[logback-user] Deadlocks with logback /RollingFileAppender

[Manoj Khangaonkar]

Hi,


This has posted by many other in the past as is evident from google search.


LOGBACK-891 is a one open JIRA that points to this issue.


Under heavy load, a multi threaded Java service deadlocks and hangs. And all there threads

are blocked in LOG.info calls. One log call seems to hold a lock. Every other log call is waiting on that lock.


I have posted a stack snippet from my thread dump on LOGBACK-891


Is anybody who encountered this issue still on the mailing list ? If so, are there any suggestions or work arounds.


Are there any best practices for using logback Loggers in a multithreaded program.


Committers,


Most prior jira or discussions - seem to end with "This cannot happen. We need a test case ". Is that still the status of

this issue ?


The issue is real and I could try to produce a test case that mirrors my case. But just so as to not reinvent the wheel, if any learning

or fixes came out of prior reports of deadlocks, can you please share them.


regards


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qos.ch/pipermail/logback-user/attachments/20180911/0f6ed794/attachment.html>