I very much appreciate all of your assistance with this, but I'm still
not sure I entirely understand why this is failing. The persistence
works for me, using the mstshash (username) cookie, as long as the
server with my running session is alive and well in the farm -- even
when I reconnect from a different IP address. However, once I cause that
server to fail a health check, and it drops from the pool, I can no
longer reconnect to that session through HAProxy. Instead, I am
redirected to one of the currently active (healthy) pool members.
Is there a specific statement that I'm missing from my config that would cause this to happen? Is there anything else you can think of to try in this situation?
Thanks,
Chris
-----Original Message-----
From: Willy Tarreau [mailto:w#1wt.eu]
Sent: Monday, March 21, 2011 3:03 PM
To: Shaw, Christopher
Cc: haproxy#formilux.org
Subject: Re: Question about using "force-persist"
On Mon, Mar 21, 2011 at 12:13:40PM -0500, Shaw, Christopher wrote:
> We are using the "mstshash" cookie, as opposed to the "msts" cookie.
> Is there a difference in functionality between those two cookies? Is
> there one we should be using over the other?
Yes, mstshash contains the user name, which can only be used as a load balancing method. If the size of the server farm changes (eg: change in weight or in number of servers UP), the load balancing will change. The msts cookie contains the server's IP address, which can be used as a persistence method, whatever the number of servers, because haproxy is able to spot the correct server based on the cookie.
Regards,
Willy
This archive was generated by hypermail 2.2.0 : 2011/03/22 19:00 CET