Andre Ruiz wrote:
As I understand, linux raid tools operate on partitions only, not on
full disks, as hardware based solutions.
That's what I was afraid of. Thanks for the info. I will have to
figure out a way to increase the size of the partition in the fly when
the time comes that we need to do such a thing.
In your case, it seems the raid logic would be inside virtual
machines, is this correct? You would have double network traffic. You
can create one big partition of PV type, and create more small
partitions of PV type and make them join the VG each time you grow
your disks. The raid would be on each filesystem of LVM. Still a pain
in the arse.
Yes, the RAID logic is inside the virtual machines. Double network
traffic is no problem as I have two gigabit network interfaces which I
can dedicate to the SAN. Yes, creating partitions and adding them to
the VG each time is possible but a pain. I'll have to see what else I
can work out.
Hard to switch to raid on the AoE server?
I already have RAID5 on the AoE server. But what if the CPU/RAM/mobo
on the AoE server dies? :) This is why one side of the mirror is on
one AoE server and the other side of the mirror is on another. If the
cpu node dies I can restart the domain on another one very quickly.
The only single point of failure in my cluster is the switch and
eventually I might even get another one and put half the disk nodes on
one and half on the other and cross connect the cpu nodes. :)