[Drbd-dev] [RFC for-6.2/block V2] block: Change the granularity of io ticks from ms to ns

[Chaitanya Kulkarni]

On 12/7/22 14:32, Gulam Mohamed wrote:
> As per the review comment from Jens Axboe, I am re-sending this patch
> against "for-6.2/block".
> 

why is this marked as RFC ? are you waiting for something more to get
resolved so this can be merged ?

> 
> Use ktime to change the granularity of IO accounting in block layer from
> milli-seconds to nano-seconds to get the proper latency values for the
> devices whose latency is in micro-seconds. After changing the granularity
> to nano-seconds the iostat command, which was showing incorrect values for
> %util, is now showing correct values.
> 
> We did not work on the patch to drop the logic for
> STAT_PRECISE_TIMESTAMPS yet. Will do it if this patch is ok.
> 
> The iostat command was run after starting the fio with following command
> on an NVME disk. For the same fio command, the iostat %util was showing
> ~100% for the disks whose latencies are in the range of microseconds.
> With the kernel changes (granularity to nano-seconds), the %util was
> showing correct values. Following are the details of the test and their
> output:
> 
> fio command
> -----------
> [global]
> bs=128K
> iodepth=1
> direct=1
> ioengine=libaio
> group_reporting
> time_based
> runtime=90
> thinktime=1ms
> numjobs=1
> name=raw-write
> rw=randrw
> ignore_error=EIO:EIO
> [job1]
> filename=/dev/nvme0n1
> 
> Correct values after kernel changes:
> ====================================
> iostat output
> -------------
> iostat -d /dev/nvme0n1 -x 1
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.08    0.05   0.06   128.00   128.00   0.07   6.50
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.08    0.06   0.06   128.00   128.00   0.07   6.30
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.06    0.05   0.06   128.00   128.00   0.06   5.70
> 
>  From fio
> --------
> Read Latency: clat (usec): min=32, max=2335, avg=79.54, stdev=29.95
> Write Latency: clat (usec): min=38, max=130, avg=57.76, stdev= 3.25
> 
> Values before kernel changes
> ============================
> iostat output
> -------------
> 
> iostat -d /dev/nvme0n1 -x 1
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.08    0.06   0.06   128.00   128.00   1.07  97.70
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.08    0.06   0.06   128.00   128.00   1.08  98.80
> 
> Device            r_await w_await aqu-sz rareq-sz wareq-sz  svctm  %util
> nvme0n1              0.08    0.05   0.06   128.00   128.00   1.06  97.20
> 
>  From fio
> --------
> Read Latency: clat (usec): min=33, max=468, avg=79.56, stdev=28.04
> Write Latency: clat (usec): min=9, max=139, avg=57.10, stdev= 3.79
> 
> Changes in V2:
> 1. Changed the try_cmpxchg() to try_cmpxchg64() in function
>     update_io_ticks()as the values being compared are u64 which was giving
>     a build error on i386 and microblaze
> 
> Signed-off-by: Gulam Mohamed <gulam.mohamed at oracle.com>
> ---

I believe it has no effect on the overall performance, if so I'd
document that.

Based on the quantitative data present in the commit log this
looks good to me, I believe you did audit all drivers and places
in the block layer.

Looks good.

Reviewed-by: Chaitanya Kulkarni <kch at nvidia.com>

-ck