Fixed Bugs and Known Issues in 5.1

                                               

Summary: A list of fixed bugs and known issues for GigaSpaces 5.1 (build 1603).

Overview

Below is a list of fixed bugs since GigaSpaces 5.0 SP1 (build 1410), and a list of known issues:

Core

  • CORE-1 – fixed problem when using JRockit that causes a JVM crash.
    BEA JRockit(R) version – R26.0.0-189-53463-1.5.0_04-20051122-2040-linux-ia32. See
    http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6380091. The problem occurs when a non-existent thread is being signaled.
  • CORE-3 – fixed problem when using the sync-replication cluster schema with transactions.
    Added new cluster property – <hold-txn-lock> controls the behavior of the order of the locking sequence between the source space, target space, and client.
  • CORE-4 – fixed problem when using the IWorker and space filter with a secured space. A non-secured proxy is provided to allow these plugins to interact with the space.
  • CORE-5 – fixed problem with double Replication-Input-Filter invocations. Fixed problem when using the ReplicationInputFilter in sync-rec-ack mode.
  • CORE-73 – fixed problem with duplicated Entries or templates when performing memory recovery.
  • CORE-81 – fixed problem when using replication Matrix, that opens multicast ports when only unicast is required. You may control this behavior using:
    -Dcom.gs.replication.disable_multicast_adaptor=true
  • CORE-89 – fixed problem when using the IWorker interface. After the space restart operation, the IWorker implementation should be restarted as well.
  • CORE-95 – fixed problem when using NotifyDelegatorMultiplexer with Broadcast mode. Passing a null value for the first indexed field will register notifications to all partitions.
  • CORE-110 – fixed problem with unclear exception messages, when using timeout operations or operations that use transactions with partitioned spaces (using the hash-based policy), and broadcast-if-null mode.
  • CORE-124 – fixed problem when calling the clean operation. The client type table at the client side was not cleaned.
  • CORE-126 – fixed problem when using the Map API in Local Cache mode – self-originated take and lease expiration events were ignored.
  • CORE-136 – fixed problem when NPE is thrown as a result of using a non-clustered proxy that is retrieved from a clustered proxy for embedded space, and using the take operation with timeout.
  • CORE-137 – fixed problem with cleaning recovered Entries.
  • CORE-138 – fixed problem when EntryAlreadyInSpace is thrown as a result of a take and write operation for an Entry with the same UID under the (same) transaction.
  • CORE-160 – Fixed problem when NumberFormatException is thrown due to converting <objectID> from string to long value.
  • CORE-161 – fixed performance problem when using primary-backup cluster schema with the Jin Transaction Manager.
  • CORE-164 – fixed problem when using the primary-backup schema with FIFO-based notifications. The broadcast notify should be disabled when using this schema.
  • CORE-169 – fixed problem when a space is registered into Jini LUS with multiple ServiceItem names. The matching should be done using all ServiceItems .
  • CORE-173 – fixed problem when a transaction rolls back, causing UID conflicts.
  • CORE-196 – fixed problem when registering for the take operation notification in FIFO mode.
  • CORE-199 – fixed problem with a cluster's start time when using the sync-replicated or the async-replicated schema.
  • CORE-201 – fixed problem with ClosedChannelException thrown when the server sends a response to the disconnected client.
  • CORE-222 – fixed problem when using partitioned-backup and local space load-balancing – the wrong backup was selected in a failover.
  • CORE-225 – fixed problem when performing a query for Entries that were previously deleted and rewritten.
  • CORE-226 – fixed problem with dynamic classes loading – loading classes using Thread.currentThread().
    getContextClassLoader().loadClass(), and not Class.forName().
  • CORE-236 – fixed problem when changing the primary key value with update operations, or when assigning null as the primary key value with write operations (relevant when using CacheLoader/Store).
  • CORE-240 – fixed problem with static client TypeTable that throws java.lang.IllegalArgumentException if the space proxy has been loaded by a different ClassLoader.
  • CORE-249 – the UpdateModifiers.UPDATE_OR_WRITE modifier used when calling the update operation for a POJO.
  • CORE-253 – fixed problem when using takeMultiple() with a local space, causing a recursive call.
  • CORE-259 – fixed problem with expired Entries visible to clients when using a persistent space.
  • CORE-293 – fixed problem when the replicator shows empty messages.
  • CORE-353 – fixed problem when client performs write operations under transactions using a clustered space, throwing SyncReplicationException instead of TransactionException.
  • CORE-368 – fixed problem when using an indexed file persistent with <memory_usage> enabled, and having a low amount of JVM memory. In this case, the user's query returns partial results. This is due to the eviction of Entries in the cache, that effects the result set returned.
  • CORE-378 – fixed problem when using persistent space with force cold init – force cold init was not invoked properly.
  • CORE-397 – fixed problem when performing recovery from persistent space. Due to race conditions, an incorrect redo key is sent, causing the redo log to be out of synch.
  • CORE-410 – fixed problem when using async-replication, that throws UnknowTypeException as a result of NPE.
  • CORE-411 – fixed problem when receiving duplicated UIDs as a result of replication recovery.
  • CORE-183 – fixed problem when using FIFO with multiple transactions, causing notifications to arrive in the wrong order.
  • CORE-388 – fixed problem when restarting a cluster using persistent spaces (indexed file or JDBC SA) in the wrong order. Only the space that was shut down while running in active mode can be started first. You may control this behavior using the following property:
    com.gs.cluster.replication.secure-restart=true

