[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-API] Storage of VM metadata


  • To: xen-api@xxxxxxxxxxxxx
  • From: George Shuklin <george.shuklin@xxxxxxxxx>
  • Date: Mon, 04 Mar 2013 01:33:23 +0400
  • Delivery-date: Sun, 03 Mar 2013 21:34:35 +0000
  • List-id: User and development list for XCP and XAPI <xen-api.lists.xen.org>



On 02.03.2013 01:37, Alex O'Ree wrote:
I keep having reliability issues with XCP 1.6 resource pools. I'm
still not sure what the cause is, however the pool masters seems to
suddenly forget that there's network cards attached. This somehow
starts a chain reaction against the remaining pool members. I've also
tried the "emergency network restart" which fails completely; timeouts
and red undescriptive error messages.

I've also check for logs but was unable to find anything of value.

The only solution I've found is to reinstalled XCP from the ISO, then
remount the SR (via a NFS share). I'm ok with this, however the real
problem is that the metadata which assoicates the VMs to the VHD seems
to be getting lost in the process (shouldn't this be stored in the SR
anyhow?). I've had all members in the pool setup for automatic
metadata backups, but I don't see anything stored on the SR that would
indicate that a backup was performed. I've previously labeled all VHD
to make this process easier, however that data gets lost too!

The irritating part is that I need to mount in each VHD to a test VM,
boot it up and try to identify it, then relabel the VHD and repeat.

So here's my questions
1) how can I prevent or restore networking when this happens?
2) where are the metadata backups stored?
3) where is the metadata stored in general?
4) where are the logs related to a situation like this?

__

All pool information is replicated between master and slave and stored in /var/xapi. Main file is state.db, and there is network.db. One of the way to reset network configuration for slave is remove network.db and restart networkd.

About strange network stuff: do you rename any of networks in pool? I saw strange effects when host join to pool with renamed networks.

If you can narrow condition to cause problem, it will help to dig in. Better is instruction step-by-step to cause the bug.


_______________________________________________
Xen-api mailing list
Xen-api@xxxxxxxxxxxxx
http://lists.xen.org/cgi-bin/mailman/listinfo/xen-api


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.