|
John Reuning wrote:
>
> Ah, I completely forgot that we have 5 min. of persistence for
> session-based web applications.
5 mins is still a lot less than an hour. After 5mins all your connections
should be new (well unless you have all your connections coming from
within the persistence netmask, in which case connections will never drop,
but in that case, you will have nothing going to the new realserver).
> We use NAT, so the connection tracking/mapping might be part of the
> trouble.
actually in LVS-NAT, the director knows exactly the state of
each connection since it sees the packets in each direction.
> I've had problems with incorrect SNAT from the real servers,
> so there exists some degree of disconnect between ip_vs and netfilter.
LVS and netfilter aren't real compatible - they tread on each other's
toes - look for "Antefacto" in the HOWTO. Be careful here.
> Here's another question: at what point in the connection negotiation
> does lvs establish a connection?
tcp establishes the connection. If you mean "creates an entry in the ipvs
table" then I don't know, but I'll make a guess and let someone else step
in if I'm wrong. The only way I can see it
work is after the SYN. In LVS-DR the SYN/ACK goes out via the realserver
and it can only go out via a realserver if a realserver has been chosen,
and that can only occur if there's an entry in the ip_vs table.
> In
> what state would the real server's application have to be for the port
> to remain open, yet lvs not register new connections? Is that even
> possible?
I don't think I understand your question. The application on the realserver
is in ESTABLISHED and then what happens?
In the meantime read the section on HOW LVS works (you'll need several
doses of your favorite beverage).
Joe
--
Joseph Mack PhD, High Performance Computing & Scientific Visualization
SAIC, Supporting the EPA Research Triangle Park, NC 919-541-0007
Federal Contact - John B. Smith 919-541-1087 - smith.johnb@xxxxxxx
|
