HiveOS OOM error

[winddude]

Since the last upgrade (0.6-165@201012) it appears the kernel is running out of memory after a few hours. I have 4gb of ram, no swap.

Oct 14 23:43:07 rig-002 kernel: [ 8262.505598] PhoenixMiner invoked oom-killer:
...
Oct 14 23:43:07 rig-002 kernel: [ 8262.505753] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/,task=PhoenixMiner,pid=5604,uid=0
Oct 14 23:43:07 rig-002 kernel: [ 8262.505756] Out of memory: Kill process 5604 (PhoenixMiner) score 93 or sacrifice child
Oct 14 23:43:07 rig-002 kernel: [ 8262.505780] Killed process 5604 (PhoenixMiner) total-vm:1568504kB, anon-rss:367620kB, file-rss:0kB, shmem-rss:0kB

any idea what’s going on? This have never been an issue before and linux is usually pretty good for running on not much memory.

[winddude]

Looking like this may be an issue with phoenixminer

[HaloGenius]

… or driver issue (memory leaks) or both

[winddude]

Yup looking like mem leak, usage is climbing from aprox. 750mb until it crashes

with a lot of phoenixminer processes, not sure if this is normal

[winddude]

Strange phoenix is working on one rig after downgrading hiveOS and using an older version of phoenix. Same thing wasn’t working on the other rig so I’ve switched to teamredminer.

[pho3nix3871]

I had this issue and now it is fixed. I thought it was my USB drive (32GB) upgraded to M.2 SSD (120GB), but that’s not the issue, but it helped. Then I thought it was the miner software (phoenixminer) because the first error of OOM says to kill phoenixminer and it was using 18GB total-vm. It was crashing every 30min to 1 hour. So I switched to lolminer and no more issues.

See my screenshot.