Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
On 03/11/16 11:08 -0500, Ken Gaillot wrote: > ClusterLabs is happy to announce the first release candidate for > Pacemaker version 1.1.16. Source code is available at: > > https://github.com/ClusterLabs/pacemaker/releases/tag/Pacemaker-1.1.16-rc1 > > [...] As usual, there are COPR builds (using upstream spec file without any final touch that is usually done downstream) for easy consumption in some environments: https://copr.fedorainfracloud.org/coprs/jpokorny/pacemaker/build/473980/ I also have something to share regarding recently announced security fix in pacemaker if you are interested in Fedora: fixed packages should be available from updates-testing repo in Fedora 23 and Fedora 25, and regular updates repo in Fedora 24 at the moment. -- Jan (Poki) pgpeRMbXtWvm5.pgp Description: PGP signature ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
On 11/03/2016 08:18 PM, Kristoffer Grönlund wrote: > Klaus Wenninger writes: > >> No, nothing specific to Red Hat. >> sbd consists of a couple of processes talking to each other. >> One is the 'inquisitor' (opening the watchdog-device and >> kicking it if everything seems OK) and there are multiple >> 'watchers' for the respective instances like pacemaker >> (well the name says it), cluster (either exchanging cpg-messages >> with corosync or checking for the existence of pacemaker_remoted) >> and of course the watcher for the block-device (Red Hat >> specific in that way that it is disabled in the build ;-) ). >> > To clarify this part, we (SUSE) haven't yet integrated the fork of SBD > by Andrew which adds watchdog-only fencing, we're still relying on > having a storage device. > > I'm not yet sure to which extent that means whether this issue affects > us or not, though. Haven't really checked in detail what - apart from the obvious - has changed since https://github.com/ClusterLabs/sbd was forked from https://github.com/l-mb/sbd but sbd-pacemaker.c with the cib-connection-code is there as well - whatever that means ... But of course if you have a block-device the info gotten via that channel can of course replace health-info gotten via cib. Probably a question how you configure it ... > > Cheers, > Kristoffer > ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
Klaus Wenninger writes: > No, nothing specific to Red Hat. > sbd consists of a couple of processes talking to each other. > One is the 'inquisitor' (opening the watchdog-device and > kicking it if everything seems OK) and there are multiple > 'watchers' for the respective instances like pacemaker > (well the name says it), cluster (either exchanging cpg-messages > with corosync or checking for the existence of pacemaker_remoted) > and of course the watcher for the block-device (Red Hat > specific in that way that it is disabled in the build ;-) ). > To clarify this part, we (SUSE) haven't yet integrated the fork of SBD by Andrew which adds watchdog-only fencing, we're still relying on having a storage device. I'm not yet sure to which extent that means whether this issue affects us or not, though. Cheers, Kristoffer -- // Kristoffer Grönlund // kgronl...@suse.com ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
On 11/03/2016 07:13 PM, Adam Spiers wrote: > Klaus Wenninger wrote: >> On 11/03/2016 05:28 PM, Adam Spiers wrote: >>> Ken Gaillot wrote: ClusterLabs is happy to announce the first release candidate for Pacemaker version 1.1.16. Source code is available at: https://github.com/ClusterLabs/pacemaker/releases/tag/Pacemaker-1.1.16-rc1 The most significant enhancements in this release are: >>> [snipped] >>> * Watchdog-based fencing using sbd now works on remote nodes. >>> What were the problems with this before, exactly? Thanks! >> If you enabled just cluster-watcher on remote-nodes that >> was not much of an observation. >> >> But if you in addition enabled pacemaker-watcher then >> when the remote-node-resource switched from one >> cluster-node to another the client receiving the >> cib inside pacemaker-watcher didn't get that switch >> and still insisted on getting something via the old >> connection so that the node was reset via watchdog. >> >> Introducing a tcp-timeout derived from the >> sbd-watchdog-timeout makes the connection timeout >> and the client switches to the new control-node. >> >> So a remote-node would just be watchdog-fenced >> if the remote-node-resource doesn't reconnect