Service Grid

  • SERVICEGRID-11 – fixed problem with MemoryMonitor reports incorrect utilization.
  • SERVICEGRID-28 – fixed problem with the CLI home directory.
  • SERVICEGRID-29 – fixed problem with the Deployment Descriptor's Editor modality.
  • SERVICEGRID-36 – fixed problem with the storage capability not being registered.
  • SERVICEGRID-39 – fixed problem with a space name that cannot have " " signs.
  • SERVICEGRID-42 – fixed problem with override XML files that are not read by the Service Grid when using GSC.
  • SERVICEGRID-47 – fixed problem with GSM resetting codebase when computehttpmd is used.
  • SERVICEGRID-62 – fixed problem that occurs when Job Submission UI does not show up.
  • SERVICEGRID-64 – fixed problem with prime example – getting NotSerializableException.
  • SERVICEGRID-76 – fixed problem with Webster – connection abort: socket write error.
  • SERVICEGRID-84 – fixed problem with NPE thrown when logging an association injection failure for a broken association.
  • SERVICEGRID-90 – removed unesseary jars in spring-platform.xml.
  • SERVICEGRID-95 – fixed problem with deadlock while running Service Grid UI.
  • SERVICEGRID-97 – fixed problem when using the partitioned-sync2backup-cluster-schema – you cannot specify zero backups.
  • SERVICEGRID-6 – added Deploy directory.
  • SERVICEGRID-19 – fine grained relocation.
  • SERVICEGRID-20 – added Discovery Group Management in UI.
  • SERVICEGRID-21 – added Launch GSC and GSM services through SG UI.
  • SERVICEGRID-24 – added Create ServiceBeanContextFactory.
  • SERVICEGRID-25 – added AssociationManagement Notification.
  • SERVICEGRID-26 – combined gsc-dl.jar and gsm-dl.jar.
  • SERVICEGRID-27 – defined a codebase for GigaSpace.
  • SERVICEGRID-37 – added Colocated space.
  • SERVICEGRID-46 – using a refid when declaring an Association instead of using the OperationalString attribute.
  • SERVICEGRID-67 – simplified configuration for groups, locators and system properties for the admin tools.
  • SERVICEGRID-69 – added a directory to the GSC default classpath.

