On Wed, Dec 18, 2019 at 07:37:16PM +0100, SeongJae Park wrote:
> From: SeongJae Park <[email protected]>
>
> Each `blkif` has a free pages pool for the grant mapping. The size of
> the pool starts from zero and is increased on demand while processing
> the I/O requests. If current I/O requests handling is finished or 100
> milliseconds has passed since last I/O requests handling, it checks and
> shrinks the pool to not exceed the size limit, `max_buffer_pages`.
>
> Therefore, host administrators can cause memory pressure in blkback by
> attaching a large number of block devices and inducing I/O. Such
> problematic situations can be avoided by limiting the maximum number of
> devices that can be attached, but finding the optimal limit is not so
> easy. Improper set of the limit can results in memory pressure or a
> resource underutilization. This commit avoids such problematic
> situations by squeezing the pools (returns every free page in the pool
> to the system) for a while (users can set this duration via a module
> parameter) if memory pressure is detected.
>
> Discussions
> ===========
>
> The `blkback`'s original shrinking mechanism returns only pages in the
> pool which are not currently be used by `blkback` to the system. In
> other words, the pages that are not mapped with granted pages. Because
> this commit is changing only the shrink limit but still uses the same
> freeing mechanism it does not touch pages which are currently mapping
> grants.
>
> Once memory pressure is detected, this commit keeps the squeezing limit
> for a user-specified time duration. The duration should be neither too
> long nor too short. If it is too long, the squeezing incurring overhead
> can reduce the I/O performance. If it is too short, `blkback` will not
> free enough pages to reduce the memory pressure. This commit sets the
> value as `10 milliseconds` by default because it is a short time in
> terms of I/O while it is a long time in terms of memory operations.
> Also, as the original shrinking mechanism works for at least every 100
> milliseconds, this could be a somewhat reasonable choice. I also tested
> other durations (refer to the below section for more details) and
> confirmed that 10 milliseconds is the one that works best with the test.
> That said, the proper duration depends on actual configurations and
> workloads. That's why this commit allows users to set the duration as a
> module parameter.
>
> Memory Pressure Test
> ====================
>
> To show how this commit fixes the memory pressure situation well, I
> configured a test environment on a xen-running virtualization system.
> On the `blkfront` running guest instances, I attach a large number of
> network-backed volume devices and induce I/O to those. Meanwhile, I
> measure the number of pages that swapped in (pswpin) and out (pswpout)
> on the `blkback` running guest. The test ran twice, once for the
> `blkback` before this commit and once for that after this commit. As
> shown below, this commit has dramatically reduced the memory pressure:
>
> pswpin pswpout
> before 76,672 185,799
> after 867 3,967
>
> Optimal Aggressive Shrinking Duration
> -------------------------------------
>
> To find a best squeezing duration, I repeated the test with three
> different durations (1ms, 10ms, and 100ms). The results are as below:
>
> duration pswpin pswpout
> 1 707 5,095
> 10 867 3,967
> 100 362 3,348
>
> As expected, the memory pressure decreases as the duration increases,
> but the reduction become slow from the `10ms`. Based on this results, I
> chose the default duration as 10ms.
>
> Performance Overhead Test
> =========================
>
> This commit could incur I/O performance degradation under severe memory
> pressure because the squeezing will require more page allocations per
> I/O. To show the overhead, I artificially made a worst-case squeezing
> situation and measured the I/O performance of a `blkfront` running
> guest.
>
> For the artificial squeezing, I set the `blkback.max_buffer_pages` using
> the `/sys/module/xen_blkback/parameters/max_buffer_pages` file. In this
> test, I set the value to `1024` and `0`. The `1024` is the default
> value. Setting the value as `0` is same to a situation doing the
> squeezing always (worst-case).
>
> If the underlying block device is slow enough, the squeezing overhead
> could be hidden. For the reason, I use a fast block device, namely the
> rbd[1]:
>
> # xl block-attach guest phy:/dev/ram0 xvdb w
>
> For the I/O performance measurement, I run a simple `dd` command 5 times
> directly to the device as below and collect the 'MB/s' results.
>
> $ for i in {1..5}; do dd if=/dev/zero of=/dev/xvdb \
> bs=4k count=$((256*512)); sync; done
>
> The results are as below. 'max_pgs' represents the value of the
> `blkback.max_buffer_pages` parameter.
>
> max_pgs Min Max Median Avg Stddev
> 0 417 423 420 419.4 2.5099801
> 1024 414 425 416 417.8 4.4384682
> No difference proven at 95.0% confidence
>
> In short, even worst case squeezing on ramdisk based fast block device
> makes no visible performance degradation. Please note that this is just
> a very simple and minimal test. On systems using super-fast block
> devices and a special I/O workload, the results might be different. If
> you have any doubt, test on your machine with your workload to find the
> optimal squeezing duration for you.
>
> [1] https://www.kernel.org/doc/html/latest/admin-guide/blockdev/ramdisk.html
>
> Signed-off-by: SeongJae Park <[email protected]>
Reviewed-by: Roger Pau Monn? <[email protected]>
Thanks, and sorry for the delay!
Roger.