You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Nikita Amelchev (Jira)" <ji...@apache.org> on 2026/10/08 19:53:00 UTC
[jira] [Created] (IGNITE-29125) Fix a false critical failure of a coordinator stopped during an exchange
Nikita Amelchev created IGNITE-29125:
----------------------------------------
Summary: Fix a false critical failure of a coordinator stopped during an exchange
Key: IGNITE-29125
URL: https://issues.apache.org/jira/browse/IGNITE-29125
Project: Ignite
Issue Type: Bug
Reporter: Nikita Amelchev
Assignee: Nikita Amelchev
When a node becomes the new coordinator of an unfinished exchange and is stopped at the same moment, the restore-state request to a remote node fails because the local communication SPI has already been stopped. The failure is reported as a critical error although the node is shutting down on purpose, so the failure handler starts a second stop and, with the default handler, may halt the JVM on the stop timeout. This also makes {{TcpDiscoveryClientTopologyGapTest}} [flaky|https://ci2.ignite.apache.org/test/-2566039099717671418?currentProjectId=IgniteTests24Java8].
--
This message was sent by Atlassian Jira
(v8.20.10#820010)