Score:0

How to diagnose intermittent freezing?

in flag

My Ubuntu box has begun to freeze intermittently, and I want to understand how I can start to diagnose and troubleshoot this?

enter image description here

  • The machine is acting as a server, sitting in my garage, hosting a small number of web applications through Docker.
  • Every once in a while, the machine becomes totally unresponsive. It does not answer web requests, and I cannot even access it by SSH. After ten minutes or so, everything is fine again.
  • When I then log into the machine via SSH, I can see by the top command that something is bogging it down hard: the normal load is <0.5 but sometimes, something is pushing it over 170! No wonder it looks to be frozen!

Update:

dmesg -T | grep ill shows that one of the Docker containers was recently killed by the system, and the same happened yesterday. That would coincide with the time when the machine became responsive again.
cat /var/log/syslog did not have entries around this tiestamp.

Evidently there is a problem with a process named mattermost. This is where I must investigate further.

[Do Mär  2 12:18:39 2023] [   3858]  1000  3858   113617      243   106496        5             0 gsd-rfkill
[Do Mär  2 12:18:39 2023] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/docker/4fef02a344354214b6acb6f36102bc59e9dc50557d5c2632d70ab3daac100a5f,task=bundle,pid=1538755,uid=998
[Do Mär  2 12:18:39 2023] Out of memory: Killed process 1538755 (bundle) total-vm:1578372kB, anon-rss:904568kB, file-rss:0kB, shmem-rss:1956kB, UID:998 pgtables:2900kB oom_score_adj:0
[Do Mär  2 12:19:23 2023] nxnode.bin invoked oom-killer: gfp_mask=0x1100cca(GFP_HIGHUSER_MOVABLE), order=0, oom_score_adj=0
[Do Mär  2 12:19:23 2023]  oom_kill_process.cold+0xb/0x10
[Do Mär  2 12:19:23 2023] [   3858]  1000  3858   113617      243   106496        5             0 gsd-rfkill
[Do Mär  2 12:19:23 2023] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/docker/83ee973eca02114fae81d46c67fd021c6fad40eb124609eff8f6067ff22335dc,task=mattermost,pid=3800325,uid=2000
[Do Mär  2 12:19:23 2023] Out of memory: Killed process 3800325 (mattermost) total-vm:25852356kB, anon-rss:4436936kB, file-rss:0kB, shmem-rss:0kB, UID:2000 pgtables:9268kB oom_score_adj:0
[Fr Mär  3 10:31:08 2023] systemd invoked oom-killer: gfp_mask=0x1100cca(GFP_HIGHUSER_MOVABLE), order=0, oom_score_adj=0
[Fr Mär  3 10:31:08 2023]  oom_kill_process.cold+0xb/0x10
[Fr Mär  3 10:31:08 2023] [   3858]  1000  3858   113617      245   106496        5             0 gsd-rfkill
[Fr Mär  3 10:31:08 2023] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/docker/6ceafe411e36e8c1ceb396ae0fe9e0772399b4d981f0c3b9f108c4b637300d6f,task=bundle,pid=2216864,uid=998
[Fr Mär  3 10:31:08 2023] Out of memory: Killed process 2216864 (bundle) total-vm:1627016kB, anon-rss:922008kB, file-rss:0kB, shmem-rss:1956kB, UID:998 pgtables:2932kB oom_score_adj:0
[Fr Mär  3 10:31:46 2023] nxnode.bin invoked oom-killer: gfp_mask=0x1100cca(GFP_HIGHUSER_MOVABLE), order=0, oom_score_adj=0
[Fr Mär  3 10:31:46 2023]  oom_kill_process.cold+0xb/0x10
[Fr Mär  3 10:31:46 2023] [   3858]  1000  3858   113617      245   106496        5             0 gsd-rfkill
[Fr Mär  3 10:31:46 2023] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/docker/66ecc2904f86807abf8e5a3e5fb9310028a03c19d09f6b8973eb2c6801c943f7,task=mattermost,pid=2425866,uid=2000
[Fr Mär  3 10:31:46 2023] Out of memory: Killed process 2425866 (mattermost) total-vm:101008668kB, anon-rss:4685940kB, file-rss:0kB, shmem-rss:0kB, UID:2000 pgtables:10188kB oom_score_adj:0
Artur Meinild avatar
vn flag
Maybe include what's happening in the `syslog` or `dmesg` around the times when it becomes unresponsive?
Torben Gundtofte-Bruun avatar
in flag
Mange tak, @ArturMeinild - that was a helpful question that led me to find a memory-eating process. *Why* that happens is a different topic!
Artur Meinild avatar
vn flag
Sure - at least the invocation of `oom-kill` explains why the system becomes unresponsive.
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.