I have installed SQM-1.4.2 in a solaris environment, based on 4 netra machines 
that are load-balanced by a cisco balancer in a 6500 switch.

In short, while Squirrel works fine on one machine, with an acceptable speed, 
i see a slowness in loading the message list (i.e. the standard frame after 
login) when i use all the 4 nodes; waiting time of the orded of 1 or 2 
minutes.

all the nodes share an NFS volume for the session files so the sessions should 
be migrating from one node to the other along with the http connection by the 
load balancing switch.

in fact, i have *no* problems with the migration of the sessions, only this 
SLOWNESS that i see when i go from one node to another.

from tracing i see that the slowness is only when i jump from one node to the 
other, and when i am in the middle of the http/1.1 persistent connection i 
get the usual speed.

i have investigated the more obvious DNS errors to no avail.

all the nodes in the apache config have a servername that squash the node name 
to "webmail" and all the configuration of squirrell use the "webmail" name 
that maps for the world to the load balancer virtual address and for every 
node to itself via an /etc/hosts entry.

i have checked the session files and here i can see the base url as 
http://webmail..... which is correct. 

has anyone seen this behavior before?

bye all
 Andrea


-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive?  Does it
help you create better code?  SHARE THE LOVE, and help us help
YOU!  Click Here: http://sourceforge.net/donate/
--
squirrelmail-users mailing list
List Address: [EMAIL PROTECTED]
List Archives:  http://sourceforge.net/mailarchive/forum.php?forum_id=2995
List Info: https://lists.sourceforge.net/lists/listinfo/squirrelmail-users

Reply via email to