cause the
JVM to exit with an error that requires the node to be removed and reinstalled.
The work-around is to ensure that no break-points are set in a debugger
when shutting down a JVM.
This issue will be resolved in a future release.
FLUENCY5688
JAVASERVICE:
A fatal exception like the following can be seen when using the
jniutils.cpp:138: debugger to examine fields in a managed object following an exception.
failed C++
assertion 'env>ExceptionCheck WARNING in native method: JNI call made with exception pending
() == JNI_FALSE'
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
jniutils.cpp:138: failed C++ assertion 'env->ExceptionCheck() ==
JNI_FALSE'
utl_dump_stack: Start of stack trace for pid 9742, tid 9832
#0 /opt/kabira/kis/distrib/kabira/devlib/libutils_g.so:utl_dump_stack
#1 /opt/kabira/kis/distrib/kabira/devlib/libsyssrv_g.so:dse_bomb(char
const*, char const*, int, DSEBombType)
#2
/opt/kabira/kis/distrib/kabira/devlib/libjavaservice_g.so:JNIUtil::attach
ProxyObject(JNIEnv_*, _jobject*, _jfieldID*)
#3
/opt/kabira/kis/distrib/kabira/devlib/libjavaservice_g.so:JNIUtil::getSlot
Data(JNIEnv_*, _jobject*, _jfieldID*)
#4
/opt/kabira/kis/distrib/kabira/devlib/libjavaservice_g.so:KTVM_GetByt
eField
#5
/opt/kabira/distrib/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/am
d64/libjdwp.so [0x7f7018ae2739]
#6
/opt/kabira/distrib/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/am
d64/libjdwp.so [0x7f7018abe0ab]
#7
/opt/kabira/distrib/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/am
d64/libjdwp.so [0x7f7018ac9806]
#8
/opt/kabira/distrib/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/am
d64/libjdwp.so [0x7f7018adc3c1]
#9
/opt/kabira/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/amd64/se
rver/libjvm.so [0x7f7042849ae7]
#10
/opt/kabira/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/amd64/se
rver/libjvm.so [0x7f7042a880c8]
#11
/opt/kabira/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/amd64/se
rver/libjvm.so [0x7f7042a88218]
#12
/opt/kabira/kis/3rdparty/linux/tibcojre/1.7.0_09_x86_64/lib/amd64/se
rver/libjvm.so [0x7f7042969c00]
#13 /lib/libpthread.so.0 [0x7f70452693f7]
#14 /lib/libc.so.6:clone
WARNING in native method: JNI call made with exception pending
This is caused by a limitation in the debugger integration with JNI.
There is no workaround to this issue.
This issue may be resolved in a future release.
FLUENCY5615
FLUSH: improve Oneways queued to cache never objects will be discarded at commit
handling of one- time.
way methods on
This behavior may be changed in a future release.
cache never
objects
FLUENCY5593
CLASSLOADER:
methods
inherited from
non-managed
base classes not
distributed
Methods inherited from non-managed types are not correctly invoked
on remote nodes if it is a distributed object.
USAGE: help for
scripted plugins
not displayed
The help command on the snapshot and runtimestatistics
administration targets does not work with the domainname parameter.
For example:
FLUENCY5461
The work-around is to override all of the parent types methods and just
invoke them via super in the child type.
This may be resolved in a future release.
administrator adminport=N domainname=Development help snapshot
There is no work-around for this issue.
This will be corrected in a future release.
