From: robj (rob@warnerbeach.com)
Date: Wed Aug 17 2005 - 11:43:20 GMT-3
Hi,
silly question perhaps but why dont you run them as a clustered pair and use
vrrp to loadbalance the flows for different apps whilst at the same time
segeregate the traffic at layer 2 via vlans at the front and back
I use this in a production enivoroment across multiple data centers and it
works well . i also use the box to box redundant solution in a non split
site enviroment and this also works well for resillience
take a look at this link
http://www.cisco.com/en/US/products/hw/contnetw/ps792/products_installation_
and_configuration_guides_list.html
The reason why you have an stp issue is that the CSS is a switch itself! and
you are introducing a loop by not having the
devices segregated from each other
HTH
Rob
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
muath_thebest@hotmail.com
Sent: 17 August 2005 14:57
To: ccielab@groupstudy.com
Subject: CSS11500
I have 2 content switches the first one do load balancing on COMM servers
the other Contenet switch do load balancing on the application servers.The
Content switch comes with only 2 ports CSS 11500. no switch modules are
installed on the CSS applience. All servers are connected to same L2 switch
and all of them have addresses from the same subnet range.
the input to the content switches coming from firewall to a L2 switch which
will distribute a connection to each Content switch"Load Balancer" .
both Load balancers have VIP from the same subnet range as the server
subnet. NO vlans is configures in this scenario.
now, we faced a spanning tree problem that one of the load balancer work at
the time while the othe one has one of its ports in bloking state.
so can you provide me of the solution of how to let the 2 load balancer work
in parallel with no ports disabled by spantree.
This archive was generated by hypermail 2.1.4 : Sun Sep 04 2005 - 17:01:19 GMT-3