For more information regarding the security incident at F5, the actions we are taking to address it, and our ongoing efforts to protect our customers, click here.

Forum Discussion

Andy_McGrath's avatar
Andy_McGrath
Icon for Cumulonimbus rankCumulonimbus
Nov 30, 2015

Cisco ACE 'partial-threshold 25 back-inservice 49' on F5 LTM

Working on a Cisco ACE to F5 LTM migration and come across the following Cisco ACE config that is part of a 4 server Server Farm (pool): partial-threshold 25 back-inservice 49 This keeps the ...
application delivery
LTM
Andy_McGrath's avatar
Andy_McGrath
Icon for Cumulonimbus rankCumulonimbus
Dec 03, 2015

To update anyone who might come across this in the future found several solutions:

 

  1. iCall script solution takes pool out of service if 3 of the 4 members fail. Didn't like this solution as a second script would be needed to brink the pool 'up' once at least 2 of the 4 members as available and this could create support and management issues (downside of iCall)

     

  2. Use GTM to monitor the backend servers via a second Virtual Server under a Generic Host Server, so running 4 monitors on GTM set to At Least Active 2 members to mark the Virtual Server as up. Then set a Virtual Server Dependency from the original Virtual Server to the newly created one. Only the original Virtual Servers are configured under the GTM Pool.

     

Option 2 works really well as if the LTM Virtual Server is disabled or fails then the GTM takes the Virtual Server 'down'. Also if 3 of the 4 backend servers fail the second Virtaul Server in GTM is taken 'down' and as the pool member Virtual Server is dependent on the second Virtaul Server this is marked 'down'.

 

In our case as the GTM cannot directly monitor the backend servers we have setup the original LTMs under 'Prober Pools' on the GTM so all monitoring is done from the LTM for both the LTM pool members and the GTM Virtual Servers.