[
https://issues.apache.org/jira/browse/SLING-11793?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17695683#comment-17695683
]
Joerg Hoh commented on SLING-11793:
-----------------------------------
This approach limits the number of messages (each of unlimited length);
wouldn't it make more sense to limit the total size of the message (e.g. by
number of characters or number of bytes)?
> Limit log messages via JobExecutionContext.log()
> ------------------------------------------------
>
> Key: SLING-11793
> URL: https://issues.apache.org/jira/browse/SLING-11793
> Project: Sling
> Issue Type: Improvement
> Components: Event
> Reporter: Rishabh Kumar
> Priority: Major
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Currently, every log message passed via JobExecutionContext.log() is appended
> to previous messages and then stored in the repository. This can bloat the
> repository and is discouraged as described in JavaDoc:
> {quote}A job consumer can use this method during job processing to add
> additional information about the current state of job processing. As calling
> this method adds a significant overhead it should only be used to log a few
> statements per job processing. If a consumer wants to output detailed
> information about the processing it should persists it by itself and not use
> this method for it. The message and the arguments are passed to the
> MessageFormat class.{quote}
> Some job implementations ignore this advice and still log potentially many
> messages during execution.
> {color:#172b4d}The Sling Job implementation should ignore further log
> messages when a threshold is reached. This may be configurable to make it
> backward compatible{color}.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
