Messages in this thread | | | Date | Thu, 9 Aug 2012 16:38:24 -0600 | From | "Jim Schutt" <> | Subject | Re: [RFC PATCH 0/5] Improve hugepage allocation success rates under load V3 |
| |
On 08/09/2012 02:46 PM, Mel Gorman wrote: > On Thu, Aug 09, 2012 at 12:16:35PM -0600, Jim Schutt wrote: >> On 08/09/2012 07:49 AM, Mel Gorman wrote: >>> Changelog since V2 >>> o Capture !MIGRATE_MOVABLE pages where possible >>> o Document the treatment of MIGRATE_MOVABLE pages while capturing >>> o Expand changelogs >>> >>> Changelog since V1 >>> o Dropped kswapd related patch, basically a no-op and regresses if fixed (minchan) >>> o Expanded changelogs a little >>> >>> Allocation success rates have been far lower since 3.4 due to commit >>> [fe2c2a10: vmscan: reclaim at order 0 when compaction is enabled]. This >>> commit was introduced for good reasons and it was known in advance that >>> the success rates would suffer but it was justified on the grounds that >>> the high allocation success rates were achieved by aggressive reclaim. >>> Success rates are expected to suffer even more in 3.6 due to commit >>> [7db8889a: mm: have order> 0 compaction start off where it left] which >>> testing has shown to severely reduce allocation success rates under load - >>> to 0% in one case. There is a proposed change to that patch in this series >>> and it would be ideal if Jim Schutt could retest the workload that led to >>> commit [7db8889a: mm: have order> 0 compaction start off where it left]. >> >> On my first test of this patch series on top of 3.5, I ran into an >> instance of what I think is the sort of thing that patch 4/5 was >> fixing. Here's what vmstat had to say during that period: >> >> <SNIP> > > My conclusion looking at the vmstat data is that everything is looking ok > until system CPU usage goes through the roof. I'm assuming that's what we > are all still looking at.
I'm concerned about both the high CPU usage as well as the reduction in write-out rate, but I've been assuming the latter is caused by the former.
<snip>
> > Ok, this is an untested hack and I expect it would drop allocation success > rates again under load (but not as much). Can you test again and see what > effect, if any, it has please? > > ---8<--- > mm: compaction: back out if contended > > ---
<snip>
Initial testing with this patch looks very good from my perspective; CPU utilization stays reasonable, write-out rate stays high, no signs of stress. Here's an example after ~10 minutes under my test load:
2012-08-09 16:26:07.550-06:00 vmstat -w 4 16 procs -------------------memory------------------ ---swap-- -----io---- --system-- -----cpu------- r b swpd free buff cache si so bi bo in cs us sy id wa st 21 19 0 351628 576 37835440 0 0 17 44394 1241 653 6 20 64 9 0 11 11 0 365520 576 37893060 0 0 124 2121508 203450 170957 12 46 25 17 0 13 16 0 359888 576 37954456 0 0 98 2185033 209473 171571 13 44 25 18 0 17 15 0 353728 576 38010536 0 0 89 2170971 208052 167988 13 43 26 18 0 17 16 0 349732 576 38048284 0 0 135 2217752 218754 174170 13 49 21 16 0 43 13 0 343280 576 38046500 0 0 153 2207135 217872 179519 13 47 23 18 0 26 13 0 350968 576 37937184 0 0 147 2189822 214276 176697 13 47 23 17 0 4 12 0 350080 576 37958364 0 0 226 2145212 207077 172163 12 44 24 20 0 15 13 0 353124 576 37921040 0 0 145 2078422 197231 166381 12 41 30 17 0 14 15 0 348964 576 37949588 0 0 107 2020853 188192 164064 12 39 30 20 0 21 9 0 354784 576 37951228 0 0 117 2148090 204307 165609 13 48 22 18 0 36 16 0 347368 576 37989824 0 0 166 2208681 216392 178114 13 47 24 16 0 28 15 0 300656 576 38060912 0 0 164 2181681 214618 175132 13 45 24 18 0 9 16 0 295484 576 38092184 0 0 153 2156909 218993 180289 13 43 27 17 0 17 16 0 346760 576 37979008 0 0 165 2124168 198730 173455 12 44 27 18 0 14 17 0 360988 576 37957136 0 0 142 2092248 197430 168199 12 42 29 17 0
I'll continue testing tomorrow to be sure nothing shows up after continued testing.
If this passes your allocation success rate testing, I'm happy with this performance for 3.6 - if not, I'll be happy to test any further patches.
I really appreciate getting the chance to test out your patchset.
Thanks -- Jim
| |