FLUENCY5447
JAVASERVICE:
Java supports multi-dimensional arrays where each dimension has
Crash when using different
"ragged" multi- sized arrays:
dimensional
arrays.
raggedArray = new byte[2][];
raggedArray[0] = new byte[2];
raggedArray[1] = new byte[8];
However, this will fail at runtime with an exception like this:
FATAL ERROR in native method: Non-array passed to JNI array
operations
at com.kabira.platform.ManagedObject.setReference(Native
Method)
at MyArray.$_init_Impl(Demo.java:14)
at MyArray.<init>(Demo.java:13)
at Demo$1.run(Demo.java:67)
at com.kabira.platform.Transaction.execute(Transaction.java:457)
at com.kabira.platform.Transaction.execute(Transaction.java:510)
at Demo.main(Demo.java:63)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown
Source)
at java.lang.reflect.Method.invoke(Unknown Source)
at
com.kabira.platform.MainWrapper.invokeMain(MainWrapper.java:65)
Node vash failed: Engine x::x with process exit status (6) triggered
abort of node. Node must be reinstalled.
The work-around is to only use multi-dimensional arrays with the same
dimensions.
This will be resolved in a future release.
FLUENCY5400
HA: timer started HA timers started in the constructor of a partitioned class which
in notifier
extends com.kabira.platform.swtimer.TimerNotifier, is not properly
constructor not replicated, and becomes lost upon fail over.
replicated
The work-around is to create the timer notifiers outside of a
constructor.
This will be resolved in a future release.
FLUENCY5287
QUERY: update
in RTC lost if
concepts
reloaded via a
query
Updates to a concept in an RTC are lost if the concept is reloaded via a
query in the same RTC.
The work-around is to flush the concept updates to shared memory by
calling a catalog function with the concept as a parameter before doing
the query.
This may be corrected in a future release. However, it is recommended
that applications do not load a concept into working memory multiple
times in an RTC to avoid the performance impact of loading a concept
into working memory.
FLUENCY5137
DEPLOY:
Application parameters with escaped double quotes do not get passed
application
properly to the application. Examples:
parameters with
embedded
double quote
java -jar deploy.jar MyMain \"param2\"
mishandled
java -jar deploy.jar MyMain param1 "this is \"param2\""
There is no work-around to this issue.
This may be resolved in a future release.
FLUENCY5013
DATAGRID:
Exception
com.tibco.as.spa
ce.ASException:
INVALID_OBJECT
(metaspace_inva
lid) during jvm
termination.
Intermittent data grid errors are seen when shutting down a node. They
look something like:
2012-09-14 15:14:31.567 [
shutdown-hook-job] INFO
runtime.cluster.om.daoprovider - Disconnecting from metaspace
2012-09-14 15:14:32.050 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
DEBUG
runtime.service - CACHEPOLLER : policy3ukpolicy3uk.0.$smtimeouts.CachePoller, Polling... 0, <= 1347632072050
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - java.lang.RuntimeException:
com.tibco.as.space.ASException: INVALID_OBJECT (metaspace_invalid)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.rmi.RemoteFilterRunner.invoke(SourceFi
le:116)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.rmi.RemoteFilterRunner.entrySet(Source
File:46)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.dao.impl.tibas.ASControlDao.entrySet(S
ourceFile:481)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.cluster.scheduler.impl.DefaultCachePolle
r.getExpiredKeys(DefaultCachePoller.java:111)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.cluster.scheduler.impl.DefaultCachePolle
r.run(DefaultCachePoller.java:82)
2012-09-14 15:14:32.074 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.session.BEManagedThread.execute(BEManaged
Thread.java:458)
2012-09-14 15:14:32.089 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.session.BEManagedThread.run_repeating(BEMa
nagedThread.java:436)
2012-09-14 15:14:32.089 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.session.BEManagedThread.run(BEManagedThre
ad.java:299)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - Caused by:
com.tibco.as.space.ASException: INVALID_OBJECT (metaspace_invalid)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
Native.API_Space_InvokeMembers(ApiSpace.cpp:481)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
Native.Java_com_tibco_as_space_impl_NativeImpl_spaceInvokeMemb
ers(JRemoteService.cpp:62)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.as.space.impl.NativeImpl.spaceInvokeMembers(Native
Method)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.as.space.impl.ASSpace.invokeAll(ASSpace.java:776)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.as.space.impl.ASSpace.invokeSeeders(ASSpace.java:740)
2012-09-14 15:14:32.090 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.as.space.impl.ASSpace.invokeSeeders(ASSpace.java:734)
2012-09-14 15:14:32.094 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.as.kit.rmi.InvocationHelper.buildAndInvoke(SourceFile:1
40)
2012-09-14 15:14:32.094 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - at
com.tibco.cep.runtime.service.rmi.RemoteFilterRunner.invoke(SourceFi
le:111)
2012-09-14 15:14:32.097 [y3uk-policy3uk.0.$smtimeouts.CachePoller]
ERROR
root - ... 7 more
2012-09-14 15:14:32.317 [
shutdown-hook-job] INFO
runtime.cluster.om.daoprovider - Disconnected from metaspace
These errors can be ignored.
