Patch Releases

Potential index corruption in Jump-Skip index, caused by an issue introduced in versions 2024.2.10 / 2025.1.1, has been fixed.

LDAP extension now supports both LDAP and local login authentication

OAuth extension can bypass the Axon Server login page by setting the property axoniq.axonserver.enterprise.oauth2.auto-redirect=true

Improved performance for persistent stream operations

New property axoniq.axonserver.event.stream-threads-per-context (default: 1). Increase value to improve performance with many persistent streams or stream segments

Fixed a race condition in commit index updates during replication

Resolved exception when multiple clients disconnect simultaneously

Fix for starting Axon Server with missing replication logs

Improvements for searching events in the UI:

Rebalance connected clients now only requests a reconnect from one client at a time

Update H2 database version to 2.3.232

Fix testing request handlers from the UI

Add topology change listener to the control channel to send changes in handler subscriptions to client applications

Fix for imbalance in the routing of queries across multiple query handler instances

Fix for too many events sent in install snapshot (since 2025.1.0)

Performance improvements for appending events for a DCB context

Performance improvements reading aggregate events from a follower node

Fix removing a node from the cluster does not properly remove the node (since 2025.0.1)

Fix reconnect button in UI always disabled when access control is enabled

Fix for high CPU usage when another node in the cluster shuts down

Improve performance for appending events through a non-leader node

Update the Spring Boot version to resolve CVE-2035-22235

Potential index corruption in Jump-Skip index, caused by an issue introduced in version 2024.2.10, has been fixed.

LDAP extension now supports both LDAP and local login authentication

Improved performance for persistent stream operations

New property axoniq.axonserver.event.stream-threads-per-context (default: 1). Increase value to improve performance with many persistent streams or stream segments

Fixed a race condition in commit index updates during replication

Resolved exception when multiple clients disconnect simultaneously

Fix for starting Axon Server with missing replication logs

Improvements for searching events in the UI:

Rebalance connected clients now only requests a reconnect from one client at a time

Update H2 database version to 2.3.232

Fix concurrency issue in reading the term of a replication item while the latest segment is being closed

Upgrade Spring-Boot to 3.3.13 to address CVE-2025-48988

Fix for imbalance in the routing of queries across multiple query handler instances

Performance improvements reading aggregate events from a follower node

Replace CLOB/BLOB fields in configuration DB with varchar/varbinary fields to avoid database size increase

Fix timing issue with closing indexes on event transformation

Prevent too many requests from UI to Axon Server

Fix removing a node from the cluster does not properly remove the node (since 2024.2.7)

Fix reconnect button in UI always disabled when access control is enabled

Fix for high CPU usage when another node in the cluster shuts down

Improve performance for appending events through a non-leader node

Update the Spring Boot version to resolve CVE-2035-22235

Fix null pointer exception in install snapshot for _admin replication group

Update axios version used in legacy UI

Fix increasing H2 database size

Fix issue with persistent streams not showing correctly with LDAP extension

Fix for incorrect output in CLI for add node to replication group request

Updated LDAP extension to resolve CVEs in included libraries

Updated Spring Boot patch level to fix CVE-2025-22228

Updated checks for available diskspace. Default warning level for replication log and event store filesystems is now 5 GB. Error level changed to 512 MB, for event store filesystems 512 MB.

Fixed cluster initialization with access control enabled with a specified context

No longer log the license content when license is obtained from Axon Console

Update Spring Boot version to 3.3.9 to resolve vulnerabilities in Netty-handler and JSON-smart libraries

Update gRPC version to 1.71.0

Fixes in Create context dialog for creating context and replication group and creating context in replication groups with multiple node roles

Fix in persistent streams to avoid loop when resetting the stream position immediately after merging segments

Updated access control for purging the event store in dev mode. The action now requires the CONTEXT_ADMIN role. Renamed the parameter targetContext to context on the /v1/public/purge-events endpoint.

Fixed a bug that caused problems when creating a new context in the presence of a user with permissions on *.

Adjusted configdb parameters, so that the database is closed only once. This resolves exceptions being logged at shutdown.

OAuth extension now takes context path into account when generating URLs.

OAuth extension supports setting the desired algorithm for id token signature using the property idTokenSignatureAlgorithm.

Fixed auto-cluster issue when internal-domain property is set in axonserver properties.

