NimbostratusA customer of us wants a redirect to a webpage in case of a LB_FAILED and a redirect to different webpage in case of maintenance.
I managed to create the fallback in case of maintenance by creating a monitor to a specific file on the backend nodes checking the content to a specific word.
Is it possible, and if yes how, to create a second option to redirect in case of a failure?
Altostratus
NimbostratusI read the above post but did not yet see how it solves my problem.
CumulonimbusNimbostratusI read the above post but did not yet see how it solves my problem.
CumulonimbusExample is:
when LB_FAILED { if { [active_members [LB::server pool]] != 0 } { HTTP::fallback "; } else { HTTP::fallback "; } }
Cumulonimbusis above iRule worked for you?
NimbostratusDo I interpret the above rule correctly that you are saying?
If loadbalancing fails with still server(s) in the pool fall back to test1
If loadbalancing fails with NO server(s) in the pool fall back to test2
CumulonimbusYep.correct.
AltostratusExample is:
when LB_FAILED { if { [active_members [LB::server pool]] != 0 } { HTTP::fallback "; } else { HTTP::fallback "; } }
Altostratusis above iRule worked for you?
NimbostratusDo I interpret the above rule correctly that you are saying?
If loadbalancing fails with still server(s) in the pool fall back to test1
If loadbalancing fails with NO server(s) in the pool fall back to test2
AltostratusYep.correct.
NimbostratusI think I found the/a solution to work with the monitor by combining your info on LB_FAILED section and the info from:
https://support.f5.com/csp/article/K12818 and https://devcentral.f5.com/codeshare?sid=693
With the monitor by using both 'receive string' and 'receive disable string' I can get the state of the pool in 'disable' or 'down'. Then using 'disable' in the when HTTP_REQUEST and 'down' in the when LB_FAILED section to redirect to different webpages.
Thanks for your input.