Cluster

  • APP-151 – fixed problem with the <repl-original-state> element which had no effect on runtime, since it was located at the wrong location in the cluster schema – it should be under the <repl-policy>.
  • CORE-91 – fixed problem when using the primary-backup schema using the persistent schema. Data was not recovered properly during the restarting of the space.
  • APP-189 – added <recovery-chunk-size> to the replication section in the ownership-replicated schema.
  • APP-265 – spring GigaSpaces load-balancing through remoting.
    Solved the problem with using GigaSpaces remoting through a Spring engine, to get load-balancing.
  • APP-275 – added support for the Space URL's create=false attribute.
    With previous versions, if create was used, an embedded space proxy was created. The create=false allows you to get a previously created embedded proxy, ignoring the passed Space URL properties, such as the space schema.
  • APP-291 – fixed problem with <GigaSpaces Root>/services/gigaspaces.config (a Jini service configuration file) using hard-coded values.
    The space schema, space URL, etc. used hard-coded values, causing the gsServer to ignore the Space URL passed or set in the gsServer. All the relevant values use variables.
  • APP-306 – fixed problem with the ping command. Jini Unicast was not supported.
  • APP-334 – fixed problem when loading the static cluster configuration XML file using http.
  • APP-353 – added the ability to pass cluster schema variable values using a properties file or a properties object.

Persistency

  • DB-27 – fixed problem when flushing data into the file when performing commit.
  • DB-28 – fixed problem when converting ByteArray to Double.
  • DB-60 – fixed problem with JDBC SA when constructing a dynamic query.
  • DB-21 – fixed problem when using the enum data type with a persistent space while using JDBC SA.

Query Engine

  • DB-86 – fixed problem when using ORDER BY with multiple columns.
  • DB-93 – fixed problem when using isSigned causing NPE.
  • DB-5 – fixed problem using a query that includes a combination of parenthesis, or an operation with an AND operation. For example:
    select * from tab where (col1 = 1) and (col3 = 3) and (col2=1 or col2=2 or col2=3)
  • DB-10 – fixed problem when using take and TakeMultiple with SQLQuery.
  • DB-18 – fixed problem when using GROUP BY with null fields.
  • DB-19 – fixed problem when using SUM.
  • DB-29 – fixed problem when using COUNT and GROUP BY. For example:
    select count(*), col1 from tab group by col1
  • DB-31 – fixed problem when using FIFO with indexed file persistency.
  • DB-45 – fixed problem when performing Insert of an integer value into a double column using SQL – the value is saved as an integer.
  • DB-47 – fixed problem when performing field quality queries. For example:
    select * from place p1,place p2 where p1.country=p2.country
  • DB-48 – fixed problem when using the same column twice as part of the WHERE phrase with the AND condition. For example:
    select * from table where col>1 AND col >1
  • DB-49 – fixed problem when * appears as part of the table name. For example:
    select t_date.*, t from t_date where t<='01:00:00'
  • DB-50 – fixed problem when comparing boolean values. For example:
    select * from t_bit where b1=false or b2=true
  • DB-51 – fixed problem when an unclear message is thrown when having three tables join.
  • DB-56 – fixed problem when using Alias as part of the select.
  • DB-58 – fixed problem when having a query with with ")" and AND, causing a Too many children on binary expression error.
  • DB-59 – Fixed problem when a query uses equals without alias in the join query causing an Unknown table in condition error.
  • DB-61 – improved the error message when using non-primitive objects (non Java.Lang.*) in an Entry, combined with an SQLQuery; and the non-primitive objects class implements the Comperable interface.
  • DB-69 – added support for ascending and descending sort using JDBC API.
  • DB-83 – fixed problem when NPE is thrown with strings compare query:
    select p1.country from place p1,place p2 where p1.country=p2.country
  • DB-95 – fixed problem when having a Delete ... Where... query throwing ClassCastException. For Example:
    DELETE from Message WHERE Message.m_counter<10
  • DB-99 – fixed problem when the SELECT query causes a Too many children on binary expression error.
  • DB-108 – added support for the IN SQL keyword for the JavaSpaces SQLQuery and the JDBC API.
  • DB-109 – fixed problem with non equal conditions.
  • DB-112 – fixed problem when the GROUP BY column name is not in the result, causing an exception to be thrown. For example:
    select sum(col1) from Tab group by col2
  • DB-118 – fixed problem when having a dynamic query using delete and rownum. For example:
    delete from tab rownum=1 and col1< ?
  • PACK-53 – fixed problem when using the JDBC driver – groups name was not supported.

