Re: [lvs-users] ipvs connections sync and CPU usage

To: Aleksey Chudov <aleksey.chudov@xxxxxxxxx>
Subject: Re: [lvs-users] ipvs connections sync and CPU usage
Cc: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
From: Julian Anastasov <ja@xxxxxx>
Date: Wed, 28 Dec 2011 23:20:11 +0200 (EET)

On Wed, 28 Dec 2011, Aleksey Chudov wrote:

> > no matter the sync period. Who has more Active conns? Master or Backup?
> Rechecked connection counters after 24h. So they are more accurate.
> ip_vs_sync_conn patched, HZ/10, sync_threshold "3  100", "port 0 patch"
>                 PersistConn ActiveConn InActConn
> Master          3216362        4294764      4459451
> Backup          3157132        3984597      4436592
>                       98.1608%     92.7628%    99.4815%
> >> Are there any advantages in the reduction of persistence connection 
> >> timeouts and tcp, tcpfin timeouts?
> > I'm not sure the timeout values play too much for the sync traffic.
> Currently,
> > sync traffic depends on packet rates and the threshold+period
> configuration.
> > Another option is to add time-based alternative. 
> As far as I understand small connection timeouts can reduce the size of the
> connections table.
> So the question is, does the size of the connections table impact on CPU
> utilization?

        Currently, sync traffic happens on received packets.
After the last packet for the state, it does not matter for
syncing to backup how much time we wait for next packets.
Of course, with lower timeouts we have less connections,
performance can come from the decreased connection lookup


Julian Anastasov <ja@xxxxxx>

Please read the documentation before posting - it's available at: mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to

<Prev in Thread] Current Thread [Next in Thread>