This will be corrected in a future release.
FLUENCY4813
JDWP: Setting a
field in a
Managed object
without an active
transaction
results in
'Unexpected
JDWP Error: 115'
Setting a field in a managed object without an active transaction in a
Java debugger causes errors.
For example, if mo is managed object, doing the following in a
debugger:
set mo.m_id = 456
causes these errors to be reported.
main[1] com.sun.jdi.InternalException: Unexpected JDWP Error: 115
mo.m_id = 456 = null
main[1] WARNING in native method: JNI call made with exception
pending
WARNING in native method: JNI call made with exception pending
WARNING in native method: JNI call made with exception pending
...
There is no work-around for this issue.
This may be resolved in a future release.
FLUENCY4714
DEPLOY: Error
when overriding
global variable
on the
command-line.
There is currently no documented support for overriding global
variables on the command line.
The work-around is to use a properties file to set the global variable,
and
specify the properties file on the deployment command-line like so:
java -jar deploy.jar servicename=abc P.ear descriptor=P.cdd -p
<properties file>
Support for this feature will be added in a future release. The final
support may change the command syntax used in the work-around
described above.,
FLUENCY4569
FLUENCY4537
ASTA: The BE-X
administrator is
not rebranded
The BE-X Administrator was not rebranded. It still references
ActiveSpaces Transactions.
CLASSLOADER:
autoboxing and
valueof
ClassCastExcepti
on
Using valueOf to update an item in an array of Long in a Managed
Object fails with a ClassCastException. For example, this snippet fails
with:
This will be corrected in a future release.
java.lang.ClassCastException: [J cannot be cast to [Ljava.lang.Long;
import com.kabira.platform.Transaction;
import com.kabira.platform.annotation.Managed;
public class Bug
{
public static void main(String[] args)
{
new Transaction()
{
public void run()
{
Unhappy clown = new Unhappy();
System.out.println(clown.longs[0]);
}
}.execute();
}
@Managed
static class Unhappy
{
Long longs[] = new Long[1];
Unhappy()
{
longs[0] = Long.valueOf(42);
}
}
There is no work-around for this issue.
This will be resolved in a future release.
FLUENCY4489
MONITOR:
Cannot display
multidimensional
arrays.
The shared memory monitor does not handle java arrays with multiple
dimensions, i.e. running the following example:
import com.kabira.platform.Transaction;
import com.kabira.platform.annotation.Managed;
@Managed
class Foo
{
public char[][] m_char;
Foo()
{
m_char = new char[5][5];
char val = 'a';
for (int i = 0; i < 5; i++)
{
for (int j = 0; j < 5; j++)
{
m_char[i][j] = val;
val++;
}
}
}
}
public class Demo
{
public static void main(String[] args)
{
new Transaction("Create")
{
public void run()
{
new Foo();
}
}.execute();
System.out.println("Main: done");
}
}
Then running the monitor and navigating to the foo instance will result
in a
Address out of bounds error displayed for m_char.
A workaround is to use swdiag to display the object data.
This will be corrected in a future release.
