Hi Mark,
Am Mittwoch, den 13.01.2010, 17:38 -0500 schrieb Yeary, Mark A.:
> Hi Thomas,
>
> I should have been more clear in my explanation. We have a two
> physical node drbd cluster with GFS2. We have tried to add two more
> nodes by increasing the gfs2 journal to four, the two in the drbd
> cluster and two separate nodes; all four with gfs2 formatting. All four
> nodes are connected via virtual machine (KVM) virtio disks.
So your hosts are vms?
> We can
> access them but they do not maintain synchronization.
.. because they all access _one common_ blockdevice passed through by
kvm? Two of them play drbd with this device while the others do not?
> Becoming frustrated (only a little :-), we went back to the two
> with drbd and gfs2. We then tried to connect with two more nodes (a
> total of four) via iscsi. That didn't seem to work either.
Why?
Just dd if=/dev/zero of=/MOUNTPOINT_GFS/10GB bs=1M count=10000
And add this as a Lun to e.g. IET
> Anything
> you can tell us would be helpful :-)
We should get more clearness about what you need - what is your goal?
Do you want to access a Filesystem from all this hosts?
Well known NFS with or GFS/OCFS2 on DRBD with iSCSI, AoE or FCoE
could be a Solution...
Or do you want to Share your _Raw Storage_?
iSCSI, AoE, FCoE on LVM on top of DRBD would be nice...
>
> Truly,
>
> Mark
>
cu,
thomas
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
