Collector instability in 6.4.4-kb.5

Collector instability in 6.4.4-kb.5


@Alon_Ashkenazi (and everyone),

I’ve been testing against 6.4.4-kb.5 on my local server on and off for a few days, and I’m seeing some new, strange behavior with the collector service. It now takes “a while” to stop. Sometimes it doesn’t seem to shutdown, and when I re-start, it comes up in a bad state.

indeni@ind-local:~/indeni_vpn$ sudo service indeni-collector stop
Stopping indeni-collector............
Not stopped; may still be shutting down or shutdown may have failed

After restart, sometimes I get this (even after waiting 10 mins):


Sometime this (even though I can ssh into the device just fine from the server):


As I’m testing, I usually just bounce the indeni-server and collector when I need to restart things, but I’ve been having these troubles, so I’ve been resorting to the imanage full restart (triton) of all services. I’m seeing these same problems even there (even though I don’t get any errors in the imanage service restart triton output).


Installed Indeni packages:
indeni-backup 6.4.4-kb.5
indeni-cognito 6.4.4-kb.5
indeni-collector 6.4.4-kb.5
indeni-ds 6.4.4-kb.5
indeni-insight-analytics 6.4.4-kb.5
indeni-server 6.4.4-kb.5
indeni-triton 6.4.4-kb.5
indeni-vigile 6.4.4-kb.5
indeni-walt 6.4.4-kb.5



I saw the service stop issue with 6.4.3 in the Lab.
So the issue is probably related to changes that were already introduced in that release.
This might be associated with a process that is identified correctly as a service.
When you run into issues, please check the actual process status using “ps aux | grep java”.


On top of what @Shouky_Dan said, I believe that sudo service indeni-collector status would be able to determine if the collector is still running or not (it basically uses ps).

If the collector doesn’t stop you could always forcibly kill it with kill -9 <pid> (find the process ID with ps -fa | grep indeni-collector).