R
Railwayβ€’2y ago
honzasterba

103 Killed

I thought that the limit on container memory was 8G, yet my processes are getting killed when reaching 4G, could it be that the last data point missing from the usage chart?
10 Replies
angelo
angeloβ€’2y ago
I dont think so. This is most likely an issue with your code.
honzasterba
honzasterbaβ€’2y ago
Hmm, ok, I have some idea, will investigate well the usage chart is peaking to 7GB some of the attempts, and adding the potentionally missing lib did not change anything, so I am pretty sure I am running out of memory but if you could check to confirm project/76e242e9-db08-4279-a677-6c0063fdb9ce/service/e5852326-cf78-4686-b1bf-bd7534119d29 that would be awesome
angelo
angeloβ€’2y ago
Sure thing Honza! This is my last case for today, but for you, anything πŸ™‚ Oh this is interesting so you are saying it kills when it hits these memory spikes?
angelo
angeloβ€’2y ago
honzasterba
honzasterbaβ€’2y ago
yes (sorry for late reply, it was late) @Angelo just pinging if you had a chance to look into this
angelo
angeloβ€’2y ago
No, not since I went to sleep. However, I don't seem to have any leads here. Is this some sort of cron that is running and is injesting some data?
angelo
angeloβ€’2y ago
also noticing it didn't happen since? Its leading me to believe that its an application issue :\
honzasterba
honzasterbaβ€’2y ago
as I said, I am pretty sure this is being killed because its trying to use too much RAM. I can fire of the job thats causing this at any time πŸ™‚ but it has not ran since
angelo
angeloβ€’2y ago
Yea, but thats what is very confusing- the RAM spiking behavior shouldn't behave differently on our platform vs. your local machine I assume it takes 8GB for you? Locally?
honzasterba
honzasterbaβ€’2y ago
Good question, I have not ran this exact workload locally.