JMS

  • APP-2 – fixed problem with JMS logging issues – when logs directory was not found.
  • APP-3 – fixed problem with a null destination in the producer. In send() use the passed destination name.
  • APP-179 – fixed problem with ClassCastException when receiving an ObjectMessage.
    The ObjectMessage did not handle the serialization of the object body correctly. Fixed the JMS examples to support ObjectMessage as well.
  • APP-213 – added support for the JMS StreamMessage.
  • APP-5 – fixed problem when the JMSReplyTo header field returns null.
  • APP-7 – implemented hashCode() and equals() for the com.j_spaces.jms.GSTopicImpl and com.j_spaces.jms.
    GSQueueImpl, using the destination name.
    This allows a partitioned space to load-balance topic and queue messages implicitly into different partitions, based on the destination name.
  • APP-6 – marked the JMSDestination field as indexed to support partitioned space.
  • APP-14 – fixed problem with asynchronized received messages. Using separate thread to release the notify thread.
  • APP-17 – fixed problem when using the GSConnectionFactory.findSpaceProxy(). Subsequent calls after the initial call didn't update the internal proxy.
  • APP-329 – fixed problem with anonymous publishers when no destination is passed to the publisher.
  • APP-12 – compression for the JMS Text message body data is supported. The message body is compressed only if its size is larger than a configured value (threshold). The message body is compressed when it is sent and decompressed when message it is received.
    This property is configured via the system property com.gs.jms.compressionMinSize. The value assigned should be in bytes. The defualt value is 500000 (0.5 MB).
  • APP-296 – added system vars placements into the jms-config.xml for the following elements:
    • com.gs.jms.container-prefix – default is myJMS_container.
    • com.gs.jms.space-prefix – default is myJMS.
    • com.gs.jms.url – default is /./myJMS?schema=cache.
    • com.gs.jms.debug – default is false.
  • APP-303 – direct JNDI InitialContext lookup for JMS admin objects is now supported.
    Added ability to create an InitialContext (using com.sun.jndi.rmi.registry.RegistryContextFactory) and lookup all JMS administrated objects using the full binded name.
    For example, to look up connection factory which has been registered while loading a space:
    initialContext.lookup("GigaSpaces;myJMS_container;myJMS;GSQueueConnectionFactoryImpl");
    initialContext.lookup("GigaSpaces;myJMS_container;myJMS;GSTopicConnectionFactoryImpl");

    Or for topic/queue discovery:

    initialContext.lookup("GigaSpaces;myJMS_container;myJMS;jms;destinations;MyQueue");
    initialContext.lookup("GigaSpaces;myJMS_container;myJMS;jms;destinations;MyTopic");
    These admin objects need to be registered already when a space is loaded.
    In order to load an embedded space together with the JMS client application (for scenarios such as P2P), you need to use methods like GSJMSAdmin.getTopicConnectionFactory() in order to load the space, and make sure
    the <overwrite-space-url> in the config/jms-config.xml is set with an embedded space URL value.

