January 7, 20197 yr So This happen like 2 month ago, sometime, all the docker container will halt (It just stop doing whatever it's doing and freeze) and after a while, it will start working again. One time that this happen, I was running apt-get in one of the container and apt-get freeze but it continue extracting the packages after it unfreeze Usually this isn't a problem for me, but recently, I migrated my Home Assistant Server From vSphere to unRaid and I realize that this is and huge issue when I wake up at the middle of the night and I can't turn the lights on or when I'm trying to drive to work, and my front door won't open Tower Diagnostic Attached tower-diagnostics-20190107-2052.zip
January 7, 20197 yr I'm experiencing this issue as well. I will add that while all the dockers are unresponsive, all my drives are spun up.
January 8, 20197 yr Author 8 hours ago, earthworm said: I'm experiencing this issue as well. I will add that while all the dockers are unresponsive, all my drives are spun up. I tried recreating the dockermachine img, haven't freeze so far. PS. My drive are all spun up and VMs are working as well.
January 8, 20197 yr Author Another UPDATE: All of my cpu and memmory usage seems to be at almost 100% when it froze and go down to 60% when it stop freezing Edited January 8, 20197 yr by Siwat2545
January 8, 20197 yr A memory usage of 99% means overcommitting the system. What happens if you let only a few docker containers run?
January 8, 20197 yr Do you think it's one of your VM's (eg. "Siwat PC") hogging the system? Looking though my own syslog I found the following: Jan 7 09:54:55 StorageMain kernel: Out of memory: Kill process 8878 (mono) score 456 or sacrifice child Jan 7 09:54:55 StorageMain kernel: Killed process 8878 (mono) total-vm:7063192kB, anon-rss:3708676kB, file-rss:0kB, shmem-rss:4kB Jan 7 09:54:56 StorageMain kernel: oom_reaper: reaped process 8878 (mono), now anon-rss:0kB, file-rss:0kB, shmem-rss:4kB I'm thinking maybe one of my dockers is misbehaving.
January 9, 20197 yr Author Interesting, I've assigned a sum of 30GB to VMs and all of the leftover for docker, I will now try running only one docker, will get back to you, PS. The freezing happen more often after running "docker build" command
January 9, 20197 yr Author 7 hours ago, earthworm said: Do you think it's one of your VM's (eg. "Siwat PC") hogging the system? Looking though my own syslog I found the following: Jan 7 09:54:55 StorageMain kernel: Out of memory: Kill process 8878 (mono) score 456 or sacrifice child Jan 7 09:54:55 StorageMain kernel: Killed process 8878 (mono) total-vm:7063192kB, anon-rss:3708676kB, file-rss:0kB, shmem-rss:4kB Jan 7 09:54:56 StorageMain kernel: oom_reaper: reaped process 8878 (mono), now anon-rss:0kB, file-rss:0kB, shmem-rss:4kB I'm thinking maybe one of my dockers is misbehaving. "Siwat PC" CPU Usage is at 16% when the docker freeze and stay at 6% even when the unraid dashboard says it is at 100% Edited January 9, 20197 yr by Siwat2545
January 9, 20197 yr Author I will try to screenshot the "Processes" page when it freeze, is there a way to view the docker cpu usage? Edited January 9, 20197 yr by Siwat2545
January 9, 20197 yr Author I ran top command and see user 998 taking up a lot of the memmory and with pid, I pin it down to gitlab and when I stop it, my memmory usage went down a ton(top, mem graph attached), will see if it freeze again( when I stop it, my home assistant automations fire a lot quicker, and my webserver is more responsive, plex media server decode in an instant) Here's the after Edited January 9, 20197 yr by Siwat2545
Archived
This topic is now archived and is closed to further replies.