Re: [linux-lvm] lvmpolld causes high cpu load issue
Il 2022-08-17 20:54 Zdenek Kabelac ha scritto: https://github.com/prajnoha/sid Thanks for sharing. From the linked page: "SID positions itself on top of udev, reacting to uevents. It is closely interlinked and cooperating with udev daemon. The udev daemon is enhanced with specialized sid udev builtin command that is used to communicate with SID. SID also listens to udev uevents issued by udev daemon which in turn triggers further processing." For my untrained eyes, it seems an *additional* component relying on udev itself, so I am not sure on how it should solve the issue at hand. But hey - anything that improve current sore spots is more than welcomed! Regards. -- Danti Gionatan Supporto Tecnico Assyoma S.r.l. - www.assyoma.it email: g.da...@assyoma.it - i...@assyoma.it GPG public key ID: FF5F32A8 ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 19:35 Gionatan Danti napsal(a): Il 2022-08-17 17:26 Zdenek Kabelac ha scritto: I like the general idea of the udev watch. It is the magic that causes newly created partitions to magically appear in the system, which is Would disabling the watch rule be a reasonable approach in this case? If the user want to scan a new device, it only needs to issue partprobe or kpartx, or am I missing something? Before diving into these 'deep waters' - I'd really like to first see if the problem is still such an issue with our upstream code base. There have been lot of minor optimization committed over the time - so the amount of fired 'watch' rules should be considerably smaller then the version mentioned in customer issue. There is on going 'SID' project - that might push the logic somewhat further, but existing 'device' support logic as is today is unfortunate 'trace' of how the design should not have been made - and since all 'original' programmers left the project long time ago - it's non-trivial to push things forward. Well, this is not good news. Just for my education, it is possibile to run a modern linux distro without udev at all? I still remember when the new cool thing for device autodiscovery was devfs (with some distro - like gentoo - taking the alternative approach to simply tarrig & untarring much of the entire /dev/ files to prepopulate the major+minor number... > We just hope the SID will make some progress (although probably small one at the beginning). Any info on the project? https://github.com/prajnoha/sid Zdenek ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Il 2022-08-17 17:26 Zdenek Kabelac ha scritto: I like the general idea of the udev watch. It is the magic that causes newly created partitions to magically appear in the system, which is Would disabling the watch rule be a reasonable approach in this case? If the user want to scan a new device, it only needs to issue partprobe or kpartx, or am I missing something? There is on going 'SID' project - that might push the logic somewhat further, but existing 'device' support logic as is today is unfortunate 'trace' of how the design should not have been made - and since all 'original' programmers left the project long time ago - it's non-trivial to push things forward. Well, this is not good news. Just for my education, it is possibile to run a modern linux distro without udev at all? I still remember when the new cool thing for device autodiscovery was devfs (with some distro - like gentoo - taking the alternative approach to simply tarrig & untarring much of the entire /dev/ files to prepopulate the major+minor number...) We just hope the SID will make some progress (although probably small one at the beginning). Any info on the project? Thanks. -- Danti Gionatan Supporto Tecnico Assyoma S.r.l. - www.assyoma.it email: g.da...@assyoma.it - i...@assyoma.it GPG public key ID: FF5F32A8 ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
-BEGIN PGP SIGNED MESSAGE- Hash: SHA512 On Wed, Aug 17, 2022 at 05:26:08PM +0200, Zdenek Kabelac wrote: > Dne 17. 08. 22 v 15:41 Martin Wilck napsal(a): > > On Wed, 2022-08-17 at 14:54 +0200, Zdenek Kabelac wrote: > > > Dne 17. 08. 22 v 14:39 Martin Wilck napsal(a): > > > > > > > > > Let's make clear we are very well aware of all the constrains > > > associated with > > > udev rule logic (and we tried quite hard to minimize impact - > > > however udevd > > > developers kind of 'misunderstood' how badly they will be impacting > > > system's > > > performance with the existing watch rule logic - and the story kind > > > of > > > 'continues' with 'systemd's' & dBus services unfortunatelly... > > > > I dimly remember you dislike udev ;-) > > Well it's not 'a dislike' from my side - but the architecture alone is just > missing in many areas... > > Dave is a complete disliker of udev & systemd all together :) I find udev useful for physical devices, but for virtual devices it is a terrible fit. It is far too slow and full of race conditions. Ideally, device-mapper ioctls would use diskseq instead of major+minor number everywhere, and devices would be named after the diskseq. > > I like the general idea of the udev watch. It is the magic that causes > > newly created partitions to magically appear in the