Re: [PATCH 0/2] ext4: Fix performance regression with mballoc

[Jan Kara <jack@suse.cz>]

On Wed 24-08-22 12:40:10, Jan Kara wrote:
> Hi Stefan!
> 
> On Wed 24-08-22 12:17:14, Stefan Wahren wrote:
> > Am 23.08.22 um 22:15 schrieb Jan Kara:
> > > Hello,
> > > 
> > > So I have implemented mballoc improvements to avoid spreading allocations
> > > even with mb_optimize_scan=1. It fixes the performance regression I was able
> > > to reproduce with reaim on my test machine:
> > > 
> > >                       mb_optimize_scan=0     mb_optimize_scan=1     patched
> > > Hmean     disk-1       2076.12 (   0.00%)     2099.37 (   1.12%)     2032.52 (  -2.10%)
> > > Hmean     disk-41     92481.20 (   0.00%)    83787.47 *  -9.40%*    90308.37 (  -2.35%)
> > > Hmean     disk-81    155073.39 (   0.00%)   135527.05 * -12.60%*   154285.71 (  -0.51%)
> > > Hmean     disk-121   185109.64 (   0.00%)   166284.93 * -10.17%*   185298.62 (   0.10%)
> > > Hmean     disk-161   229890.53 (   0.00%)   207563.39 *  -9.71%*   232883.32 *   1.30%*
> > > Hmean     disk-201   223333.33 (   0.00%)   203235.59 *  -9.00%*   221446.93 (  -0.84%)
> > > Hmean     disk-241   235735.25 (   0.00%)   217705.51 *  -7.65%*   239483.27 *   1.59%*
> > > Hmean     disk-281   266772.15 (   0.00%)   241132.72 *  -9.61%*   263108.62 (  -1.37%)
> > > Hmean     disk-321   265435.50 (   0.00%)   245412.84 *  -7.54%*   267277.27 (   0.69%)
> > > 
> > > Stefan, can you please test whether these patches fix the problem for you as
> > > well? Comments & review welcome.
> > 
> > i tested the whole series against 5.19 and 6.0.0-rc2. In both cases the
> > update process succeed which is a improvement, but the download + unpack
> > duration ( ~ 7 minutes ) is not as good as with mb_optimize_scan=0 ( ~ 1
> > minute ).
> 
> OK, thanks for testing! I'll try to check specifically untar whether I can
> still see some differences in the IO pattern on my test machine.

I have created the same tar archive as you've referenced (files with same
number of blocks) and looked at where blocks get allocated with
mb_optimize_scan=0 and with mb_optimize_scan=1 + my patches. And the
resulting IO pattern looks practically the same on my test machine. In
particular in both cases files get allocated only in 6 groups, if I look
at the number of erase blocks that are expected to be touched by file data
(for various erase block sizes from 512k to 4MB) I get practically same
numbers for both cases.

Ojaswin, I think you've also mentioned you were able to reproduce the issue
in your setup? Are you still able to reproduce it with the patched kernel?
Can you help debugging while Stefan is away?

								Honza

-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR