Gluster users - Jul 2018 - Gluster FS 3.10.2: degraded rsync performance

Dear All,

after using Gluster FS 3.10.2 for a few months in a production environment, I
noticed that rsync performance degraded terribly.
During last period, rsync copies were very slow: the maximum throughput value
was about 150KB/s.

So I decided to reboot all gluster servers. After that, rsync performance
returned to the normality.
Why this happened? What can I do in order to avoid rsync performance degradation
and a periodical reboot of the entire cluster?

Thank you very much.
Regards,
Mauro

Dear Users,

after a few days, the rsync issue is again here.
Anyone of you experienced the same problem?
It seems that it is not a network issue or an hardware related problem. If I use
scp command, the throughput is good (but I lose all rsync command benefits).

Thank you in advance.
Mauro
> Il giorno 10 lug 2018, alle ore 00:16, Mauro Tridici <mauro.tridici at
cmcc.it> ha scritto:
> 
> Dear All,
> 
> after using Gluster FS 3.10.2 for a few months in a production environment,
I noticed that rsync performance degraded terribly.
> During last period, rsync copies were very slow: the maximum throughput
value was about 150KB/s.
> 
> So I decided to reboot all gluster servers. After that, rsync performance
returned to the normality.
> Why this happened? What can I do in order to avoid rsync performance
degradation and a periodical reboot of the entire cluster?
> 
> Thank you very much.
> Regards,
> Mauro
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20180712/b56efd6c/attachment.html>