Hi, On Mon, Jan 24, 2011 at 06:24:31PM +0100, Ante Karamatic wrote: > Hi all > > I've set up clvmd and corosync. It's working fine when all the nodes are > in the cluster. But, when I cleanly reboot/poweroff one of the nodes, > locking of clustered volume group isn't working well. > > I've 'greped' trough archives and found out that I'm not alone. Whole > problem is nicely explained at: > > http://www.mentby.com/xinwei-hu/vgscan-fails-when-other-nodes-quit-cleanly.html > > Basically, vgscan returns 'cluster request failed: Host is down'. I've > reverted the commit Xinwei mentions and everything is working fine now. > > Still, I'm not sure that's the right approach. I'm using 2.02.54 > version, but I've also tried latest versions. Unfortunately with the > same results. clvmd/lvm is compiled with --with-clvmd=all. CKPT service > is enabled in corosync. > > Could someone give me an input on what's going on here and where should > I look to fix this?
I think that you should post to another list where LVM is discussed. This is only cluster libraries (clplumbing), LRM, stonith, and resource agents. Thanks, Dejan > Thank you > > -- > Ante Karamatic <iv...@ubuntu.com> > > _______________________________________________________ > Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev > Home Page: http://linux-ha.org/ _______________________________________________________ Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev Home Page: http://linux-ha.org/