FLUENCY4474
RT: Creating
key'd instances
via reflection
fails.
Attempting to create a managed object with keys defined using
reflection fails with a duplicate key exception.
There is no work-around for this issue.
This will be resolved in a future release.
FLUENCY3956
ASTA:
reset/delete
buttons handled
incorrectly
following drag &
drop
Changing the order of items in dialog boxes that support drag & drop
does not managed the reset and delete button states correctly. The
reset button is not enabled, and in some cases the delete button is
incorrectly enabled.
The work-around is to close and re-open the dialog.
This issue will be resolved in a future release.
FLUENCY3859
FLUENCY3757
STATISTICS:
object update
latency does not
include
partitioned
objects
The object update latency reported for distributed and partitioned
objects does not include the network latency.
There is no work-around for this issue.
This will be corrected in a future release.
ASTA:
The performance monitors in the administrator report a
Channel.Security. Channel.Security.Error when accessed by a client in a different domain
Error starting
than the node is running.
monitors
There is no work-around for this issue.
This wil be resolved in a future release.
FLUENCY3678
DATAGRID: Crash Attaching JProfiler 6.2.4 to a running AST JVM caused a SEGV crash due
during JProfiler to a library symbol conflict between JProfiler and DataGrid.
startup
There is no work-around for this issue.
A future release of JProfiler (version 6.2.5) will contain a fix for this
issue.
FLUENCY3436
DATAGRID:
spurious output
to stdout
Certain log messages from the data grid library will display to the
console, and cannot be disabled. The "member left:", "ip_address:",
and "Connected metaspace ..." messages should be ignored. This will be
corrected in a future release.
FLUENCY3408
ASTA: numeric
and data
columns in tables
not sorted
correctly
Tables displaying numeric and date data are not sorted correctly. They
are currently sorted in character sort order.
ASTA: support
multi-master
merge
There is no support for resolving a multi-master scenario from
ActiveSpaces® Transactions Administrator.
FLUENCY3382
There is no work around for this issue.
This will be resolved in a future release.
The work-around is to use the administrator command line.
This issue will be resolved in a future release.
FLUENCY3293
DATAGRID: class If the class-load action for a keyed, grid-backed type occurs outside any
load for key
transaction context, and there are instances of the type to be restored
restore requires from the data grid, a java.lang.NoClassDefFoundError may be thrown.
a transaction
This occurs as AST will automatically restore any instances of keyed
types that are found in the data grid. Currently, this restore action
requires a transaction. This restriction will be removed in a future
release.
FLUENCY3225
DATAGRID:
cardinality for
flushable types
updated at
Cardinality for data grid backed managed objects that are not always
cached in shared memory does not include any objects created in the
current transaction - it only returns the cardinality for completed
transactions.
prepare time
There is no work-around for this inconsistency with non-data grid
backed managed objects.
This will be resolved in a future release.
FLUENCY3181
DATAGRID:
When running a query against a space, certain illegal query filters will
invalid query not not fail as expected. Specifically, a query filter that consists of a single
token will find zero elements, rather than returning an error. An
failing
example of a such a filter would be query space name=foo filter=abc.
There is no work-around for this issue.
This will be resolved in a future release.
FLUENCY3073
KDM:
configuration
restore: incorrect
date conversion
when timezone
offsets of -00xx
The Domain Manager's configuration cache incorrectly converts
negative timezone offsets in a node's installation time to a positive
offset. For example a node installation timezone offset of -0030 is
converted to 0030. This may cause configuration restore to incorrectly
restore, or not, configuration to a node.
There is no work-around for this issue.
This will be corrected in a future release.
FLUENCY2855
KDM: cannot
remove switch
using domain
manager
It is not possible to remove a node using the domain manager. Attempt
to do so will fail with the following error message:
switchadmin servicename=domainmanager domainnode=replica stop
switch
guest@kabira-server:~$ switchadmin servicename=domainmanager
domainnode=replica stop remove
Invalid response data stream received from remote node.
