RE: [Xen-users] Live migration - Networking take 2 minutes or so to return & MAX_SKBUFF_ORDER error

["Robert Dunkley" <Robert@xxxxxxxxx>]

Hi Martin,

Thanks for the help. 

I don't know if this is related but I see very odd outgoing ping results
from within the Windows VMs (Pinging the Windows VM from outside is
fine), here is an example, the times are wrong, it is definitely not
taking 8 seconds for some pings, the weird high number pings seem to
skip through very fast as if the system does not bother even sending the
ping).
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64
Reply from 10.0.0.2: bytes=32 time=82987ms TTL=64
Reply from 10.0.0.2: bytes=32 time=82987ms TTL=64
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64
Reply from 10.0.0.2: bytes=32 time=-83558ms TTL=64
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64
Reply from 10.0.0.2: bytes=32 time<1ms TTL=64

An RDP connection seems to cut network migration time down to about
20seconds so at least it helps. Pinging the router from within the VM
during migration means I only lose 4 pings and about 5 seconds downtime.
I'm guessing this issue might be down to the cheap router and switches
we use in the office, hopefully it will be OK once it is in a production
environment with much better Cisco gear.

The only other possible cause I can think of is that I get an increase
MAX_SKBUFF_ORDER error on boot:
Aug 12 09:51:52 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:51:53 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:51:55 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:51:55 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:51:59 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:52:03 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:52:03 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:52:03 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER.
Aug 12 09:52:08 mrhappy kernel: Attempt to allocate order 5 skbuff.
Increase MAX_SKBUFF_ORDER. 

It does not seem to recur after boot though. Is this SKBUFF error
anything to worry about?

Thanks,

Rob


-----Original Message-----
From: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
[mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Martin
Hierling
Sent: 11 August 2008 18:54
To: xen-users
Subject: Re: [Xen-users] Live migration - Networking take 2 minutes or
so toreturn

Hi,

> 2003 machine is very fast and I can VNC into it fine but networking
takes 2
> minutes or so to come back on line. Disabling the nic and re-enabling
it
> within the Windows VM fixes the problem straight away. Is there
something
> wrong with my Xen Config? Any know fixes for this?

perhaps your switch didnt get it? You move the MAC address from one
port to another, so cam table has to be rebuild. Easiest way, start a
RDP session during migration and generate some traffic or get a vnc
console and ping your gateway, for example. does it fix the problem?

regards Martin

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users

The SAQ Group
Registered Office: 18 Chapel Street, Petersfield, Hampshire. GU32 3DZ
SEMTEC Limited trading as SAQ is Registered in England & Wales
Company Number: 06481952

http://www.saqnet.co.uk AS29219
SAQ Group Delivers high quality, honestly priced communication and I.T. 
services to UK Business.
DSL : Domains : Email : Hosting : CoLo : Servers : Racks : Transit : Backups : 
Managed Networks : Remote Support.

Find us in http://www.thebestof.co.uk/petersfield


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users