Space Browser

  • APP-18 – fixed problem with wrong extended indexing value in the Classes view.
  • APP-38 – the com.gs.jini_lus.groups property is used instead of the com.gs.browser.lookup.groups. com.gs.browser.lookup.groups has been removed from the SpaceBrowser script.
  • APP-20 – fixed problem with the Create New Space window – the Persistent Options section was editable by default, even though the Persistent checkbox was unchecked.
  • APP-224 - fixed problem with viewing secured spaces – ClassCastException was thrown while trying to add the space node to the tree.
  • APP-37 – fixed problem with wrong selected tree nodes when clicking the cluster member.
  • APP-287 – fixed problem when the benchmark read operation invoked called using a transaction.
  • APP-42 – fixed problem when Jini LUS is not running, preventing the space browser from displaying the cluster members automatically.
  • APP-47 – changed the background color of the cluster nodes to system gray.
  • APP-91 – fixed problem when running containers that aren't running are still displayed as part of the existing containers list.
  • APP-130 – fixed problem with NPE thrown when creating a cluster, using the primary_backup schema.
  • APP-304 – fixed problem with 100% CPU utilization when benchmark is stopped using the Stop button.
  • APP-107 – SpaceBrowser.bat script starts using javaw.exe and not java.exe.
  • APP-138 – fixed problem with NPE when clicking the Next button in the first panel of the cluster wizard.
  • APP-157 – fixed problem with the Cluster root node that collapses when another cluster space member is discovered.
  • APP-428 – fixed problem with browser responsiveness when using the address bar.
  • APP-154 – fixed problem with 100% CPU utilization when trying to add a tree node, causing infinite NPE.
  • APP-97 – fixed the default URL address to: rmi://host/myCache_container/myCache.
  • APP-486 – fixed problem when the same space node is displayed twice.
  • APP-476 – fixed problem when using Windows with German Regional Options. The benchmark graph component does not display the graph.
  • APP-456 – fixed problem when creating an embedded space in the browser – the progress bar freezes.
  • APP-457 – fixed problem when shutting down a container and exiting using File or Exit, causing the browser to freeze.
  • APP-11 – added a table that displays all registered spaces and their properties. Clicking the Spaces root node displays this table.
  • APP-127 – added the ability to monitor the Mirror Service using the Cluster View.
  • APP-311 – added the ability to clean statistics in the statistics view.
  • APP-44 – added support for multi-space queries in the Query view. Added a toggle button in order to set one of the following query states: Single-space query, or multi-space query.
  • APP-46 – added support to copy a URL from the Cluster View Members table to the clipboard.

POJO & Spring

  • APP-106 – added template null value support for native types.
  • APP-86 – added Spring 2.0 support.
  • APP-240 – added LeaseContext CLSs for update operations with POJOs.
    An instance of this class is returned as a result of a write call with a POJO, allowing you to distinguish between write and update operations when using the com.j_spaces.core.IJSpace#write(Object entry, Transaction transaction, long lease, long timeout, int modifiers) call or the com.j_spaces.core.IJSpace#write(Object entry, Transaction transaction, long lease) call.
    If a write operation is conducted internally, the LeaseContext holds the Lease object. If an update is conducted internally, the getOriginalObject returns the previous value of the updated object:
    Message o = new Message ("A");
    LeaseContext lease = space.write(o, null, Lease.FOREVER);
    LeaseContext lp = (LeaseContext) lease;
    String uid = lp.getUID();
    Object original  = lp.getOriginalObject();
  • APP-85 – spring 2.0 Declarative Cache support.
  • APP-31 - fixed problem with with converting POJOs to Entrys – super class indexed fields were not handled properly.
  • APP-26 – added a javaSpaceInterceptor synchronous mode example.
  • APP-228 – fixed problem with the UnsupportedOperationException when using the notify operation with the Spring gigaspacestemplate example
  • APP-254 – transaction example config xml has been simplified .

Configuration, API and Tools

  • APP-50 – fixed problem when suing JProperties – XML tags with no value were ignored.
  • APP-109 – modified the gsInstance to use a -classpath by default, and not the -jar. This will load schema files from the <GigaSpaces Root>\config\schemas, and not from the jar.
  • APP-52 – fixed problems when the SpaceFinder or CacheFinder ignore a properties file URL property.
  • APP-57 – fixed problem when StringIndexOutOfBoundsException is thrown while locating the GigaSpaces Root directory.
  • APP-90 – fixed problem when loading the container file (a list of spaces when running multiple spaces as part of the same container) – the container file loaded as a resource from the classpath.
  • APP-165 – fixed problem when a null.xml file is created under the bin directory after creating or destroying a space.
  • APP-25 – Jini Lookup group uses a GigaSpaces version and not a GigaSpaces build number.
  • APP-55 – renamed NO-Serialization to Native-Serialization.
  • APP-79 – added the ability to get a collocated space proxy using IJSpace.getEmbeddedProxy().
  • APP-186 – added the option to display ClusterPolicy configuration at run time using the com.gigaspaces.container.level=CONFIG property.
    This enables the cluster policy runtime configuration report. This report displays the runtime configuration values of the various cluster data structures, such as: replication (synchronized or asynchronized), failover, load-balancing, Mirror Service, etc.
    This report is different from the schema XML files runtime configuration report, that is generated using the com.gs.env.report=true system property (or via the Space Browser), since the schema XML files report display only the actual configuration file structures which might be different from the actual runtime config. Also, in the cluster policy report, you can see all the settings, even those which have not been configured through the cluster configuration file, but are dynamically injected using properties.
  • APP-119 – fixed problem when calling the SpaceConfig.getDCacheProperties() method – default values are returned instead of the values passed in the URL properties file.
  • APP-150 – fixed problem using a secured space – FileNotFoundException is thrown when restarting a container with a secured space. A space was not displayed.
  • APP-404 – fixed problem when starting a space using the gsServer script with groups in the Space URL. The space starts using the specified group, while the Jini LUS runs using different group.
  • APP-423 – fixed problem with the list command: list -listLUS output was missing a service attributes list and a port of the LUS service.
  • APP-122 – added support for passing additional arguments to the gsInstance and the gsServer.
    The gsInstance as well as the gsServer can have 0-3 arguments. If no argument is passed, the default space URL will be used.
    If arguments are passed, an empty " " must be provided to use the default value:
    gsInstance "/./newSpace?schema=persistent" "../../classes" "-DmyOwnSysProp=value -DmyOwnSysProp2=value"
    gsInstance "" "" "-DmyOwnSysProp=value -DmyOwnSysProp2=value"