The work around is to execute the remove directly to the target node
without the domain manager.
This will be resolved in a future release.
FLUENCY2509
JAVASERVICE:
The -showversion JVM option will cause the Fluency JVM to fail during
JVM fails to start startup.
if given the -
showversion JVM There is no work-around for this issue.
option
This will be corrected in a future release.
FLUENCY2491
NODE: display
node does not
show node
description if
MDNS not
running
switchadmin display switch and Kabira Manager do not include the
node description if service discovery is not running.
With MDNS running here is the output from display switch.
switchadmin adminport=2001 display switch
Node administration for primary:
Host: localhost
Administration Port: 2001
Service Name: "primary"
Node Name: "primary"
Engine fluency::fluency
State: running
Build type: DEVELOPMENT
Contains JVM: No
Principals: guest
Active Principal: guest
Application: kabira/fluency
Node Path: /opt/kabira/run/fluency/nodes/primary
Installation Path: /opt/kabira/kis
Node Description: Fluency Development
Here is the output without MDNS running.
switchadmin adminport=2001 display switch
Node administration for primary:
Host: localhost
Administration Port: 2001
Service Name: "None"
Node Name: "primary"
Engine fluency::fluency
State: running
Build type: DEVELOPMENT
Contains JVM: No
Principals: guest
Active Principal: guest
Application: kabira/fluency
Node Path: /opt/kabira/run/fluency/nodes/primary
Installation Path: /opt/kabira/kis
There is not work-around for this issue.
This will be corrected in a future release.
FLUENCY2482
ADMINISTRATIO If the MDNS daemon is disabled, no error message is displayed to the
N: no error
user when using servicename with administrator. administrator just
message
returns a non-zero error code. For example:
reported if
MDNS disabled
administrator servicename=xxxx display node
echo $?
1
An error message will be returned in a future release.
FLUENCY2446
JAVASE: JDWP
can hang during
A Java debugger can hang if the wrong transaction is chosen by the
JDWP library when changing threads while single stepping through a
stepping in
debugger
method in a transaction. When the wrong transaction is selected, the
debugger will hang on the next single step.
The work-around is to not switch threads while single stepping through
a method in a transaction.
This will be resolved in a future release.
FLUENCY2426
KDM: cannot
start node
through KDM
Attempting to start a node through KDM will fail. For example,
executing this command:
switchadmin hostname=kabira-server adminport=2000
domainnode=replica username=guest password=guest start switch
will fail with:
Invalid response data stream received from remote node.
The work around is to send the command directly to the node being
started, by passing KDM. For example:
switchadmin hostname=kabira-server adminport=2003
username=guest password=guest start switch
Waiting for application to start
Components started
Loading configurations
Auditing security configuration
Host: kabira-server
Administration Port: 2003
Service Name: "replica"
Node Name: "replica"
This will be resolved in a future release.
FLUENCY1925
JAVASERVICE:
invalid JVM
options causes
node to crash
Invalid JVM options causes the node to crash. For example (note
duplicate -Xms options):
2010-01-21 13:21:08.162242|KVM|INFO
|31637|ktvmmgr.cpp(1655)|Creating a JVM
VM options:
Dcom.kabira.application.class.path=/opt/kabira/run/fluency/nodes/pri
mary/classpaths/snippets_jar3
-Xms256m
-Xms512m
agentlib:jdwp=transport=dt_socket,address=63315,server=y,suspend=n
-Dcom.kabira.application.detach=true
Djava.system.class.loader=com.kabira.platform.classloader.ClassLoader
-Dcom.kabira.platform.buildType=DEVELOPMENT
-Dcom.kabira.platform.management.port=2001
,,.
