If you were on rc0 and upgraded to rc1, then you are hitting BZ 1684029 I believe. Can you please upgrade all the nodes to rc1, bump up the op-version to 60000 (if not already done) and then restart glusterd services to see if the peer rejection goes away?
On Thu, Mar 14, 2019 at 7:51 AM Guillaume Pavese < [email protected]> wrote: > putting [email protected] in the loop > > Guillaume Pavese > Ingénieur Système et Réseau > Interactiv-Group > > > On Thu, Mar 14, 2019 at 11:04 AM Guillaume Pavese < > [email protected]> wrote: > >> Hi, I am testing gluster6-rc1 on a replica 3 oVirt cluster (engine full >> replica 3 and 2 other volume replica + arbiter). They were on Gluster6-rc0. >> I upgraded one host that was having the "0-epoll: Failed to dispatch >> handler" bug for one of its volumes, but now all three volumes are down! >> "gluster peer status" now shows its 2 other peers as connected but >> rejected. Should I upgrade the other nodes? They are still on Gluster6-rc0 >> >> >> Guillaume Pavese >> Ingénieur Système et Réseau >> Interactiv-Group >> >> >> On Wed, Mar 13, 2019 at 6:38 PM Niels de Vos <[email protected]> wrote: >> >>> On Wed, Mar 13, 2019 at 02:24:44AM +0000, [email protected] >>> wrote: >>> > SRC: >>> https://build.gluster.org/job/release-new/81/artifact/glusterfs-6.0rc1.tar.gz >>> > HASH: >>> https://build.gluster.org/job/release-new/81/artifact/glusterfs-6.0rc1.sha512sum >>> >>> Packages from the CentOS Storage SIG will become available shortly in >>> the testing repository. Please use these packages to enable the repo and >>> install the glusterfs components in a 2nd step. >>> >>> el7: >>> https://cbs.centos.org/kojifiles/work/tasks/3263/723263/centos-release-gluster6-0.9-1.el7.centos.noarch.rpm >>> el6 >>> <https://cbs.centos.org/kojifiles/work/tasks/3263/723263/centos-release-gluster6-0.9-1.el7.centos.noarch.rpmel6>: >>> >>> https://cbs.centos.org/kojifiles/work/tasks/3265/723265/centos-release-gluster6-0.9-1.el6.centos.noarch.rpm >>> >>> Once installed, the testing repo is enabled. Everything should be >>> available. >>> >>> It is highly appreciated to let me know some results of the testing! >>> >>> Thanks, >>> Niels >>> _______________________________________________ >>> packaging mailing list >>> [email protected] >>> https://lists.gluster.org/mailman/listinfo/packaging >>> >> _______________________________________________ > maintainers mailing list > [email protected] > https://lists.gluster.org/mailman/listinfo/maintainers >
_______________________________________________ Gluster-users mailing list [email protected] https://lists.gluster.org/mailman/listinfo/gluster-users
