We are running GAD on a Windows NLB across 2 servers (primary/primary).
If gad fails we run a script to disable the node for that server (detect code 5064), therefore all traffic is sent to the alternative node.
However we want to to detect that gad has started again, then run a script to enable the node for that server again.
The problem is the detect code (5060) when the application is started is generated twice for gad, once when it is started, then after initialization. if we enabled the node at initial start up, agents would not be able to connect until after the initialization. We want to run the enable node script after initialization (or second start up).
is there way round for this?
The status change from initialization to running seems to be 43-20620 - but when i create an alarm condition (in GA), no active alarm is generated.
scs logs below:
12:00:48.759 Standard {gad host} gad1 GCTI-00-05064 Application terminated due to internal condition
12:00:48.759 ### LOCAL MESSAGE#=5064, 'Application terminated due to internal condition'
12:01:44.602 #### App{114,gad1} status changed to APP_STATUS_RUNNING mode=PRIMARY, [prev=APP_STATUS_INITIALIZING]
12:01:44.602 Trace {framework host} scs GCTI-43-20620 [Application gad1]: LiveStatus changed to APP_STATUS_RUNNING from APP_STATUS_INITIALIZING
12:01:44.602 ### LOCAL MESSAGE#=20620, '[Application gad1]: LiveStatus changed to APP_STATUS_RUNNING from APP_STATUS_INITIALIZING'
The problem seems to be that the status change is based on the scs application, not gad however can it be restricted to just gad? We dont want the enable script running when any application changes from initializing to running.
Thanks.
