Re: [Xen-users] Performance of live-migration on the xl-stack with 10GB-lines

["Hildebrand, Nils (BIT II 9)" <Nils.Hildebrand@xxxxxxxxxxx>]

Hi Ian,

netcat suffers from the same limitation: Just a single CPU.

I think setting up a socket pair on IP-level would be something that would 
work, since the IP stack is able to use multiple CPUs.

Or is this the same as passing an empty "-s" option?

The manpage is not that clear in that point. What does xl expect "the other 
end" to do when specifying "-s"?

Is there a technical detail description about how "xl migrate" works on 
sending/receiving side?

Kind regards

Nils




-----UrsprÃngliche Nachricht-----
Von: Ian Campbell [mailto:ian.campbell@xxxxxxxxxx] 
Gesendet: Mittwoch, 13. Mai 2015 11:29
An: Hildebrand, Nils (BIT II 9)
Cc: 'xen-users@xxxxxxxxxxxxx'
Betreff: Re: [Xen-users] Performance of live-migration on the xl-stack with 
10GB-lines

On Wed, 2015-05-13 at 07:36 +0000, Hildebrand, Nils (BIT II 9) wrote:
> What other alternatives exist to make use of the full available
> bandwidth for live-migrations on the xl-stack?

xl migrate takes a -s option to provide an alternative ssh command. So
you can pass a script which uses whatever transport you like, e.g.
netcat based to an xinetd based listener on the tgt host or something
more complex involving ssh to launch the listener and netcat to pump the
data over to it etc.

If you do go down this route it would be very useful to see a right up
(perhaps on the wiki) so I have something concrete to point the next
person at.

Ian.


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