INFO: task blocked for more than 120 seconds.
|This article/section is a stub — probably a pile of half-sorted notes, is not well-checked so may have incorrect bits. (Feel free to ignore, fix, or tell me)|
Under heavy IO load on servers you may see something like:
INFO: task nfsd:2252 blocked for more than 120 seconds. "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
...probably followed by a call trace that mentions your filesystem, and probably io_schedule and sync_buffer.
The code behind this sits in hung_task.c and is relatively new (added somewhere around 2.6.30). It is a kernel thread that detects tasks that stays in the D state for a while, typically meaning it is waiting for IO. It complains when it sees a process has been waiting on IO so long that the whole process has not been scheduled for 120 seconds (default).
It is not a crash. That is, unless you have set sysctl_hung_task_panic (in which case your host is now panicked) it's only meant as an indication of where the program was when it had to wait.
- if it happens nightly, it's probably some cron job, and load from something like updatedb.
- most likely to happen to a process that was ioniced into the idle class, in which case this this message indicates expectable ioniceness behaviour when there is something else that does IO fairly continuously for at least 120 seconds.
- if not, this can easily mean your IO system is slower than your IO use -- often specifically caused by overhead, such as that from head seeking.
- tweaking the linux io scheduler for the device may help (See Computer_data_storage_-_General_&_RAID_performance_tweaking#OS_scheduling)
- if your load is fairly sequential, you may get some relief from using the noop io scheduler (instead of cfq - though note that that disables ionice)
- if it's relatively random upping the queue depth may help
- if it happens on a fileserver, you may want to consider spreading to more fileservers, or using a parallel filesystem
- NFS seems to be a common culprit, probably because it's good at filling the writeback cache, something which implies blocking while writeback happens - which is likely to block various things related to the same filesystem. (verify)
- I've seen this mention kjournald, when the underlying RAID array was itself blocked by a desktop drive with bad sectors.