Changelog
This section lists the changes that were made in InsightEdge Platform 12.3.x.
12.3.1
This service pack was released in July 2018.
Improvements
GS-13518 : Sort results for REST operations related to Spaces and Processing Units.GS-13511 : CLI Auto Complete.GS-13510 : Simplified the CLI command “pu decrement” to receive only one parameter: pui_id.GS-13513 : Enhance CLI output to display “N/A” instead of “-1” where relevant.GS-13515 : Upgrade Apache Curator to 4.0.1.GS-13485 : Improve memory allocation by pre-allocating ArrayList instances in the SQL execution engine.GS-13519 : Upgrade RocksDB to 5.11.3.GS-13524 : Display blobstore operation statistics in the Web Management Console.GS-13532 : Enable configuring the MINIMAL_BUFFER_DIFF_TO_ALLOCATE property of the OffHeapMemoryPool.GS-13552 : Share a single instance of Curator Framework client to support large-scale deployments.GS-13551 : Handle Apache Curator leader election events in a single threaded queue.GS-13557 : Configure Spark to include InsightEdge dependencies on startup.GS-13564 : Upgrade Spring to 4.3.17.GS-13562 : Upgrade Apache Zookeeper to 3.4.12 and Netty to 3.10.6.Final.GS-13569 : Log fills up with “unmarshalling failure” messages when a cluster partition member is unreachable.GS-13572 : XAP and InsightEdge demo command should fail if XAP_MANAGER_SERVERS is configured.GS-13571 : Enable building Python notebook in build process for InsightEdge Zeppelin tutorial.GS-13581 : Upgrade Apache Spark to version 2.3.1.GS-13428 : Enhance REST API and CLI to support primary zones.GS-13542 : Display redo log information in the Web Management Console.GS-13579 : Remove dashes from new CLI output to simplify output parsing.GS-13582 : Support public and private IP addresses for Docker deployment.GS-13563 : Decouple JPA integration from OpenSpaces module.
Resolved Issues
GS-13520 : Could not display the off-heap columns in the Web Management Console (in systems with MemoryXtend).GS-13505 : Running the demo command in the CLI using off-heap generates a NoClassDefFoundError.GS-13502 : Provisioning a Space instance fails on “Identified another participant with the same name for Space”.GS-13517 : Index intersection optimization did not work as expected.GS-13512 : Insightedge CLI command fails to start Spark master and Spark worker when XAP_MANAGER_SERVERS is not configured.GS-13509 : Unhealthy Space remains in Stopped state indefinitely after a network failure.GS-13508 : Apache ZooKeeper sometimes fails to create a data directory on startup.GS-13521 : Space instance recovery fails on “Failed while getting participants from zookeeper server”.GS-13486 : IllegalArgumentException: A metric named process_cpu_time-total already exists.GS-13521 : Space instance recovery fails on “Failed while getting participants from zookeeper server”.GS-13486 : IllegalArgumentException: A metric named process_cpu_time-total already exists.GS-13451 : The SQL query “SELECT COUNT (…) GROUP BY” doesn’t work on nested properties in the Web Management Console.GS-13537 : The SQL query “SELECT COUNT (…) GROUP BY” doesn’t work in the Web Management Console when it includes enums.GS-13469 : Remoting service routes incorrectly when the routing key is long, and if the value is higher than the max integer.GS-13544 : The XAP Manager mishandles pending requests when its leadership is relinquished.GS-13548 : Storing the blobStoreVersion not implemented in MemoryXtend off-heap driver.GS-13539 : The Java 8 LocalDateTime is not shown in the Web Management Console object inspector.GS-13496 : The Replication Channels metrics for partition clusters are not being reported.GS-13547 : ServerTypeDesc tree was not cloned correctly.GS-13559 : Display blobstore operation statistics in the GigaSpaces Management Center.GS-13531 : SpaceTypeDescriptorBuilder fails when explicitly indexing the routing key using deprecated values.GS-13541 : Manager class path does not include cert file located in ‘com.gigaspaces.lib.opt.security’ directory.GS-13565 : Space instance becomes backup although it is offered leadership by Apache Zookeeper.GS-13478 : Change operation replicated to mirror can cause NPE in certain failover scenarios.GS-13546 : Thexap space runCLI command fails when XAP_MANAGER_SERVERS is defined.GS-13543 : Queries using the DISTINCT keyword do not work in the Web Management Console.GS-13538 : Web Management Console displayed different values in the Monitoring tab and the Space statistics view for the number of data objects written to a Space.GS-13577 : The ‘pu undeploy’ CLI command returns an error when the resource JAR file is not found in the resources directory.GS-13584 : The ‘xap pu run’ CLI command might leave orphan sub-processes in some scenarios.