    In this case, the default space URL and classpath will be used, and new system properties will be appended to the Java command.
    3 arguments can be passed through the command line:

    • Argument 1 – the space URL (surrounded by ""), is set into the SPACE_URL variable.
      If "" is passed, the system will use the default space URL.
    • Argument 2 – classpath is appended to the default classpath.
      This argument should be surrounded by "".
    • Argument 3 – any additional Java command line arguments such as system properties. This argument should be surrounded by "".
  • APP-155 – the default schema now uses NO-Serialization mode (0), like the cache schema. You may need to add your class Entries to the Space JVM classpath as a result of this change, or use Full-Serialization mode (2).
  • APP-281 – fixed problem when using the timeout parameter Space URL – it was ignored.
  • APP-402 – fixed problem with WebSphere Integration – solved the compilation problem with TestServlet.java.
  • APP-220 – added support Command Window Title for gsServer.bat and gsInstance.bat.
  • PACK-68 – added the EXT_JAR variable to all scripts classpath.
  • APP-80 – added API methods to return the Space URL object, which the space or container is initialized with. For more details, see:
    IJSpace.getURL() and IJSpaceContainer.getURL()

CPP & .Net

The CPP and .Net APIs include the additional classes:

Known Issues

  • APP-77 – support multi-NIC in embedded Webster and Reggie.
  • APP-78 – ClassCircularityError is thrown when starting Jini service running inside J2EE + 1.4, or when gsServer + 1.4 + semi-dynamic clustering.
  • APP-204 – <proxy-broadcat-threadpool-min-size> and <proxy-broadcat-threadpool-max-size> are not created in the load-bal-policy section for static cluster configuration.
  • APP-212 – the tag hold-txn-lock must be added or created to both appropriate cluster schema files, and cluster static configuration files.
  • APP-391 – versioning and optimistic locking with POJO are not supported.
  • APP-411 – POJO performance is about 50% of a standard space Entry, both in remote and embedded mode.
  • APP-368 – NPE while invoke example: Advanced\Messaging_Grid\JMS\Simple\bin\run_TopicDurableSubscriber.bat.
  • APP-369 – NPE while invoke example: examples\Advanced\Messaging_Grid\JMS\Simple\Transacted.
  • APP-496 – if the gs.xml has the same attribute defined as part of class and its sub-class, the Entry as part of the space has this attribute displayed twice as part of the Browser Classes view.
  • CORE-414 – space.clean() causes the JVM to hand when using IBM JDK 1.4.1.
  • CORE-474 – when performing update under a transaction, and then performing read or take under the same transaction, the Entry that is returned does not include the update. This will be fixed in future releases.
  • CORE-69 – the LeaseRenewalManager causes the CPU to consume 100%. This might happen when defining the LeaseRenewalManager object as static.
  • CORE-467 – null templates are not supported in broadcast notify. This limitation will be fixed in future versions.
  • DB-16 – SELECT ALL is not supported.
  • DB-20 – the JDBC API does not support the DISTINCT keyword.
  • DB-32 – the JDBC API does not allow updating an object using its UID and version.
  • DB-128 – <GigaSpaces Root>\GenericJDBCProperties\OracleProperties\types.properties should have the following:
    binary=raw(2000)
    longBinary=long raw

    Instead of:

    binary=BLOB
    longBinary=BLOB

    Note that in this case you can have only one complex object as part of your Entry.

