Re: numa involved in instability and swap usage despite RAM free?
On 27/06/2018 07:52, Steve Kargl wrote: > On Tue, Jun 26, 2018 at 02:39:27PM -0700, Adrian Chadd wrote: >> On Mon, 25 Jun 2018 at 11:23, Steve Kargl >> wrote: >>> >>> On Sun, Jun 24, 2018 at 12:03:29PM +0200, Alexander Leidinger wrote: I don't have hard evidence, but there is enough "smell" to open up a discussion... Short: Can it be that enabling numa in the kernel is the reason why some people see instability with zfs and usage of swap while a lot of free RAM is available? >>> >>> Interesting observation. I do have NUMA in my kernel, and swap >>> seems to be used instead of recycling freeing inactive memory. >>> Top shows >>> >>> Mem: 506M Active, 27G Inact, 98M Laundry, 2735M Wired, 1474M Buf, 1536M Free >>> Swap: 16G Total, 120M Used, 16G Free >From someone that has had memory issues since 10.1 (bug 194654), I have recently realised something that seems to make some sense to me. The arc_max setting is a limit of zfs arc and this ram gets wired to prevent it swapping, this makes sense. The vm.max_wired is a value that I had thought was ignored but now I see that these are two values of wired memory which are not connected. max_wired appears to default to 30% of kmem_size. Both of these values are added together to be reported in vm.stats.vm.v_wire_count which is the wired value shown by top. This appears to be the reason that I can see 9G wired when max_wired is at 5G The implications of this is that together (arc_max + max_wired) can be set to more than the physical installed ram. I can verify that with 8G installed and the two values add up to more than 7G you get no choice but a hard reset. Since upgrading to 16G I have been more vigilant and not allowed more than 10G to be wired so haven't had that problem in a year and a half. With the default arc_max usually set to ram minus 1G and max_wired at 5G it is easy to see that the current defaults are dangerous. I have not seen max_wired mentioned in relation to zfs but it seems that it should be considered when setting arc_max to prevent over wiring ram. Close to three weeks ago I applied review D7538 to my everyday desktop running stable/11. Until a few days ago I had no swap usage which is now at 9M. In the last few years of monitoring wired usage to try and find a solution I have not seen less than 1G of swap usage after an hour of uptime. If nothing else D7538 makes arc more willing to be released. -- FreeBSD - the place to B...Storing Data Shane Ambler ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: numa involved in instability and swap usage despite RAM free?
On Tue, Jun 26, 2018 at 02:39:27PM -0700, Adrian Chadd wrote: > On Mon, 25 Jun 2018 at 11:23, Steve Kargl > wrote: > > > > On Sun, Jun 24, 2018 at 12:03:29PM +0200, Alexander Leidinger wrote: > > > > > > I don't have hard evidence, but there is enough "smell" to open up a > > > discussion... > > > > > > Short: > > > Can it be that enabling numa in the kernel is the reason why some > > > people see instability with zfs and usage of swap while a lot of free > > > RAM is available? > > > > Interesting observation. I do have NUMA in my kernel, and swap > > seems to be used instead of recycling freeing inactive memory. > > Top shows > > > > Mem: 506M Active, 27G Inact, 98M Laundry, 2735M Wired, 1474M Buf, 1536M Free > > Swap: 16G Total, 120M Used, 16G Free > > > > Perhaps, I don't understand what is meant by inactive memory. I > > thought that this means memory is still available in the buffer > > cache, but nothing is current using what is there. > > > > Aren't there now per-domain VM counters you can query via sysctl? > Maybe they'd help in diagnosing what's going on. > I upgraded to a r335642 yesterday. I haven't seen the swapping problem, yet; although I've tried to force it. There are 158 sysctl knobs that contain the string "vm". Do you have a pointer any particular one to monitor? -- Steve ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: numa involved in instability and swap usage despite RAM free?
Hi, Aren't there now per-domain VM counters you can query via sysctl? Maybe they'd help in diagnosing what's going on. -adrian On Mon, 25 Jun 2018 at 11:23, Steve Kargl wrote: > > On Sun, Jun 24, 2018 at 12:03:29PM +0200, Alexander Leidinger wrote: > > > > I don't have hard evidence, but there is enough "smell" to open up a > > discussion... > > > > Short: > > Can it be that enabling numa in the kernel is the reason why some > > people see instability with zfs and usage of swap while a lot of free > > RAM is available? > > Interesting observation. I do have NUMA in my kernel, and swap > seems to be used instead of recycling freeing inactive memory. > Top shows > > Mem: 506M Active, 27G Inact, 98M Laundry, 2735M Wired, 1474M Buf, 1536M Free > Swap: 16G Total, 120M Used, 16G Free > > Perhaps, I don't understand what is meant by inactive memory. I > thought that this means memory is still available in the buffer > cache, but nothing is current using what is there. > > -- > Steve > ___ > freebsd-current@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: numa involved in instability and swap usage despite RAM free?
On Sun, Jun 24, 2018 at 12:03:29PM +0200, Alexander Leidinger wrote: > > I don't have hard evidence, but there is enough "smell" to open up a > discussion... > > Short: > Can it be that enabling numa in the kernel is the reason why some > people see instability with zfs and usage of swap while a lot of free > RAM is available? Interesting observation. I do have NUMA in my kernel, and swap seems to be used instead of recycling freeing inactive memory. Top shows Mem: 506M Active, 27G Inact, 98M Laundry, 2735M Wired, 1474M Buf, 1536M Free Swap: 16G Total, 120M Used, 16G Free Perhaps, I don't understand what is meant by inactive memory. I thought that this means memory is still available in the buffer cache, but nothing is current using what is there. -- Steve ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: numa involved in instability and swap usage despite RAM free?
Alexander Leidinger Alexander at leidinger.net wrote on Sun Jun 24 10:03:49 UTC 2018 : > Short: > Can it be that enabling numa in the kernel is the reason why some > people see instability with zfs and usage of swap while a lot of free > RAM is available? [It will likely be a few months before I again have access to the environment these notes are based on. It has been about a month since I last had access.] On a AMD Ryzen Threadripper 1950X (16 cores, 2 HW threads per core) I enabled: options NUMA options MAXMEMDOM=2 in fairly recent times. This is a UFS context, not a ZFS one. I'd not been explicitly controlling how things run (so using defaults). This is head with debugging disabled (via including GENERIC and overriding). I did not have the swap usage problem with doing many buildworld buildkernel (self hosted and the cross builds for several targets). Nor when did a poudriere bulk -a (with ALLOW_MAKE_JOBS=yes ). This was a FreeBSD native boot context at the time. (I usually have run the same drives under Hyper-V but have not seen the problem there either.) For native FreeBSD I used -j32 (in buildworld/buildkernel terms but also for the bulk -a) and for under-Hyper-V I used -j28 . 96 GiBytes of ECC RAM total (48 GiBytes/NUMA-node). I'm not sure how common NUMA being enabled is, nor how common various MAXMEMDOM settings are. I'd not be surprised if various folks reporting problems had not explicitly enabled NUMA, nor set some explicit MAXMEMDOM figure. It may be that they all have ZFS in common in fairly recent times. (I'm ignoring examples of long-latency I/O on the same device as some swap partitions that are in use: this gets into Out Of Memory process killing without the swap being mostly used. Some reports of swap problems have this sort of issue involved on small systems unlikely to be using ZFS.) === Mark Millard marklmi at yahoo.com ( dsl-only.net went away in early 2018-Mar) ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"