12.3.0
The main version was released in March 2018.
Features and Enhancements
GS-13438 : Unified Command Line Interface.GS-13488 : New MemoryXtend storage driver for off-heap memory.GS-13500 : When killing an external process, avoid SIGKILL on systems which do not support it (Windows).GS-13474 : Docker Images for InsightEdge and XAP.GS-13490 : Revise Space index types to provide option for reduced memory footprint.GS-13489 : Reduce @SpaceId(autogenerate=false) memory footprint.GS-13456 : Upgrade Apache Zookeeper to 3.4.10.GS-13400 : Replication redolog compaction for mirror target.GS-13477 : Created v2 of XAP Manager REST API.GS-13491 : Renamed /deployments to /pus in v2 of XAP Manager REST API.GS-13480 : Limit Lookup Locator Discovery interval to 1 minute instead of 1 hour.GS-13482 : Change default of org.jini.rio.monitor.pendingRequestDelay to 1 minute instead of 10 minutes.GS-13470 : Upgrade packaged Jetty server to 9.2.24.GS-13458 : Upgrade Spring Framework to 4.3.13.GS-13457 : Upgrade Spring Security to 4.2.3.GS-13464 : Enhance gs-agent to Support running Web Management Console (–webui).GS-13473 : Enhance CachePolicy configuration to use fluent syntax.GS-13463 : Remove obsolete security algorithm to align with current standards.GS-13102 : Added ability to monitor connection status for local views.GS-12800 : Improved AbstractEventListenerContainer monitoring via public getter methods.GS-13224 : Improve log message to include table name when initial load fails due to wrong column name.GS-13452 : Improve responsiveness of remote statistics gathering in Admin API.GS-13136 : Make instance variables in DefaultHibernateSpaceDataSourceConfigurer protected.GS-13411 : User-defined cache criteria for hot data in MemoryXtend.GS-13436 : Added more GSM state information to application dump.GS-13442 : Reduce the footprint of the Lookup Service template cache.GS-13302 : When the query result set includes only indexes, fetch results from off-heap memory instead of disk.GS-12365 : Asynchronous execution should always return immediately.GS-13417 : Add metrics to Lookup Service to facilitate troubleshooting of performance issues.GS-13389 : Lookup Service (LUS) enhancements to improve stability in environments with large clusters.GS-13433 : Enhance LRMI network filter buffer allocation strategy by adding support for large objects (>10MB).GS-13418 : Skip login page if Web UI is in non-secured mode.
Resolved Issues
GS-12946 : Loop of restarts of PU instance in same container cause memory leak.GS-13484 : Missing partition instance not provisioned after network disconnected and reconnected.GS-13483 : Relocation listener is not called during failure to destroy instance.GS-13454 : Elastic Processing Unit failed to fully deploy with service-limit=1 even when there were empty GSCs running.GS-13300 : Eager Scale Strategy did not evenly redistribute partitions after a cluster node disconnected and reconnected.GS-13465 : Use XapNetworkInfo.getHost() instead of InetAddress.getLocalHost().GS-13462 : XAP Manager doesn’t support com.gigaspaces.lib.opt.security system property for providing an alternative path for custom security JARs.GS-13461 : Add missing xap-admin JAR for default XAP.NET application.GS-13453 : Inefficient use of underlying data structure slows down iterator performance.GS-13284 : In rare scenarios, a transactional Polling container may drop events.GS-13236 : The Admin API blocks itself, becoming unresponsive and exploding memory usage.GS-13440 : Incomplete recovery of processing units after healing of GSM (if there are 3 or more).GS-13281 : Querying a type with enum on a re-deployed space returns incorrect results.GS-13383 : Insightedge shutdown command on Mac systems doesn’t terminate the gs-agent.GS-13416 : Lookup service improvements resulted in loss of client notifications in rare scenarios.GS-13371 : REST calls to Spark failed on machine failover.GS-13317 : No proactive verification of user credentials when logging into secured web-ui.GS-13427 : The REST manager specification contained the wrong request status values.GS-13430 : Inconsistent Request ID type in Manager REST API (string vs. numeric).GS-13204 : The ProcessingUnitInstance.relocationAndWait() may not abort after timeout, causing a permanent blocking state.
See also:
The complete list of changes is also available in JIRA .