Florian, Thanks again OK, what you're saying that highly available iSCSI target cannot be created if DRBD is running in Dual-Primary mode between node1 and node2 and either node1 or node2 is an iscsi target /dev/drbd0 with ocfs2 on top of it for the remaining node3.
Is this the correct assumption or not? This DRBD behavior in regard to iSCSI or OCFS2 is not documented anywhere, so please clarify this for the last time. Best regards and thanks, Alex --- On Wed, 1/27/10, Florian Haas <[email protected]> wrote: From: Florian Haas <[email protected]> Subject: Re: [Linux-HA] ERROR: Device "drbd0": another node is heartbeating in our slot! To: "General Linux-HA mailing list" <[email protected]> Date: Wednesday, January 27, 2010, 4:51 PM On 2010-01-27 17:38, Hunny Bunny wrote: > Hello Florian, > Thanks for your prompt reply. > However, your "Won't work" doesn't give me any definitive answers to this > problem. > Is it a deficiency of DRBD or OCFS2 which doesn't allow to achieve such a > simple task as > sharing OCFS2 partition on a top of /dev/drbd0 block between three nodes? > I admit, I might don't understand something obvious to you. > But in this case scenario how would you share OCFS2 partition on top of DRBD > between three nodes? > > Thank you in advance for any ideas, I believe Lars already explained that in considerable detail why it's not working. And if you want OCFS over 3 nodes _and_ DRBD redundancy, create a highly available iSCSI target (using regular Single-Primary DRBD), then mount that iSCSI target from 3 nodes. Cheers, Florian -----Inline Attachment Follows----- _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