2010-01-21 13:21:08.199367|KVM|INFO
|31637|notifiers.cpp(130)|installNotifiers called with VMID 31621
2010-01-21 13:21:08.236792|USR|FATAL|31637|engine.cpp(74)|An
illegal call to exit() was detected in process 31621 with active engine
instances. You must use the stop() method in swbuiltin to terminate
engines.
This will be handled more gracefully in future releases.
FLUENCY1922
JAVASERVICE:
iterative JVM
start/stop
memory leak
There is a shared memory leak of approximately 200 kilobytes, seen
when JVMs are launched, stopped and then left stopped. Some of this
leak also remains even after the JVM is removed.
This is only an issue in the development process where many JVMs are
deployed (one for each launch of the application from the deploy tool).
This will be resolved in a future release.
FLUENCY1129
CLASSLOADER:
inherited fields
used in keys are
not audited for
"final"
An inherited field used in a Java key will not be audited for "final"-ness.
For instance, the following should not be legal:
public class A {
int i;
}
@Managed
@Key(name ="test", fields = {"i"})
public class B extends A { }
FLUENCY619
ASTA: domain
manager node
running ASTA can
be
stopped/remove
d
It is possible to stop/remove the Domain Manager node that runs the
ActiveSpaces® Transactions Administrator. If this happens the web
client becomes non-responsive. The Domain Manager node should not
be stopped using ActiveSpaces® Transactions Administrator.
The workaround is to restart the Domain Manager node if this happens.
Stopping the Domain Manager node using ActiveSpaces® Transactions
Administrator will be prevented in a future release.
FLUENCY544
CLASSLOADER: Arrays of enums stored in a ManagedObject store a null as the first
can't store a null value in the enum declaration.
reference in a
Java enum array For example, the following assert will fail.
enum Relic { BELL, BOOK, CANDLE };
class Holder extends ManagedObject
{
Relic[] m_relic;
void foo()
{
m_relic = new Relic[10];
assert( m_relic[0] == null );
// m_relic[0] is actually set to BELL
}
}
This problem will be corrected in a future release.
FLUENCY542
CLASSLOADER: Arrays of strings in a ManagedObject store a null as an empty string.
can't store a null
For example, the following assert will fail.
reference in a
Java String array
public class StringHolder extends ManagedObject
{
String[] m_string;
void foo()
{
m_string = new String[10];
assert( m_string[0] == null );
//
// m_string[0] will return a valid string of length 0
//
}
}
This problem will be corrected in a future release.
FLUENCY287
ASTA: Browser
refresh logs you
Using ActiveSpaces® Transactions Administrator with any browser,
when you reload the current page by any means (for example, clicking
on the browser's "Reload" button, or pressing ctrl-R), ActiveSpaces®
out
Transactions Administrator logs you out of all nodes and returns you to
the original login screen.
This issue will be fixed in a future release.
FLUENCY271
ASTA: apache
processes not
stopped when
remove switch
used
Stopping a domain manager node using switchadmin remove switch
does not stop the Apache server processes.
The work-around is to use switchadmin stop switch before switchadmin
remove switch, or to manually kill the Apache server processes.
This will be resolved in a future release.
FLUENCY84
JAVASERVICE:
Invalid
com.sun.manage
ment.jmxremote.
