[ejabberd] Ejabberd shared roster performance

Evgeny Khramtsov xramtsov at gmail.com
Fri Jan 5 11:07:21 MSK 2018


Tue, 2 Jan 2018 18:04:10 -0800
Gregory Makarov <gmakarov at gmail.com> wrote:

> Hi,
> 
> I've done some load testing of Ejabberd and I see some strange
> results.
> 
> Ejabberd 17.08 on two node cluster with 4CPU cores (2.60GHz) and 8GB
> of memory on each node.
> 5000 online users in 250 shared rosters (20 users per shared roster).
> "roster" table is empty. Users change their presence approximately
> each minute (Away -> DND -> Away).
> CPU usage on both nodes - approximately 350%.
> 
> Why CPU usage is so high?

Because the implementation is suboptimal. This is a well-known issue.


More information about the ejabberd mailing list