[
https://issues.apache.org/jira/browse/CAMEL-26339?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Claus Ibsen resolved CAMEL-26339.
---------------------------------
Resolution: Fixed
Fixed by https://github.com/apache/camel/pull/28556 (merged to main as
b18e4520df39).
> camel-plc4x - the triggered consumer logs a WARN with a stack trace on every
> failed poll
> ----------------------------------------------------------------------------------------
>
> Key: CAMEL-26339
> URL: https://issues.apache.org/jira/browse/CAMEL-26339
> Project: Camel
> Issue Type: Improvement
> Reporter: Claus Ibsen
> Assignee: Serdar Gökay
> Priority: Minor
> Fix For: 4.24.0
>
>
> When the PLC cannot be reached, the triggered consumer of camel-plc4x (a
> {{trigger}} on the endpoint, which polls every {{period}} ms) fails every
> poll. Each failed poll goes to the consumer's exception handler, which logs a
> WARN with the full stack trace ({{Failed polling endpoint: ...}}). With
> {{period=100}} that is ten stack traces a second for as long as the PLC is
> down.
> With the default {{autoReconnect=false}} each poll also logs {{Connection
> lost and auto-reconnect is turned off, shutting down Plc4XEndpoint}} from
> {{Plc4XEndpoint.reconnectIfNeeded()}}, and stops the endpoint again, while
> the route keeps polling.
> Users cannot calm this down with the scheduled poll options
> ({{backoffErrorThreshold}}, {{backoffMultiplier}}), because {{Plc4XEndpoint}}
> is a {{DefaultEndpoint}} and does not offer them.
> This was seen in the review of CAMEL-25617
> (https://github.com/apache/camel/pull/28387), on a real S7-1200 and on
> protocol simulators, and was left for a follow-up.
> Possible directions:
> * Log the first failed poll of a run of failures at WARN with the stack
> trace, the following ones without the stack trace or at DEBUG, and log once
> at INFO when a poll succeeds again.
> * With {{autoReconnect=false}}, log the shutdown once, and stop polling (or
> fail the route) rather than repeating it on every poll.
> * Optionally offer the backoff options of a scheduled poll consumer on the
> endpoint.
> The other consumers and the producer share {{reconnectIfNeeded()}}, so they
> are worth a look too.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
