Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi Trey, On 11/18/11 07:22, Trey Dockendorf wrote: > In preparation to begin testing DRBD I've found my superiors have new > requirements. The desired setup is 3 nodes in the cluster, replicating > a LVM volume between all 3. The volume will contain QCOW2 images for > KVM, and to have concurrent access I've been looking at using GFS2. OK, so if it's 3 nodes that a hypervisor is meant to run on, then that rules out putting DRBD-based storage and the hypervisors on the same boxes. (It certainly doesn't rule out the use of DRBD altogether; see below). > The potential complication is all 3 servers are not identical. Two are > identical, Dell 2950s. The other is a new Dell R510. The 2950s run 6 x > SATA 7200RPM in RAID 6, and the R510 has it's system on RAID1 SAS and 6 > x SAS 10000RPM in RAID 6. Is it correct that with DRBD, the combination > of mismatched performance of the disk I/O would be a problem? How much > more difficult is a 3 node cluster over 2 node? For multiple-node write access, it's impossible. What DRBD currently supports is adding a third node in a stacked configuration, but that is only useful for backup and DR purposes. You can't really think of the third node as a fully-fledged member of the virtualization cluster. > Also, if I'm able get an iSCSI, what role would DRBD play in the model > of 3 servers w/ shared storage? I assume to allow concurrent access to > the same iSCSI space, that I would have to still use a cluster file > system (live migration). Would DRBD then be used to replicate the > system partitions or with KVM is it only useful to replicate the VM > store when not using shared storage? You can put your iSCSI on DRBD-backed storage, and then use that iSCSI target as centralized storage for your hypervisors. You may not even need to set up cLVM. That's what this talk explains: http://www.hastexo.com/content/roll-your-own-cloud (Free-of-charge registration required, or just use your Google Profile or WordPress account, or anything else that supports OpenID, to log in.) You can also take a look at this Tech Guide which I wrote while working at Linbit, which is still hosted on their site: http://www.linbit.com/en/education/tech-guides/highly-available-virtualization-with-kvm-iscsi-pacemaker/ > With or without a shared storage device (most likely without), how would > failover work for the virtual servers? Is that where Pacemaker comes > in? Basically a way to trigger the still-alive servers to bring up the > VMs that were running on the failed server. Yes, watch the talk. :) Cheers, Florian -- Need help with High Availability? http://www.hastexo.com/now