[
https://issues.apache.org/jira/browse/CAMEL-26794?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Work on CAMEL-26794 started by Claus Ibsen.
-------------------------------------------
> camel-cli-connector - an Error in its scheduled task stops the status updates
> and the CLI actions for good
> ----------------------------------------------------------------------------------------------------------
>
> Key: CAMEL-26794
> URL: https://issues.apache.org/jira/browse/CAMEL-26794
> Project: Camel
> Issue Type: Bug
> Components: camel-cli-connector
> Reporter: Claus Ibsen
> Assignee: Claus Ibsen
> Priority: Minor
> Fix For: 4.24.0
>
>
> The file transport runs its work (actions, the status file, traces) from one
> scheduleWithFixedDelay task, and the WebSocket transport its heartbeat and
> debug tasks the same way. They catch Exception, but not Error. A
> ScheduledExecutorService cancels a task for good once it throws, so a single
> OutOfMemoryError on the connector thread (a short memory spike while the
> status is collected) ends it, even when the JVM recovers:
> * the status file is never written again, so camel get, camel ps and the TUI
> show the numbers of that moment from then on
> * action files are no longer read: camel stop, reset-stats, a send, etc. do
> nothing
> * the integration no longer notices that its lock file is deleted
> The integration itself keeps routing.
> The tasks should survive any Throwable: they are the only way the Camel CLI
> can reach and stop the integration. The error is logged (inside its own try,
> as logging can fail too under memory pressure) and the task runs again on its
> next delay.
> _Claude Code on behalf of davsclaus_
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
