|
If any process interrupts the service, causing the memory high and is required to kill the process, it can be done automatically with an automation stitch.
The following article can be used to find out the process behind the high memory incident: Troubleshooting Tip: How to do initial troubleshooting of high memory utilization issues (conserve ...
Once the process is being identified, it can be set in the following template to kill it during the high memory incident.
config system automation-action
edit "kill_process"
set action-type cli-script
set script "fnsysctl killall <process_name>" <-- Input the process name; Ex: httpsd, cw_acd etc.
next
end
config system automation-trigger
edit "high_memory_trigger"
set event-type low-memory
next
end
config system automation-stitch
edit "auto_kill_process"
set trigger "high_memory_trigger"
config actions
edit 1
set action "kill_process"
next
end
next
end
There are impacts on killing a process such as the daemon being restarted and in the case of a production network, it can restart the process as well as sessions. Before using it, the scope and the risk have to be evaluated.
If the high memory event halts the firewall and all the services, killing the process can be helpful in that scenario.
Related articles:
Technical Tip: Free up memory to avoid conserve mode
Technical Tip: How conserve mode is triggered
|