within >> time - regardless which node it is running now. >> >> Actually that commit in pacemaker should be beneficial >> for tooling run on remote-nodes - via proxy - in general. > Thanks a lot for this info! I have to admit I don't fully understand, > because I don't know what pacemaker-watcher and cluster-watcher are. > Are they specific to Red Hat? No, nothing specific to Red Hat. sbd consists of a couple of processes talking to each other. One is the 'inquisitor' (opening the watchdog-device and kicking it if everything seems OK) and there are multiple 'watchers' for the respective instances like pacemaker (well the name says it), cluster (either exchanging cpg-messages with corosync or checking for the existence of pacemaker_remoted) and of course the watcher for the block-device (Red Hat specific in that way that it is disabled in the build ;-) ). > > ___ > Users mailing list: Users@clusterlabs.org > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
Klaus Wenninger wrote: > On 11/03/2016 05:28 PM, Adam Spiers wrote: > > Ken Gaillot wrote: > >> ClusterLabs is happy to announce the first release candidate for > >> Pacemaker version 1.1.16. Source code is available at: > >> > >> https://github.com/ClusterLabs/pacemaker/releases/tag/Pacemaker-1.1.16-rc1 > >> > >> The most significant enhancements in this release are: > > [snipped] > > > >> * Watchdog-based fencing using sbd now works on remote nodes. > > What were the problems with this before, exactly? Thanks! > If you enabled just cluster-watcher on remote-nodes that > was not much of an observation. > > But if you in addition enabled pacemaker-watcher then > when the remote-node-resource switched from one > cluster-node to another the client receiving the > cib inside pacemaker-watcher didn't get that switch > and still insisted on getting something via the old > connection so that the node was reset via watchdog. > > Introducing a tcp-timeout derived from the > sbd-watchdog-timeout makes the connection timeout > and the client switches to the new control-node. > > So a remote-node would just be watchdog-fenced > if the remote-node-resource doesn't reconnect within > time - regardless which node it is running now. > > Actually that commit in pacemaker should be beneficial > for tooling run on remote-nodes - via proxy - in general. Thanks a lot for this info! I have to admit I don't fully understand, because I don't know what pacemaker-watcher and cluster-watcher are. Are they specific to Red Hat? ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
On 11/03/2016 05:28 PM, Adam Spiers wrote: > Ken Gaillot wrote: >> ClusterLabs is happy to announce the first release candidate for >> Pacemaker version 1.1.16. Source code is available at: >> >> https://github.com/ClusterLabs/pacemaker/releases/tag/Pacemaker-1.1.16-rc1 >> >> The most significant enhancements in this release are: > [snipped] > >> * Watchdog-based fencing using sbd now works on remote nodes. > What were the problems with this before, exactly? Thanks! If you enabled just cluster-watcher on remote-nodes that was not much of an observation. But if you in addition enabled pacemaker-watcher then when the remote-node-resource switched from one cluster-node to another the client receiving the cib inside pacemaker-watcher didn't get that switch and still insisted on getting something via the old connection so that the node was reset via watchdog. Introducing a tcp-timeout derived from the sbd-watchdog-timeout makes the connection timeout and the client switches to the new control-node. So a remote-node would just be watchdog-fenced if the remote-node-resource doesn't reconnect within time - regardless which node it is running now. Actually that commit in pacemaker should be beneficial for tooling run on remote-nodes - via proxy - in general. > > ___ > Users mailing list: Users@clusterlabs.org > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Pacemaker 1.1.16 - Release Candidate 1
Ken Gaillot wrote: > ClusterLabs is happy to announce the first release candidate for > Pacemaker version 1.1.16. Source code is available at: > > https://github.com/ClusterLabs/pacemaker/releases/tag/Pacemaker-1.1.16-rc1 > > The most significant enhancements in this release are: [snipped] > * Watchdog-based fencing using sbd now works on remote nodes. What were the problems with this before, exactly? Thanks! ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org