hazelcast/hazelcast v5.2.4

on GitHub

This document lists the enhancements and fixed issues for the Hazelcast Platform 5.2.4 release. The numbers in the square brackets refer to the issues and pull requests in Hazelcast's GitHub repository (github.com/hazelcast/hazelcast).

Enhancements

• Updated the version of jackson-core dependency to 2.15.2. [#24730]
• Hazelcast was sending requests to Kubernetes API when deploying an application with embedded Hazelcast and service-dns (DNS lookup mode) specified to a Kubernetes cluster. This was causing the requests to be unsuccessful and the application not to start. This mechanism has been improved by creating Kubernetes client only for the DNS lookup mode. [#24045]

Fixes

• Fixed an issue where some of the members in a Hazelcast cluster deployed on Kubernetes (as a statefulset) shut down with en exception in a delayed manner. [#24709]
• Fixed an issue where Jet job snapshots could be prematurely deleted after a restart of a cluster, having lossless restart enabled. [#24576]
• Fixed an issue where the SELECT COUNT(DISTINCT COLUMN) query for maps was producing incorrect results. [#24490]
• Fixed various issues in [Health Monitor] including incorrect metric names. [#24634]
• Fixed an issue where the REST calls were failing for Hazelcast clusters with TLS v1.3 configured, and deployed on Kubernetes. [#24624]
• Fixed an issue where SQL statements were failing when a class (to determine the fields of a key/value pair) no longer exists but the mapping is still valid. [#24043]
• Fixed an issue where lastAccessTime and expirationTime was not updated when an entry is accessed via executeOnEntries. [#24012]