Re: [PATCH 0/6] Improve visibility of writeback

[Kemeng Shi]

on 3/21/2024 1:22 AM, Jan Kara wrote:
> On Wed 20-03-24 19:02:16, Kemeng Shi wrote:
>> This series tries to improve visilibity of writeback. Patch 1 make
>> /sys/kernel/debug/bdi/xxx/stats show writeback info of whole bdi
>> instead of only writeback info in root cgroup. Patch 2 add a new
>> debug file /sys/kernel/debug/bdi/xxx/wb_stats to show per wb writeback
>> info. Patch 4 add wb_monitor.py to monitor basic writeback info
>> of running system, more info could be added on demand. Rest patches
>> are some random cleanups. More details can be found in respective
>> patches. Thanks!
>>
>> Following domain hierarchy is tested:
>>                 global domain (320G)
>>                 /                 \
>>         cgroup domain1(10G)     cgroup domain2(10G)
>>                 |                 |
>> bdi            wb1               wb2
>>
>> /* all writeback info of bdi is successfully collected */
>> # cat /sys/kernel/debug/bdi/252:16/stats:
>> BdiWriteback:              448 kB
>> BdiReclaimable:        1303904 kB
>> BdiDirtyThresh:      189914124 kB
>> DirtyThresh:         195337564 kB
>> BackgroundThresh:     32516508 kB
>> BdiDirtied:            3591392 kB
>> BdiWritten:            2287488 kB
>> BdiWriteBandwidth:      322248 kBps
>> b_dirty:                     0
>> b_io:                        0
>> b_more_io:                   2
>> b_dirty_time:                0
>> bdi_list:                    1
>> state:                       1
>>
>> /* per wb writeback info is collected */
>> # cat /sys/kernel/debug/bdi/252:16/wb_stats:
>> cat wb_stats
>> WbCgIno:                    1
>> WbWriteback:                0 kB
>> WbReclaimable:              0 kB
>> WbDirtyThresh:              0 kB
>> WbDirtied:                  0 kB
>> WbWritten:                  0 kB
>> WbWriteBandwidth:      102400 kBps
>> b_dirty:                    0
>> b_io:                       0
>> b_more_io:                  0
>> b_dirty_time:               0
>> state:                      1
>> WbCgIno:                 4284
>> WbWriteback:              448 kB
>> WbReclaimable:         818944 kB
>> WbDirtyThresh:        3096524 kB
>> WbDirtied:            2266880 kB
>> WbWritten:            1447936 kB
>> WbWriteBandwidth:      214036 kBps
>> b_dirty:                    0
>> b_io:                       0
>> b_more_io:                  1
>> b_dirty_time:               0
>> state:                      5
>> WbCgIno:                 4325
>> WbWriteback:              224 kB
>> WbReclaimable:         819392 kB
>> WbDirtyThresh:        2920088 kB
>> WbDirtied:            2551808 kB
>> WbWritten:            1732416 kB
>> WbWriteBandwidth:      201832 kBps
>> b_dirty:                    0
>> b_io:                       0
>> b_more_io:                  1
>> b_dirty_time:               0
>> state:                      5
>>
>> /* monitor writeback info */
>> # ./wb_monitor.py 252:16 -c
>>                   writeback  reclaimable   dirtied   written    avg_bw
>> 252:16_1                  0            0         0         0    102400
>> 252:16_4284             672       820064   9230368   8410304    685612
>> 252:16_4325             896       819840  10491264   9671648    652348
>> 252:16                 1568      1639904  19721632  18081952   1440360
>>
>>
>>                   writeback  reclaimable   dirtied   written    avg_bw
>> 252:16_1                  0            0         0         0    102400
>> 252:16_4284             672       820064   9230368   8410304    685612
>> 252:16_4325             896       819840  10491264   9671648    652348
>> 252:16                 1568      1639904  19721632  18081952   1440360
>> ...
> 
> So I'm wondering: Are you implementing this just because this looks
> interesting or do you have a real need for the functionality? Why?
Hi Jan, I added debug files to test change in [1] which changes the way how
dirty background threshold of wb is calculated. Without debug files, we could
only monitor writeback to imply that threshold is corrected.
In current patchset, debug info has not included dirty background threshold yet,
I will add it when discution of calculation of dirty background threshold in [1]
is done.
The wb_monitor.py is suggested by Tejun in [2] to improve visibility of writeback.
The script is more convenient than trace to monitor writeback behavior of the running
system.

Thanks

[1] https://lore.kernel.org/lkml/a747dc7d-f24a-08bd-d969-d3fb35e151b7@xxxxxxxxxxxxxxx/
[2] https://lore.kernel.org/lkml/ZcUsOb_fyvYr-zZ-@xxxxxxxxxxxxxxx/
> 
> 								Honza
>