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

Re: [Xen-API] [XCP-1.6-BETA] SXM migration - ping timeouts


  • To: xen-api@xxxxxxxxxxxxx
  • From: George Shuklin <george.shuklin@xxxxxxxxx>
  • Date: Wed, 10 Oct 2012 00:30:35 +0400
  • Delivery-date: Tue, 09 Oct 2012 20:30:45 +0000
  • List-id: User and development list for XCP and XAPI <xen-api.lists.xen.org>

It's really strange. xs tools for pv-domain is just simple shell script to report vm internals back to XCP.

You can fake it presence with this script (run on host where VM resides):
http://wiki.xensource.com/wiki/XAPI_fake_presence_of_PV_drivers

But behavior is really funny. May be PV presence cause live migration and 'no guest tools' cause nomal migration (with pause and without iterative incremental memory transfer)?

On 09.10.2012 23:23, Burnie wrote:
When migrating a VM from one standalone server to another, I noticed
that there's a huge gain in making sure the VM has xs-tools installed:

VM: Centos 6.3 x86_64 - PV

- with xs-tools installed:
206 packets transmitted, 205 received, 0% packet loss, time 205680ms
rtt min/avg/max/mdev = 0.204/48.439/197.497/78.066 ms

- without xs-tools installed:
209 packets transmitted, 157 received, 24% packet loss, time 208432ms
rtt min/avg/max/mdev = 0.217/66.975/196.220/84.136 ms

In both cases the VM seems to appear in XC within 1-2 seconds on the
new server, but without xs-tools, it doesn't seem to do any networking.

I dont know if this is normal or not, but I guess it has something to
do with some optimized drivers/modules?


_______________________________________________
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®.