Moved option to export traces to Jaeger to an extension. To migrate, move the jar files in the extensions/axon-server-extensions-jaeger-2024.2.2.zip to the exts directory in the Axon Server installation directory.

update packaged versions of LDAP and OAuth extensions

Redistribute clients across Axon Server nodes when a node is restarted with access control enabled

UI fixes for non-admin users:

Axon Server node should return information about the clients connected to that node only

Backing up the RAFT log entries should take into account replication group rather than context

Record handling duration for initial query of a subscription query

Persistent streams improvements:

Redistribute clients across Axon Server nodes when a node is restarted

Event processor operations fail when the processing group contains a forward slash

Potential replication issue when trying to apply events for already closed contexts during shutdown of Axon Server

Increased maximum length for the username to 255 characters

Update the event store size when a new index file is created

Visual improvements in search table: headers not visible by default & action not visible by default

Improved logging in the event store

Stop replication applying process when the replication group is stopped

Global Index pre-load for configured contexts

Revert optimization in replication from version 2023.2.4, as it could lead to a node entering fatal state

Stop Axon Server from redirecting a client to a node that is in fatal state

Reduce communication between the leader and follower and logging when a node is starting up

Search page improvements

Set correct permissions for persistent stream API calls

Add validation of newly created index files

Allow non-pristine clusters to connect to Console

Fix the event store size in the context page

Fix for a problem starting up Axon Server with plugins configured

Removed race condition causing a possible delay in receiving the first event on a newly registered event handler

Improve the diagnostics package to contain full log information when "logging.config" property is set -

Add an option to reduce the number of global index segments Axon Server checks when the first event for a new aggregate is stored. This can be configured globally with the property "axoniq.axonserver.event.global-index-segments-check" or on a context level with the property "event.global-index-segments-check". The value is the number of global index segments to check, with a minimal value of 2.

Fix for Control DB migration in case of plugin configuration properties with long values

Updating a license through Axon Console now takes effect immediately

Improved distribution of queries to different instances of the query handlers

Updating a license through Axon Console now takes effect immediately

Reduced memory usage for internal communication

Reduced the number of threads used with a large number of contexts

UI improvements

axoniq.axonserver.accesscontrol.same-origin=false (true sets the X-Frame-Options header to SAMEORIGIN)

axoniq.axonserver.event-store-background-thread-count=8

axoniq.axonserver.event-store-processors-thread-count=8

Fix the increasing number of threads on the running Axon Server nodes when one node in the cluster is down.

Small fixes in the replication process:

remove delay in starting to synchronize with a node that is far behind

improve the performance for a follower catching up with the leader

prevent situations where a follower attempts to apply replication log entries that were already included in a snapshot

Fix for authentication issue when multiple applications have the same token

UI, copy token to clipboard fails when not running on a trusted URL

UI, improved validations for applications, replication groups and contexts operations

Improved handling for missing connection to Axon Console

Support for Google Marketplace licenses

Axon Server now performs a clean shutdown when it was started with an incorrect node name or internal hostname/port

Redistribute clients across Axon Server nodes when a node is restarted with access control enabled

Redistribute clients across Axon Server nodes when a node is restarted

Event processor operations fail when the processing group contains a forward slash

Potential replication issue when trying to apply events for already closed contexts during shutdown of Axon Server

Revert optimization in replication from version 2023.2.4, as it could lead to a node entering fatal state

Stop Axon Server from redirecting a client to a node that is in fatal state

Reduce communication between the leader and follower and logging when a node is starting up

Improved distribution of queries to different instances of the query handlers

Prevent stale threads when an Axon Server node closes the connection to another node

Clean up metrics from disconnected clients

prevent WARN log messages when a query completed message was received from an unexpected client

gRPC version updated to 1.65.1

Add an option to reduce the number of global index segments Axon Server checks when the first event for a new aggregate is stored. This can be configured globally with the property "axoniq.axonserver.event.global-index-segments-check" or on a context level with the property "event.global-index-segments-check". The value is the number of global index segments to check, with a minimal value of 2.

Reduced memory usage for internal communication

Fix the increasing number of threads on the running Axon Server nodes when one node in the cluster is down.

Small fixes in the replication process:

remove delay in starting to synchronize with a node that is far behind

improve the performance for a follower catching up with the leader

prevent situations where a follower attempts to apply replication log entries that were already included in a snapshot

Fix for authentication issue when multiple applications have the same token

Increasing number of threads on the running Axon Server nodes when one node in the cluster is down.

Fix for an error handling subscription query responses during the upgrade from a version before 2023.2.0 to 2023.2.0 or 2023.2.1.

Improved readiness probe to return 200 (OK) once the communication services are ready and the replication groups are completely initialized. The endpoint for the new readiness probe is /actuator/health/readiness.