Memory limit exceeded during the process.

[devs-asl]

Hello GCP Support Team,

I have one customizable Compute Engine on GCP like a 2core vCPU, 4GB Ram, and 50GB storage and there is one application running. Lately, one process has been consuming 80% of my machine's memory. 

The below information is given the process name and directory location for your kind perusal.

ruby -Eascii-8bit:ascii-8bit /usr/sbin/google-fluentd --log /var/log/google-fluentd/google-fluentd.log --use-v1-config --suppress-repeated-stacktrace --group root --daemon /var/run/google-fluentd/google-fluentd.pid --under-supervisor

How can we solve the above mentioned problem and advise us a better solution in that case.

Regards

ASL DEVS Team.

[dionv]

Hello dev-asl,

I would recommend first creating a snapshot of your instance before any else as precaution before updating the instance to upgrade.

[kolban]

By looking at the name of the process, it appears to be part of the Compute Engine Agent package ... see here.  I am hearing that the problem seems to have only started lately.  This makes me think that some configuration/event happened on your machine where, before that, all was running normally and after that, something changed.   If it were me in your shoes, I'd start looking through the monitoring history and audit trail of the machine.  See if we can't find an exact time when the process started consuming the excessive CPU.  Once you know the date/time, you can drill into the logs to see what happened at or around that time.

I'd also be tempted to spin up a second compute engine and re-install your app.  If your problem persists in the in the new instance, then we have a recipe we can take apart to see what happens.  If the problem is gone in the new instance, switch to using that instance.

Another thought would be to un-install the agent ... see here.  Are you actually using the agent for logging/monitoring?