Hazelcast Clustering Plugin Changelog
2.6.1 -- November 9, 2022
- [Issue #86] - Reduce overhead of loading class definitions
2.6.0 -- December 6, 2021
NOTE: This version of the plugin requires Openfire 4.7.0 or higher
- [Issue #50] - Move non-Hazelcast code to Openfire-core
- [Issue #65] - Postpone member joined event until joining node is ready to receive new data
- [Issue #69] - When recovering split-brain, let 'leave' play out before 'joining' again
- [Issue #71] - Send member joined notification across the cluster on clustering start
- [Issue #74] - Warn against usage of plugin-provided classes in Hazelcast
- [Issue #76] - Finish leftCluster handling before invoking joinCluster handlers
2.5.1 -- September 29, 2021.
NOTE: This version and lower of the plugin will not work on Openfire 4.7.0 and higher
2.5.0 -- January 2, 2020
NOTE: This version of the plugin requires Openfire 4.5.0 or higher
- [OF-1868] - Session message carbons, requested blocklist not clustered
- [Issue #35] - Remove support for Openfire < version 4.5
- [Issue #37] - Support for SystemProperty
- [Issue #39] - Expose the Cache size as a long
- [Issue #40] - Reduce level of debug logging for cluster tasks
- [Issue #42] - ClusterTask exceptions are silently lost
- [Issue #44] - Update to Hazelcast 3.12.5
2.4.2 -- April 26, 2019
- [Issue #23] - Split-brain recovery does not follow documented process
- [Issue #25] - Update to Hazelcast 3.12
2.4.1 -- March 22, 2019
- [Issue #5] - Dynamically created ClusteredCache entries not expiring
- [Issue #13] - Less dramatic log messages when running a one-node cluster
- [Issue #16] - Expose host name of cluster members on the admin UI
- [Issue #19] - Stop changing the XMPPServer NodeID
2.4.0 -- January 25, 2019
- Minimum Java requirement: 1.8
- [OF-827] - Add support for XEP-0191: Blocking Command
- [OF-1517] - Don't require i18n source files for all plugins to be encoded.
- [OF-1641] - Ensure all JSP pages have the correct contentType.
- [Issue #1] - ClusteredCacheFactory.isSeniorClusterMember() incorrectly returns false
- [Issue #3] - Node-specific cluster info page not displaying
- [Issue #8] - Update to Hazelcast 3.11.1
2.3.0 -- March 15, 2018
- [HZ-5] - The markedAsSeniorClusterMember event is not fired when a node becomes senior
- [HZ-11] - Disabling clustering never completes
- [HZ-2] - Update to Hazelcast 3.9.2
- [HZ-8] - Preserve Hazelcast network settings between plugin upgrades
2.2.4 -- October 26, 2017
- [OF-1416] - Dynamically created ClusteredCache entries not expiring
2.2.3 -- September 15, 2017
- Use an event rather than an arbitrary delay before starting clustering
2.2.2 -- August 3, 2017
- Ensure that Hazelcast backed Cache objects have the correct settings
2.2.1 -- November 4, 2016
- [OF-1210] - correct time-to-live-seconds and MaxLifetime settings for hazelcast
2.2.0 -- October 12, 2015
- [OF-953] - Updated JSP libraries.
- Requires Openfire 4.0.0
2.1.2 -- September 16, 2015
- OF-943: Fix intermittent timeout issues for synchronous remote task execution
2.1.1 -- August 11, 2015
- Updated default Hazelcast config file to match 3.5 schema
2.1.0 -- August 7, 2015
- Updated Hazelcast to latest release (3.5.1).
- Avoid reverse DNS lookups for cluster node members (patch submitted by caojianlin)
2.0.0 -- February 6, 2015
- Updated Hazelcast to latest release (3.4).
- Repackaged to use Openfire's org.jivesoftware.openfire namespace prefix (rather than com.jivesoftware.openfire).
1.3.4 -- January 10, 2015
- Updated clustering SPI to better match design and documentation.
- Fixed multiple session-related NPEs that manifest under high load/latency in the cluster.
1.3.3 -- December 16, 2014
1.3.1 -- October 15, 2014
Hazelcast plugin update for Openfire 3.10.x:
- Updated Hazelcast to release 3.3.1 (what's new).
- Enhanced plugin to implement changes in the 3.10.x Openfire Session API (OF-843).
1.2.2 -- May 6, 2014
- Fix node cleanup logic when a node leaves the cluster (OF-794)
1.2.1 -- April 10, 2014
- Updated Hazelcast to release 3.1.7 (what's new).
- Changed configuration file and README to adjust for Hazelcast schema changes.
1.2.0 -- February 10, 2014
- Fix cluster initialization logic (OF-699)
- Updated Hazelcast to release 3.1.5 (what's new).
1.1.0 -- Sep 13, 2013
1.0.6 -- May 1, 2013
- Added support for cluster time (OF-666)
hazelcast-cloud.jar to support AWS deployments (more info).
- Updated Hazelcast to release 2.5.1 (bug fixes).
1.0.5 -- March 26, 2013
JMX Support (OF-655)
- Optionally enable JMX instrumentation for Hazelcast if JMX is enabled for Openfire (via the admin console).
- Updated minimum server version to Openfire 3.8.2 (release pending)
1.0.4 -- February 6, 2013
- OF-607: Optimize JID serialization in cluster mode
- Updated minimum server version to Openfire 3.8.0 (release pending)
1.0.3 -- January 15, 2013
Minor improvements for plugin installation and upgrade:
- Modify timing for cluster startup/shutdown operations to improve plugin installation and upgrade.
- Defer initialization of cluster event dispatcher thread until cluster is actually enabled.
readme.html to describe upgrade procedures, DNS round-robin configuration, etc.
- Upgraded Hazelcast to version 2.5.
1.0.2 -- January 9, 2013
This release addresses a number of issues and other feedback received via the
posted in the Openfire community forum:
- Fixed cluster serialization issues for null collection objects.
- Improved error handling for remote sessions from invalid/offline cluster members.
- Removed extraneous IllegalStateException thrown during certain cluster tasks.
- Avoid encoding issues across cluster members by forcing UTF-8 for nodeID strings.
- Added nodeID to the system clustering overview page in the admin console.
- Updated cache configuration to prevent eviction for critical application components.
1.0.1 -- December 14, 2012
- Upgraded Hazelcast to version 2.4.1.
1.0.0 -- September 22, 2012