PR for upstream: CEPHFS related

[Ruben Gaspar]

Dear All,

Just to request if it would be possible to merge in upstream this PR/commit [1] that I was discussing with Matthias N. during last Opencast summit, in Zaragoza. 

We are kind of affected by some issue related with CEPHFS (may be [2]), which may affect other people running Opencast on CEPHFS.

Our symptoms are a huge CPU and bandwidth consumption on our admin node (just acting as admin/presentation node), from which most of it is IO wait. It happens pretty randomly, usually a clean-up of jobs and a reboot fix it for a little while.

Thank you,

Ruben

--

[1] https://zivgitlab.uni-muenster.de/educast-nrw/forks/opencast/-/commit/7ed7ad46bfd56a287e4509215a5ad707f5d5913f

[2] https://www.mail-archive.com/ceph-...@lists.ceph.com/msg53368.html

[Katrin Ihler]

Hi Ruben,

that fix is already merged, see https://github.com/opencast/opencast/pull/5682 Should be part of the next release for 14 and onward. :)

Cheers,

Katrin

--
To unsubscribe from this group and stop receiving emails from it, send an email to users+un...@opencast.org.

-- 
ELAN e.V.
Karlstr. 23
D-26123 Oldeburg

elan-ev.de

[Ruben Gaspar]

Thank you Katrin, that's great moving to 14 version in the next two
weeks. Hopefully it fixes our issue.
Cheers,
Ruben