Recent content by sengo

According to the document two_node requires hardware fencing. Are you going to support it with hw fencing?

Do you mean using:quorum {provider: corosync_votequorumtwo_node: 1}I'm no cluster expert, but how does other softwares deal with 2 node clusters (eg VMware) without fencing and quorum disks?

Not recommended is a thing. Unsupported and impossible to enable is another...

What a pity. Two node cluster are the most common configuration for us...

I see in the docs that the minimum number of nodes for an ha confiuration is 3. Is a 2 node cluster no longer supported?

Log in ram and optionally to remote syslog collector. Swap for VMs is in the VMs storage

In VMware you can install a ROK using: 52. Why do I get a “system not supported” message when I try to install HP OEM mediaon top of VMware? To allow the HP OEM media to detect that it’s being installed on HP hardware, you will need to add the followingconfiguration line to the VM configuration...

Run a i/o benchmark with and without gfs please. In my tests the performance dropped by about 70%. If you need snapshot on a block san use lvm -s and lvconvert --merge

Oh... Any chance of getting it back in the next versions?

Why did you remove fence_SCSI agent? It use quick and effective to configure with shared block sans... I don't think it was removed from upstream

I know :( But in a test lab you don't always have ipmi or other fence devices. Just don't use it in production.

Use /bin/true as fence agent.

That's the reason they pay for support. You can get support for PVE too: http://www.proxmox.com/products/proxmox-ve/subscription-service-plans

Thank you captain obvious :| How do you think the data is replicated with drbd? Avian carriers?

Thant won't work. XFS is not a cluster aware filesystem, if you mount it on more than one host you will corrupt the filesystem.