09-20-2007 02:43 AM
Primary serverfarm with stickiness (cookie insert) goes down. Backup serverfarm kicks in with sorry servers. Primary serverfarm comes back up and returning connections still get serviced by the backup farm. The reason for this is explained in the load balancing guide.
[quote]
If you want to configure a sorry server farm and you want existing connections to revert to the primary server farm after it comes back up, do not use stickiness.
[/quote]
The big questions is. How can i offer a serverfarm with stickiness and a sorry serverfarm without stickiness?
Roble
09-26-2007 07:34 AM
To instruct the ACE to replicate HTTP-cookie sticky table entries on the standby ACE, use the replicate sticky command. Use the no form of this command to restore the behavior of the ACE to the default of not replicating HTTP-cookie sticky table entries
09-26-2007 01:52 PM
To configure serverfarm with stickiness and a sorry serverfarm without stickiness, use
sticky http-cookie Mycookie group1
cookie insert Acecookie
serverfarm sf1 backup sf2
To configure serverfarm & Backup serverfarm with stickiness use
sticky http-cookie Mycookie group1
cookie insert Acecookie
serverfarm sf1 backup sf2 sticky
Syed
09-26-2007 09:52 PM
Hi Syed,
unfortunately that's the theory. If you have cookie inserts the clients are still stuck to the sorry servers once they have hit the sorry servers.
And my sticky group looks exactly like your first example. The documentation is kind off confusing. My interpretation of the quoted text was if your primary serverfarm is sticky your backup server farm will also be sticky no matter what you configure. That is actually why i asked.
If i use dynamic cookies from the application the clients hop from one rserver to another every 2nd or 3rd connection. The behavior of the stickiness and sorry servers then works like i want it but the stickiness itself is not consistent.
1.5a still had this behavior and i think this might be a bug.
Roble
09-26-2007 10:32 PM
I admit I never tested it with cookie insert.
In any case I found it.
you are hitting
CSCsi61812
Syed
09-26-2007 11:47 PM
Thanks for the quick reply, the bug is not visible for me. You seem to have other means of accessing that bug repository.
Cisco needs to issue some kind of Bugbait award i think i might qualify.
Does CSCsi61812 description say that it is fixed in A1.61?
Roble
10-04-2007 02:18 AM
Finally i have been able to check out the bug id.
According to the bug id this is "expected behavior". But i can't think of a scenario where this would be useful. So i mark this as a design flaw.
So my next step will be probably a tac call for the dynamic cookie issue and the not working stickiness.
Roble
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: