Hello all,
we need to know how to do the proper sizing of memory for rsyslog server.
We did the sizing of ActionQueueHighWaterMark according to the memory
available on the server and median message size.
Our assumption is the memory consumption should not grow "much higher" than
ActionQueueHighWaterMark*message_size.

We do run rsyslogd 8.15 version.
Just experienced issue with memory shortage and rsyslog process crash.

Some numbers:
approx. 9000 messages/minute
median message size 840 Bytes (max. message 889Bytes)
standard rsyslog memory ~630MB
rsyslog *memory before crash 3.2GB*
number of messages in memory queues (activated due to TCP forward outage)

*  1st queue 1045144  2nd queue 742223*
rsyslog threads 16
rsyslog threads on crash 20
stack size 8MB (default not lowered)

>From these numbers it seems the median message size should be around 1.3KB
(if only messages allocating the memory).

Searched for some memory consumption bugs and found this commit
https://github.com/rsyslog/rsyslog/commit/db2ceea663bafe720588d8219973f9d979b990b4
Not sure whether this bug could have such impact.

What could cause the higher(-than-expected) memory usage? Are there any
obvious errors in our assumptions? Something else we have to consider for
proper sizing?

-- 
Peter
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of 
sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE 
THAT.

Reply via email to