|
On Fri, 9 Nov 2007, Ruben Laban wrote:
> I'm aware of that. I'm no fan of 'bumping' threads to get attention in
> general, but this one has been bugging me for quite some time now.
yes it's like a thorn in a dog's paw.
>> neither end has closed the tcp connection?
>
> That's one very likely scenario. We're using LVS-DR so the
> director only sees half the traffic. Except you'd think
> (or atleast I do) that after a given timeout period, the
> entry would be purged eventually. I just checked the IPVS
> table on our backup loadbalancer, and it had ~125000
> entries! Of which ~6000 with state ESTABLISHED. The last
> failover occured october, 23rd. I'd think those entries
> should have expired by now?
I would have thought so too. Are you using our ip_vs module
or the SuSE market enhanced ip_vs? They may be the same, but
if they're different, we don't know what they did to it. I
would expect if this was a regular bug someone else would
have seen it too.
> I'm not fond of having to enable persistency on our
> loadbalancers, but the web application that's behind it,
> requires it, unfortunately. I've looked into the -SH
> scheduler as well a while ago. But I couldn't find enough
> information on it.
there's an example setup from someone who's using it to
satisfy affinity. Presumably you've found that.
> Especially the behaviour when one realserver goes
> down.
undefined for the current connection. Presumably tcpip will
drop the connection eventually. What the application does is
something else. It's whatever happens if you pull the
ethernet cable on that connection.
> Will the hashing scheme cause all connection to be kicked
> to another realserver because of the redistribution of the
> hash due to the number of realservers changing?
the failover scripts will re-run ipvsadm without the failed
realserver
Joe
--
Joseph Mack NA3T EME(B,D), FM05lw North Carolina
jmack (at) wm7d (dot) net - azimuthal equidistant map
generator at http://www.wm7d.net/azproj.shtml
Homepage http://www.austintek.com/ It's GNU/Linux!
|
