Re: [Xen-users] Network connections stalling - Debian Linux domU's

["Fajar A. Nugraha" <list@xxxxxxxxx>]

On Mon, Jul 21, 2014 at 2:38 AM, Henrik StÃrner <henrik@xxxxxxx> wrote:
> Hi Steve,
>
> on 20-07-2014 19:43, Steve Dawson wrote:
>
>> On 07/20/2014 12:35 PM, Henrik StÃrner wrote:
>
>>> To cut a long story short, the problem appears to be with the domU
>>> hosting the IMAP server (Dovecot). When a user connects to the
>>> IMAP service, the network connection quickly stalls. A "netstat" on
>>> the domU shows a very large Send-Q (i.e. data sent, but not
>>> acknowledged by the remote end of the connection) - usually around
>>> 25000, it ought to be quite small. A tcpdump also shows very large
>>> intervals between packets, 10-30 seconds is not uncommon.
>
>> It could be hyper threading, either disable it, or manually pin your
>> domU's cpus in pairs. I had a similar issue running a bunch of citrix
>> xenapp  (2008 windows terminal) servers with hyper threading enabled,
>> when the loads got to about 30% the load averages would skyrocket
>> suddenly, pinning the cpus fixed the problem.
>
> I didn't see any high load averages, just a stuck connection. Still,
> I'll try to have the hyperthreading disabled - the server is in a
> hosting center, so BIOS access requires a bit of work :-(

You could try pinning dom0 and domU to different vcpus first, to make
sure its on different physical core, e.g.
- dom0 -> cpu 0
- domU -> cpu >=2

Since hyperthread is still enabled, skip cpu 1 for now (since it's on
the same core as cpu0)

To help pinpoint the problem source, try also:
- disabling checksum offload on all interfaces
- do some tests with iperf, domU <-> domU, domU <-> dom0, domU <-> public

-- 
Fajar

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users