system, which is > > Tragedy of design comes from the plain fact that there are only 'very > occasional' consumers of all these 'collected' data - but gathering all the > info and keeping all of it 'up-to-date' is getting very very expensive and > can basically 'neutralize' a lot of your CPU if you have too many resources > to watch and keep update. > > > > very convenient for users and wouldn't work otherwise. I can see that > > it might be inappropriate for LVM PVs. We can discuss changing the > > rules such that the watch is disabled for LVM devices (both PV and LV). > > It's really not fixable as is - since of the complete lack of 'error' > handling of device in udev DB (i.e. duplicate devices, various frozen > devices...) > > There is on going 'SID' project - that might push the logic somewhat > further, but existing 'device' support logic as is today is unfortunate > 'trace' of how the design should not have been made - and since all > 'original' programmers left the project long time ago - it's non-trivial to > push things forward. What is the SID project, what are its goals, and how does it plan to achieve them? - -- Sincerely, Demi Marie Obenour (she/her/hers) Invisible Things Lab -BEGIN PGP SIGNATURE- iQIzBAEBCgAdFiEEdodNnxM2uiJZBxxxsoi1X/+cIsEFAmL9EEEACgkQsoi1X/+c IsEkLhAA0/7E1rP44bEFCK76JzKvevdlxe7sRBEPvgn/1m9SiEntC47QiEjnQeNi cI9RLmHUlpYRghfDQMq8vhKk6a+NbnGsWTx3jciqQph+5SSIfPW9VuNi9w0nlvwS GPHLweMadCblWqXh8XP2RvJx1Z1QeXZ6kYbfMjhZdxY7a/vg0rXTh0XghSyrgfYs lgFbcqdJbEX5q70OGds8rhxAbTiBKnPHh3z5aFTCN7ILXO4blRWcqhDvAk0w3SQf lt5WgDBjZ+5gv2pNiNuwZIzqsgL6FDE4CcR+7JWlAakC1GcocVp87aoiR1hNGMob ZQoGaivvIjqYwSkWUDUArS8ntcKRBr/mYBcm6WuGZFbWja6NT2tEVJ8vcXdr2x5W DoPk7Vkj/Y9pOn2kcYQMKR1mGOQhq1AwimSHuzPOeWifUWM5BOkH7hS46Tyq2bZJ BM/QjUQcnckyAgPRYu+OWP3IvfOU+bFdTKabaoNgtCT85mfgL65sr8kx23ikQeZb RQ9VcbQnJceKrNsqBnCDE4Xegh96er4Gm+68Crdgs0adHOTcyC5937PPSVy99ls8 MbkdPEVGHe4L1TS8XhI6+NCf0oaFCVE/1vKeS4yO28VbSn/N3pbhiNF6cpc0sWDg NA0mbIsl19t4j8CtXVCjPeh1+RULvXqhedQIC/xJF3FserAInkc= =J7YY -END PGP SIGNATURE- ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 15:41 Martin Wilck napsal(a): On Wed, 2022-08-17 at 14:54 +0200, Zdenek Kabelac wrote: Dne 17. 08. 22 v 14:39 Martin Wilck napsal(a): Let's make clear we are very well aware of all the constrains associated with udev rule logic (and we tried quite hard to minimize impact - however udevd developers kind of 'misunderstood' how badly they will be impacting system's performance with the existing watch rule logic - and the story kind of 'continues' with 'systemd's' & dBus services unfortunatelly... I dimly remember you dislike udev ;-) Well it's not 'a dislike' from my side - but the architecture alone is just missing in many areas... Dave is a complete disliker of udev & systemd all together :) I like the general idea of the udev watch. It is the magic that causes newly created partitions to magically appear in the system, which is Tragedy of design comes from the plain fact that there are only 'very occasional' consumers of all these 'collected' data - but gathering all the info and keeping all of it 'up-to-date' is getting very very expensive and can basically 'neutralize' a lot of your CPU if you have too many resources to watch and keep update. very convenient for users and wouldn't work otherwise. I can see that it might be inappropriate for LVM PVs. We can discuss changing the rules such that the watch is disabled for LVM devices (both PV and LV). It's really not fixable as is - since of the complete lack of 'error' handling of device in udev DB (i.e. duplicate devices, various frozen devices...) There is on going 'SID' project - that might push the logic somewhat further, but existing 'device' support logic as is today is unfortunate 'trace' of how the design should not have been made - and since all 'original' programmers left the project long time ago - it's non-trivial to push things forward. I don't claim to overlook all possible side effects, but it might be worth a try. It would mean that newly created LVs, LV size changes etc. would not be visible in the system immediately. I suppose you could work around that in the LVM tools by triggering change events after operations like lvcreate. We just hope the SID will make some progress (although probably small one at the beginning). However let's focus on 'pvmove' as it is potentially very lengthy operation - so it's not feasible to keep the VG locked/blocked across an operation which might take even days with slower storage and big moved sizes (write access/lock disables all readers...) So these close-after-write operations are caused by locking/unlocking the PVs? Note: We were observing that watch events were triggered every 30s, for every PV, simultaneously. (@Heming correct me if I'mn wrong here). That's why we would like to see 'metadata' and also check if the issue is appearing on the latest version of lvm2. Zdenek ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