port number
crashes node
An invalid value to a system property when starting jconsole, for
example:
Error: Invalid com.sun.management.jmxremote.port number:
192.168.71.128:9876
Brings down the Fluency node with this message:
2008-10-09 14:36:49.629085|KVM|INFO
|10069|ktvmmgr.cpp(1241)|Creating a JVM
VM options:
-Dcom.sun.management.jmxremote.port=192.168.71.128:9876
-Dcom.sun.management.jmxremote.authenticate=false
agentlib:jdwp=transport=dt_socket,address=16106,server=y,suspend=n
Djava.system.class.loader=com.kabira.platform.classloader.ClassLoader
-Dkabira.engine.buildType=PRODUCTION
Class path:
/opt/kabira/kis/distrib/kabira/component/ktvm/java/ktvm.jar
/opt/kabira/kis/3rdparty/generic/bcel/5.2/bcel-5.2.jar
/opt/kabira/run/fluencydev/nodes/primary/classpaths/scratch_Main5
java/swbuiltin.jar
java/kis.jar
java/appstats.jar
java/event.jar
java/scheduler.jar
java/switchadmin.jar
java/logger.jar
java/diskmonitor.jar
java/ktvmadmin.jar
java/netchannel.jar
java/ha.jar
java/kdds.jar
java/ktvmnotifier.jar
Initial load classes:
com.kabira.platform.ktvmnotifier.RequestWaiter
2008-10-09 14:36:49.642968|KVM|INFO
|10069|notifiers.cpp(43)|installNotifiers called with VMID 10069
2008-10-09 14:36:50.526578|USR|FATAL|10088|engine.cpp(73)|An
illegal call to exit() was detected in process 10069 with active engine
instances. You must use the stop() method in swbuiltin to terminate
engines.
This error will be handled without bringing the node down in a future
release.
FLUENCY63
DEPLOY: client
fails to resolve
services on
fluency.jar fails to resolve Fluency node services running on the server
appliance hosted on Mac OS X.
The work-around is to use host names and port numbers instead of
macosx
service names.
This will be resolved in a future release.
FLUENCY-7 CLASSLOADER: Attempting to integrate the NetBeans profiler with Fluency causes this
java.lang.IllegalSt fatal exception:
ateException:
recursive
invocation
[primary] Profiler Agent: Initializing...
[primary] Profiler Agent: Options: >/opt/kabira/run/fluencydev/deploy/profiler-server-linuxamd64/lib,5140<
[primary] Profiler Agent: Initialized succesfully
[primary] Error occurred during initialization of VM
[primary] java.lang.IllegalStateException: recursive invocation
[primary] at
java.lang.ClassLoader.initSystemClassLoader(ClassLoader.java:1303)
[primary] at
java.lang.ClassLoader.getSystemClassLoader(ClassLoader.java:1286)
[primary]
at java.lang.ClassLoader.defineClass1(Native Method)
[primary]
at java.lang.ClassLoader.defineClass(ClassLoader.java:620)
[primary] at
java.security.SecureClassLoader.defineClass(SecureClassLoader.java:12
4)
[primary] at
java.net.URLClassLoader.defineClass(URLClassLoader.java:260)
[primary] at
java.net.URLClassLoader.access$000(URLClassLoader.java:56)
[primary] at
java.net.URLClassLoader$1.run(URLClassLoader.java:195)
[primary]
Method)
at java.security.AccessController.doPrivileged(Native
[primary]
at
java.net.URLClassLoader.findClass(URLClassLoader.java:188)
[primary]
at java.lang.ClassLoader.loadClass(ClassLoader.java:306)
[primary] at
sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:276)
[primary]
at java.lang.ClassLoader.loadClass(ClassLoader.java:251)
[primary] at
java.lang.ClassLoader.loadClassInternal(ClassLoader.java:319)
[primary]
at java.lang.Class.forName0(Native Method)
[primary]
at java.lang.Class.forName(Class.java:247)
[primary] at
java.lang.SystemClassLoaderAction.run(ClassLoader.java:2015)
[primary]
Method)
at java.security.AccessController.doPrivileged(Native
[primary] at
java.lang.ClassLoader.initSystemClassLoader(ClassLoader.java:1311)
[primary] at
java.lang.ClassLoader.getSystemClassLoader(ClassLoader.java:1286)
[primary]
[primary] Current thread is 139832737576688
[primary] Dumping core ...
[primary] Caught fatal signal 6 (Aborted)
[primary] utl_dump_stack: Start of stack trace for pid 10825
There is no work-around for this failure.
This issue appears to be a NetBeans profiler issue.