You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@artemis.apache.org by cl...@apache.org on 2026/09/10 20:12:17 UTC
(artemis) branch main updated: NO-JIRA Fixing release notes on 2.57.0
This is an automated email from the ASF dual-hosted git repository.
asf-gitbox-commits pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/artemis.git
The following commit(s) were added to refs/heads/main by this push:
new b3b4caea0e NO-JIRA Fixing release notes on 2.57.0
b3b4caea0e is described below
commit b3b4caea0e37a48c5e4ee339595980f9aeb6283b
Author: Clebert Suconic <cl...@apache.org>
AuthorDate: Thu Sep 10 14:23:32 2026 -0400
NO-JIRA Fixing release notes on 2.57.0
(cherry picked from commit 77528a997dc90d3549773f220934d13c79a02d03)
(cherry picked from commit 26bd36570d4cbad7dbfbb9cf24a5936a0661e142)
(cherry picked from commit a00dc08bf10555345e7a130745a93c89ef852891)
---
docs/user-manual/versions.adoc | 38 ++++++++++++++++++++++++--------------
1 file changed, 24 insertions(+), 14 deletions(-)
diff --git a/docs/user-manual/versions.adoc b/docs/user-manual/versions.adoc
index c55576153f..92a21330a2 100644
--- a/docs/user-manual/versions.adoc
+++ b/docs/user-manual/versions.adoc
@@ -15,20 +15,13 @@ NOTE: Follow the general upgrade procedure outlined in the xref:upgrading.adoc#u
:sectnums!:
-== Version 2.56.0
-
-https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315920&version=12357208[Full release notes]
+== Version 2.57.0
=== Highlights
-* The xref:lock-coordination.adoc#lock-coordination-ha-with-mirroring[lock coordinators] now support a new `auto-start` attribute to control whether a coordinator begins polling for the distributed lock on broker activation, plus new `startLockCoordinator` / `stopLockCoordinator` management operations and matching `lock start` / `lock stop` CLI commands.
-* A lock coordinator that cannot reach its distributed lock store (e.g. ZooKeeper is unreachable) no longer hangs broker shutdown indefinitely.
-* The AMQP protocol module has been updated to proton-j 0.35.0 with a new `maxTransfersPerDelivery` configuration option.
-* The journal file lock health-check behavior is now configurable via the new `journal-lock-monitor-timeout` and `journal-lock-monitor-max-retries` parameters in `broker.xml`.
-See xref:persistence.adoc#configuring-the-message-journal[configuring the message journal] for details.
-* MQTT QoS 1 and QoS 2 message delivery has been redesigned for resiliency across broker restarts and client reconnects.
+* This version contains important bug fixes and updates.
-=== Upgrading from 2.55.0
+=== Upgrading from 2.56.0
* The Core protocol connect handshake packet was extended in a backward-compatible way on the wire (same packet type identifiers).
When security is enabled, Core clients that authenticate at connection time should pass credentials via `ServerLocator.setConnectionCredentials(user, password)` or `ServerLocator.createSessionFactory(user, password)`.
@@ -42,6 +35,27 @@ An old broker connecting to a broker that has already been upgraded does not sen
The upgraded broker therefore treats the connection as unauthenticated and responds with only the placeholder topology entry, not the real cluster view.
Until that old broker is upgraded as well, it may be unable to discover or connect properly to the new brokers (cluster connections, bridges, replication, backup registration, and similar Core links can fail or stall).
Upgrade every broker in the cluster, or temporarily disable `coreConnectionSecurityEnabled` on acceptors during a transitional window, before relying on cross-version cluster connectivity.
+* Server discovery is disabled by default, see the xref:clusters.adoc#server-discovery[Server Discovery] section in the user manual for security requirements and enablement.
+* Selector filters now limit `%` wildcards to 5 by default.
+Configure via system property `org.apache.activemq.artemis.selector.maxWildcards` or environment variable `ARTEMIS_SELECTOR_MAX_WILDCARDS`.
+The broker will fail to start if existing durable subscriptions exceed this limit, so increase the value before upgrading if needed.
+
+== Version 2.56.0
+
+https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315920&version=12357208[Full release notes]
+
+=== Highlights
+
+* The xref:lock-coordination.adoc#lock-coordination-ha-with-mirroring[lock coordinators] now support a new `auto-start` attribute to control whether a coordinator begins polling for the distributed lock on broker activation, plus new `startLockCoordinator` / `stopLockCoordinator` management operations and matching `lock start` / `lock stop` CLI commands.
+* A lock coordinator that cannot reach its distributed lock store (e.g. ZooKeeper is unreachable) no longer hangs broker shutdown indefinitely.
+* The AMQP protocol module has been updated to proton-j 0.35.0 with a new `maxTransfersPerDelivery` configuration option.
+* The journal file lock health-check behavior is now configurable via the new `journal-lock-monitor-timeout` and `journal-lock-monitor-max-retries` parameters in `broker.xml`.
+See xref:persistence.adoc#configuring-the-message-journal[configuring the message journal] for details.
+* MQTT QoS 1 and QoS 2 message delivery has been redesigned for resiliency across broker restarts and client reconnects.
+
+=== Upgrading from 2.55.0
+
+
* Due to https://issues.apache.org/jira/browse/ARTEMIS-6128[ARTEMIS-6128] the `Alder32` and `fileAlder32` properties have been deprecated in favor of `Adler32` and `fileAdler32` respectively.
* Due to https://issues.apache.org/jira/browse/ARTEMIS-6172[ARTEMIS-6172] the HTTP-specific transport parameters deprecated by https://issues.apache.org/jira/browse/ARTEMIS-5819[ARTEMIS-5819] (`httpClientIdleTime`, `httpClientIdleScanPeriod`, `httpResponseTime`, `httpServerScanPeriod`) are restored.
These are required so HTTP-tunneled Core connections can create response slots for server-initiated packets (for example cluster topology updates) without starving replies to client requests.
@@ -57,10 +71,6 @@ Due to the additional persistent state there will be a performance impact.
Of course, this is a traditional trade-off for increased reliability.
The exact impact will depend on your specific use-case.
Generally speaking, the recommendation is to use the lowest QoS possible to avoid unnecessary overhead.
-* Server discovery is disabled by default, see the xref:clusters.adoc#server-discovery[Server Discovery] section in the user manual for security requirements and enablement.
-* Selector filters now limit `%` wildcards to 5 by default.
-Configure via system property `org.apache.activemq.artemis.selector.maxWildcards` or environment variable `ARTEMIS_SELECTOR_MAX_WILDCARDS`.
-The broker will fail to start if existing durable subscriptions exceed this limit, so increase the value before upgrading if needed.
== Version 2.55.0
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@artemis.apache.org
For additional commands, e-mail: commits-help@artemis.apache.org