Gluster users - Feb 2018 - Thousands of EPOLLERR - disconnecting now

Hello

I have a large cluster in which every node is logging:

I [socket.c:2474:socket_event_handler] 0-transport: EPOLLERR -
disconnecting now

At a rate of of around 4 or 5 per second per node, which is adding up to a
lot of messages. This seems to happen while my cluster is idle.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20180209/59d34a04/attachment.html>

On Thu, Feb 8, 2018 at 2:04 PM, Gino Lisignoli <glisignoli at gmail.com>
wrote:
> Hello
>
> I have a large cluster in which every node is logging:
>
> I [socket.c:2474:socket_event_handler] 0-transport: EPOLLERR -
> disconnecting now
>
> At a rate of of around 4 or 5 per second per node, which is adding up to a
> lot of messages. This seems to happen while my cluster is idle.
>

This log message is normally seen repetitively when there are problems in
the network layer. Can you please verify if there are any problems in the
network layer?

Regards,
Vijay
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20180208/e6fa7074/attachment.html>

Are you running gluster version <= 3.12?

Did you happen to start seeing this flood after rebalance? I'm just trying
to eliminate you're not hitting
https://bugzilla.redhat.com/show_bug.cgi?id=1484885 .

On Fri, Feb 9, 2018 at 4:45 AM, Vijay Bellur <vbellur at redhat.com>
wrote:
>
> On Thu, Feb 8, 2018 at 2:04 PM, Gino Lisignoli <glisignoli at
gmail.com>
> wrote:
>
>> Hello
>>
>> I have a large cluster in which every node is logging:
>>
>> I [socket.c:2474:socket_event_handler] 0-transport: EPOLLERR -
>> disconnecting now
>>
>> At a rate of of around 4 or 5 per second per node, which is adding up
to
>> a lot of messages. This seems to happen while my cluster is idle.
>>
>
>
> This log message is normally seen repetitively when there are problems in
> the network layer. Can you please verify if there are any problems in the
> network layer?
>
> Regards,
> Vijay
>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://lists.gluster.org/mailman/listinfo/gluster-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20180209/b3417e9e/attachment.html>