  • DB-132 – the SQL keyword IN does not work with a string. This problem will be fixed in later releases.
  • SERVICEGRID-135 – incorrect calculation of free memory while trying to instantiate a 5 GB partition. This error adversely effects the Memory SLA feature, where the Service Grid thinks it does not have enough memory. This problem will be fixed in later releases.
    Current workarounds:
    • Do not use Memory SLA in the DDS.
    • Make sure that the -Xms value is exactly the same as the -Xmx value.
  • When using SQL with readmultiple:
    spaceInstance.readMultiple(sQuery, null, Integer.MAX_VALUE);

    and using metadata Entries, the set of results turns back, and holds no UID information.

CPP & .Net

  • CPPNET-17 - A .Net class that does not contain [GS:VersionID] will throw an exception when update is called to write it to the space. The workaround for 5.1 is to add a version id field and provide the [GS:VersionID] attribute.
  • CPPNET-41 - Use the /platform:x86 switch when compiling and running on 64-bit windows machine platforms. Example:
    csc.exe /target:exe /platform:anycpu  /out:Release/benchmark.exe /recurse:./*.cs  /reference:..\..\lib\netrt.dll;..\..\lib\GsSerializer.dll;..\..\lib\gss50.dll

Space Browser

  • APP-405 – network space table does not support secured spaces.
  • APP-420 – class attribute type for array is not displayed properly.
  • APP-421 – the benchmark graph and table do not display data when performing batch write.
  • APP-29 – in the Create New Space Advanced Mode, the User Permission under the Security tab is grayed out.
  • APP-41 – if during the View of Cluster from GUI, any exception is thrown, then the hourglass cursor remains.
  • APP-280 – the Space Browser does not support space date formats when displaying values in the query view.
  • APP-308 – the JNDI and Jini parameters in the directory services are not saved properly in the configuration file. For example, when both JNDI and Jini are true (checked), and you uncheck only the JNDI checkbox and save, both values are changed to false in the Configuration File.
  • APP-225 – ConcurrentModificationException thrown while executing 2 queries in the same time at the Query Viewer in space browser.
  • APP-252 – space browser com.j_spaces.tools.cluster.view.ServiceManagerImpl$PingEntry Entry causes Lease Expired error messages.
  • APP-269 – the space browser does not show update statistics as a result of replication.
  • APP-207 – the Cluster Wizard does not create mirror service tags, cache-loader tags, the <notify-recovery> tag, and the <ttl> tag in the repl-policy section.
  • APP-272 – when you try to create a Private Space, you got an exception and the action fails.
  • APP-494 – When running the Space Browser with the Windows XP operation system, the Space Browser causes the OS to reboot after 10-15 minutes of idle time. This problem is related to a known issue that JAVA SUN 1.5 has with the NVIDIA screen card. The solution is to upgrade the NVIDIA driver to 11/2005 and beyond.
    Or have the following JVM arguments as part of the Space Browser: sun.java2d.d3d=false.

End of Life Notice

  • As a result of moving to standard Java logging, the Log4j.jar and configurations have been removed from the product.
  • Log Buffer file for persistent space - The Mirror Service should be used instead
  • Complex column - The CacheLoader/Store should be used instead
  • Space Container log - the standard Java logging should be used instead
  • The following Generic Adapter Interfaces and classes as part of the com.j_spaces.sadapter.GenericPA package will not be supported in future versions:
    IBasicPersistence,IMultipleOperations,IQueryblePersistentAdapter,ITransactionalPersistentAdapter,PersistentContext,PersistentEntry,PersistentMatchTemplate. The CacheLoader and CacheStore interfaces should be used instead.

Labels

 
(None)