Corosync with sheepdog has 2 major problem : - corosync is limited to a "small" numbers of nodes ( technically not a problem, on "small" clusters ) - under heavy load, network fragmentation will appear
The problems and the descriptions are on the sheepdog mailing lists. Thanks for the hard work! 2014-07-14 13:02 GMT+02:00 Alexandre DERUMIER <aderum...@odiso.com>: > >> also add zookeeper support > > >>I wonder why we need that? I found this at > https://github.com/sheepdog/sheepdog/wiki/Cluster-Management-Backends-and-dual-NIC > : > > I just add support to zooker because some user request it. (So, it they > want to manually enable it, they can, without need to recompile > pve-sheepdog) > > >>> Corosync is not recommended for production use as the method used for > synchronization is susceptible to packet loss under load. For reliable > operation, zookeeper is highly recommended. > > > >But this looks like total nonsense to me?! > > I think that it make sense if you want to manage big cluster, as corosync > is limited to around 16 nodes. > > _______________________________________________ > pve-devel mailing list > pve-devel@pve.proxmox.com > http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel >
_______________________________________________ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel