Hi,
I am an ACE newbie - I have a two-tier ACE setup and I am basically trying to get the front-end ACE to divert to a sorry page if the back end servers hanging of the Back-end ACE do not reply to their probes.
I have the following setup...
Internet
|
DMZ ACE (doing SSL termination)
|
Reverse Proxy Server farm
|
Corporate LAN ACE
|
Application Server farm
DMZ ACE is probing Rev Proxy farm on TCP 2000 - and using sticky cookie insertion.
Corporate LAN ACE is probing App Server farm on TCP 2000 - and using sticky cookie insertion.
If the Application server farm becomes unavailable, I would like the DMZ ACE to detect this and then redirect the clients to a 'service unavailable' page hosted on the Reverse Proxy Servers.
My thought so far is the following...
DMZ ACE
rserver Rev_proxy1
...
rserver Rev_proxy2
...
probe icmp probe_icmp
ip address <App_Server_VIP>
serverfarm Rev_proxy_farm
probe probe_icmp
prove probe_tcp_2000
rserver Rev_proxy1, Rev_proxy2
So the above Rev_proxy_farm availability is tied to the appearance of the App Server vip due to the directed icmp probe to the Corporate LAN ACE VIP - the VIP will disappear if the App Server farm does not respond to it's TCP probe.
I am then not sure how to redirect the HTTP request to the Reverse Proxy Server seeing as though these have already been flagged unavailable.
Should I then follow 'Configuring a Sorry Server Farm' as per http://www.cisco.com/en/US/docs/interfaces_modules/services_modules/ace/v3.00_A1/configuration/slb/guide/classlb.html#wp1049254 to divert the connections from the Reverse_proxy:2000 to Reverse_proxy:3000 (which serves Service unavailable page)?
Any advice on whether this is the best way to go would be much appreciated.
Cheers,
Al