Of them then it'd smell like an issue fixed by 79dfdaccd1d5 (memcg: How many tasks are hung in mem_cgroup_handle_oom? If there were many > killing of few processes inside cgroup so some memory is freed > - problem can be resolved by raising memory limit for cgroup or > freezes until process is killed (strace cannot be terminated by > - when i try to 'strace' any of current processes, the whole strace Is the machine under global memory pressure as well? Use_hierarchy enabled? Is the memcg oom (aka memory.oom_control)Įnabled? Do you use soft limit for those groups? Is memcg swapĪccounting enabled and memsw limits in place? What is your memcg configuration? Do you use deeper hierarchies, is > random user), out of memory is usually handled ok. > strange because it happens only sometimes (about once per week on > strange problem when a cgroup gets out of its memory limit. > i'm using memory cgroup for limiting our users and having a really To: azurIt +Cc: linux-kernel, linux-mm, cgroups mailinglist 18:05 ` azurIt 1 sibling, 1 reply 172+ messages in thread 0:26 ` Kamezawa Hiroyuki 15:24 ` Michal Hocko I really wonder why OOM is _sometimes_ not resolving this (it's usually is, only sometimes not). I'm sure that all 'freezed' processes are taking very much CPU because also server load goes really high - next time i will make a screenshot of htop. The main problem is that when this problem happens, it's NOT resolved automatically by kernel/OOM and user of cgroup, where it happend, has non-working services until i kill his processes by hand. I can look also to the data of 'freezed' proces if you need it but i will have to wait until problem occurs again. When i look to a random user PID (Apache web server): >and process's oom_adj values ? (/proc//oom_adj, /proc//oom_score_adj) >What is your memcg's memory.oom_control value ? It's expected oom-killer or memory-reclaim can solve the probelm. >Then, the problem here is the one process which uses CPU. >Under OOM in memcg, only one process is allowed to work. To: Kamezawa Hiroyuki +Cc: linux-kernel, linux-mm 21:45 ` Michal Hocko 0 siblings, 1 reply 172+ messages in thread > the body of a message to More majordomo info at > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in It's expected oom-killer or memory-reclaim can solve the probelm.Īnd process's oom_adj values ? (/proc//oom_adj, /proc//oom_score_adj) Then, the problem here is the one process which uses CPU. Because processes tends to use upĬPU at memory shortage. Under OOM in memcg, only one process is allowed to work. > I'm currently using kernel 3.2.34 but i'm having this problem since 2.6.32. > I also garbbed the content of /proc//stack of freezed process: > - problem can be resolved by raising memory limit for cgroup or killing of few processes inside cgroup so some memory is freed > - when i try to 'strace' any of current processes, the whole strace freezes until process is killed (strace cannot be terminated by CTRL-c) > - current processes are freezed and taking 100% of CPU > - no new processes can be started for this cgroup It's very strange because it happens only sometimes (about once per week on random user), out of memory is usually handled ok. > i'm using memory cgroup for limiting our users and having a really strange problem when a cgroup gets out of its memory limit. 15:24 ` Michal Hocko 1 sibling, 1 reply 172+ messages in threadįrom: Kamezawa Hiroyuki 0:26 UTC ( / raw) 19:02 memory-cgroup bug 0:26 ` Kamezawa Hiroyuki I'm currently using kernel 3.2.34 but i'm having this problem since 2.6.32. I also garbbed the content of /proc//stack of freezed process: problem can be resolved by raising memory limit for cgroup or killing of few processes inside cgroup so some memory is freed when i try to 'strace' any of current processes, the whole strace freezes until process is killed (strace cannot be terminated by CTRL-c) current processes are freezed and taking 100% of CPU no new processes can be started for this cgroup I'm using memory cgroup for limiting our users and having a really strange problem when a cgroup gets out of its memory limit. 15:24 ` Michal Hocko 0 siblings, 2 replies 172+ messages in thread Memory-cgroup bug LKML Archive on help / color / mirror / Atom feed * memory-cgroup bug 19:02 azurIt
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |