RabbitMQ 3.6.12 with memory leak?

342 views
Skip to first unread message

Peter Ko

unread,
Nov 21, 2017, 1:00:36 AM11/21/17
to rabbitmq-users
We’re running a RabbitMQ cluster ("rabbitmq-server-3.6.12-1.el7" and "erlang 19.3”) with 5 nodes which runs on "CentOS Linux release 7.3".

We had an issue with RabbitMQ about two months ago. So we've upgraded our version of RabbitMQ from 3.6.10 to 3.6.12.
Here is the link to my previous post. 

It has been running okay since, but now only one of the nodes in the cluster is showing high memory consumption.
node1 2.4G 
node2 623MB 
node3 634MB
node4 512MB
node5 239MB

Current rabbitmq cluster stat.
ready 390,485 messages 
publish 13/s
deliver 11/s

Node1's memory details.
Queues 219MB queues
Binaries 41MB
System - 27MB code
             - 1.1MB atoms
     - 1.4GB other

I think RabbitMQ is somehow having an issue with returning (fully used) memory. Maybe some sort of memory leak?
The attached image is showing the ground of my thought.


Anyone has any clue why only one node in the cluster is showing such irregular memory consumption?  
Any help would be appreciated.
스크린샷 2017-11-21 오후 2.02.25.png

Michael Klishin

unread,
Nov 21, 2017, 3:17:57 AM11/21/17
to rabbitm...@googlegroups.com

Michael Klishin

unread,
Nov 21, 2017, 3:19:33 AM11/21/17
to rabbitm...@googlegroups.com
If it's just one node then see if it may be handling most of queue masters [1] or connections [2].

1. Queue Master Location in http://www.rabbitmq.com/ha.html

To post to this group, send email to rabbitm...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
--
MK

Staff Software Engineer, Pivotal/RabbitMQ
Reply all
Reply to author
Forward
0 new messages