Looks like it crashed, if I do a /etc/init.d/gluster-server stop on the third node and ps aux | grep gluster after that there is still a lot of processes listed. Should I kill everything ? On Mon, May 23, 2016 at 04:31:28PM +0200, Kevin Lemonnier wrote:> Hi, > > We have in production 3 nodes for a volume storing VM files on 3.7.6, > suddenly the third node isn't seen by the others. For exemple in the heal > infos : > > Status: Brick is not connected > > The network seems to be working fine, the 3 nodes are in cluster and can ping > each other. gluster volume status lists the 3 bricks and I restarted the daemon > on the third node to be sure, still the same result. > The brick is stored on xfs on /mnt/vg1-storage and all the files seems to be there, > it's not read only or anything. > > Where can I check ? > Thanks > > > -- > Kevin Lemonnier > PGP Fingerprint : 89A5 2283 04A0 E6E9 0111> _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users-- Kevin Lemonnier PGP Fingerprint : 89A5 2283 04A0 E6E9 0111 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: Digital signature URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160523/510e7828/attachment.sig>
Have you tried to stop all services of gluster ?? Like - glusterd , glusterfsd On 23/05/16 15:51, Kevin Lemonnier wrote:> Looks like it crashed, if I do a /etc/init.d/gluster-server stop on the third node and > ps aux | grep gluster after that there is still a lot of processes listed. > Should I kill everything ? > > On Mon, May 23, 2016 at 04:31:28PM +0200, Kevin Lemonnier wrote: >> Hi, >> >> We have in production 3 nodes for a volume storing VM files on 3.7.6, >> suddenly the third node isn't seen by the others. For exemple in the heal >> infos : >> >> Status: Brick is not connected >> >> The network seems to be working fine, the 3 nodes are in cluster and can ping >> each other. gluster volume status lists the 3 bricks and I restarted the daemon >> on the third node to be sure, still the same result. >> The brick is stored on xfs on /mnt/vg1-storage and all the files seems to be there, >> it's not read only or anything. >> >> Where can I check ? >> Thanks >> >> >> -- >> Kevin Lemonnier >> PGP Fingerprint : 89A5 2283 04A0 E6E9 0111 > > >> _______________________________________________ >> Gluster-users mailing list >> Gluster-users at gluster.org >> http://www.gluster.org/mailman/listinfo/gluster-users > > > > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users-- Anant Saraswat System Administrator Direct: +91 124 4548387 Tel: +91 124 4548383 Ext- 1018 UK: +44 845 0047 142 Ext- 5020 TBS Website <http://www.techbluesoftware.co.in> Techblue Software Pvt. Ltd The Palms, Plot No 73, Sector 5, IMT Manesar, Gurgaon- 122050 (Hr.) www.techbluesoftware.co.in <http://www.techbluesoftware.co.in> TBS Facebook <https://www.facebook.com/pages/Techblue-Software-Limited/441777369284888> TBS Twitter <https://twitter.com/TechbluSoftware> TBS Google+ <https://plus.google.com/+TechblueSoftwareCoIn> TBS Linked In <https://www.linkedin.com/company/techblue-softwares-pvt-ltd> TBS Branding <http://www.techbluesoftware.co.in> -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160523/34449b3a/attachment.html>
On 05/23/2016 08:21 PM, Kevin Lemonnier wrote:> Looks like it crashed, if I do a /etc/init.d/gluster-server stop on the third node and > ps aux | grep gluster after that there is still a lot of processes listed. > Should I kill everything ?When you create a file from the mount, does it get replicated to the third brick? In any case you can kill all brick processes of a node with `pkill glusterfsd`. Then if you restart glusterd on the same node with `service glusterd restart`, it should bring back all brick processes. You might want to just try restarting glusterd without actually killing the bricks first.> > On Mon, May 23, 2016 at 04:31:28PM +0200, Kevin Lemonnier wrote: >> Hi, >> >> We have in production 3 nodes for a volume storing VM files on 3.7.6, >> suddenly the third node isn't seen by the others. For exemple in the heal >> infos : >> >> Status: Brick is not connected >> >> The network seems to be working fine, the 3 nodes are in cluster and can ping >> each other. gluster volume status lists the 3 bricks and I restarted the daemon >> on the third node to be sure, still the same result. >> The brick is stored on xfs on /mnt/vg1-storage and all the files seems to be there, >> it's not read only or anything. >> >> Where can I check ? >> Thanks >> >> >> -- >> Kevin Lemonnier >> PGP Fingerprint : 89A5 2283 04A0 E6E9 0111 > >> _______________________________________________ >> Gluster-users mailing list >> Gluster-users at gluster.org >> http://www.gluster.org/mailman/listinfo/gluster-users > > > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160523/a2022b20/attachment.html>