[arch-general] Help diagnosing kworker 'bug'

[Anatol Pomozov]

Hi

On Tue, Aug 5, 2014 at 4:37 PM, Oon-Ee Ng <ngoonee.talk at gmail.com> wrote:
> Yesterday night I noticed (just before performing an update my conky
> showing high continuous writing to root. iotop -Pa shows this:-
>
> Total DISK READ :       0.00 B/s | Total DISK WRITE :       8.93 M/s
> Actual DISK READ:       0.00 B/s | Actual DISK WRITE:      11.06 M/s
>   PID  PRIO  USER     DISK READ  DISK WRITE  SWAPIN     IO>    COMMAND
>   112 be/4 root          0.00 B     64.91 M  0.00 % 24.34 % [kworker/u16:3]
> 11936 be/4 root          0.00 B      0.00 B  0.00 %  0.06 % [kworker/1:1]
> 28794 be/4 root          0.00 B     36.00 K  0.00 %  0.00 % [kworker/u16:1]
>
> This is after roughly 6-7 seconds, and 65 MB has already been written
> by that kworker thread.

How long this IO activity takes? Could it be some kind of automatic
defragmentation or some other internal btrfs background optimization?
A good idea is to check btrfs changelog for 3.16 kernel release.

>
> As I said, this was already happening before an upgrade. I ran the
> upgrade anyway, which upgraded linux to 3.16-2, and still got the same
> thing. Yes, I'm using [testing].
>
> Any ideas on how to proceed? Next thing I'm going to try is
> downgrading linux to 3.15, but I thought I'd post this here first in
> case I don't make it back.

'perf' is a great and very powerful tool that allow to debug problems
like this. Run '# perf top -g -p $PID' and it will show where the
process spends *cpu cycles*. It should be enough to understand what
kworker thread does. For all curious minds I highly recommend to read
this tutorial https://perf.wiki.kernel.org/index.php/Tutorial