Forum Discussion

Jelle_B__42488's avatar
Jelle_B__42488
Icon for Nimbostratus rankNimbostratus
Dec 04, 2012

multi tier health based load balancing

Hello all,

 

 

Ok I am kidna stuck how to aproach this. I have a requiment to load balance based on a return code provided by the backend application server. I can check if it is there or not based on a web GET monitor no problem , however the customer would liek a 3rd option next to up/down and that is "going down" allowing open sessions to finish but new once to terminate on other pool member.

 

 

now I suspect to use an irule to control states of the pool memebr but I cant find a good starting point and as such i am reaching out for some help.

 

The health check will return 3 values for all intensive purposes :

 

Green all good host usable

 

Yellow application is being scheduled to go down only excisting sessions may continue

 

Red not available.

 

Obviouslly the first and last are easy but how do I do the second one ?

 

2 Replies

  • Green all good host usable

     

    Yellow application is being scheduled to go down only excisting sessions may continue

     

    Red not available.

    isn't it normal behavior when setting "Green" as receive string?

     

     

    None (default)

     

    LTM will continue to send data on established connections as long as client is sending and server is responding. Connection management / recovery / cleanup is via standard TCP mechanics for both clientside and serverside flows.

     

     

    Use "None" if you don't want LTM to intervene in managing either side of the connection. Useful if your servers may not be accepting new connections, but should be allowed to continue servicing existing connections when marked DOWN. Also supports custom monitoring designed to support connection bleeding and other non-standard state management schemes.

    LTM: Action on Service Down by Deb

     

    https://devcentral.f5.com/tutorials/tech-tips/ltm-action-on-service-down

     

  • hmmm never interpreted this as such , I found a similar document but it didn't outline it like this.

     

     

    This would mean I could just tell the developers to basically allow the health check to see the service down , new connections are send to the other pool member(s), but will leave already established sessions open till terminated from either side. Worth a try (developers wont be happy having to code this on their end not ours).