On Wed, Aug 17, 2022 at 01:41:17PM +, Martin Wilck wrote: > I like the general idea of the udev watch. It is the magic that causes > newly created partitions to magically appear in the system, which is > very convenient for users and wouldn't work otherwise. I can see that > it might be inappropriate for LVM PVs. We can discuss changing the > rules such that the watch is disabled for LVM devices (both PV and LV). > I don't claim to overlook all possible side effects, but it might be > worth a try. It would mean that newly created LVs, LV size changes etc. > would not be visible in the system immediately. I suppose you could > work around that in the LVM tools by triggering change events after > operations like lvcreate. I think it's worth looking into at least. udev causes most of our major problems, and causes things to fall apart everywhere at scale. > Note: We were observing that watch events were triggered every 30s, for > every PV, simultaneously. Can you see what parts of the VG metadata are changing each time? If it's just metadata related to pvmove progress, then things are probably working as designed, and we'd just be looking for optimizations (perhaps by some code changes, or by reducing copies of metadata as suggested by Zdenek.) Dave ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 14:39 Martin Wilck napsal(a): On Wed, 2022-08-17 at 18:47 +0800, Heming Zhao wrote: On Wed, Aug 17, 2022 at 11:46:16AM +0200, Zdenek Kabelac wrote: ATM I'm not even sure if you are complaining about how CPU usage of lvmpolld or just huge udev rules processing overhead. The load is generated by multipath. lvmpolld does the IN_CLOSE_WRITE action which is the trigger. Let's be clear here: every close-after-write operation triggers udev's "watch" mechanism for block devices, which causes the udev rules to be executed for the device. That is not a cheap operation. In the case at hand, the customer was observing a lot of "multipath -U" commands. So apparently a significant part of the udev rule processing was spent in "multipath -U". Running "multipath -U" is important, because the rule could have been triggered by a change of the number of available paths devices, and later commands run from udev rules might hang indefinitely if the multipath device had no usable paths any more. "multipath -U" is already quite well optimized, but it needs to do some I/O to complete it's work, thus it takes a few milliseconds to run. IOW, it would be misleading to point at multipath. close-after-write operations on block devices should be avoided if possible. As you probably know, the purpose udev's "watch" operation is to be able to determine changes on layered devices, e.g. newly created LVs or the like. "pvmove" is special, because by definition it will usually not cause any changes in higher layers. Therefore it might make sense to disable the udev watch on the affected PVs while pvmove is running, and trigger a single change event (re-enabling the watch) after the pvmove has finished. If that is impossible, lvmpolld and other lvm tools that are involved in the pvmove operation should avoid calling close() on the PVs, IOW keep the fds open until the operation is finished. Hi Let's make clear we are very well aware of all the constrains associated with udev rule logic (and we tried quite hard to minimize impact - however udevd developers kind of 'misunderstood' how badly they will be impacting system's performance with the existing watch rule logic - and the story kind of 'continues' with 'systemd's' & dBus services unfortunatelly... However let's focus on 'pvmove' as it is potentially very lengthy operation - so it's not feasible to keep the VG locked/blocked across an operation which might take even days with slower storage and big moved sizes (write access/lock disables all readers...) So the lvm2 does try to minimize locking time. We will re validate whether just necessary 'vg updating' operation are using 'write' access - since occasionally due to some unrelated code changes it might eventually result sometimes in unwanted 'write' VG open - but we can't keep the operation blocking a whole VG because of slow udev rule processing. In normal circumstances udev rule should be processed very fast - unless there is something mis-designe causing a CPU overloading. But as mentioned already few times - without more knowledge about the case we could hardly guess exact reasoning. But we already provided useful suggestion how to reduce number of 'processed' device by udev by reduction of 'lvm2 metadata PVs' - the big reason for frequent metadata upsate would be a big segmentation of LV - but this we will not know without seeing user's 'metadata' of a VG in this case... Zdenek ___ linux-lvm mailing list linux-lvm@redhat.com https://listman.redhat.com/mailman/listinfo/linux-lvm read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 12:47 Heming Zhao napsal(a):
On Wed, Aug 17, 2022 at 11:46:16AM +0200, Zdenek Kabelac wrote:
Dne 17. 08. 22 v 10:43 Heming Zhao napsal(a):
On Wed, Aug 17, 2022 at 10:06:35AM +0200, Zdenek Kabelac wrote:
Dne 17. 08. 22 v 4:03 Heming Zhao napsal(a):
On Tue, Aug 16, 2022 at 12:26:51PM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 12:08 Heming Zhao napsal(a):
Ooh, very sorry, the subject is wrong, not IO performance but cpu high load
is triggered by pvmove.
The machine connecting disks are more than 250. The VG has 103 PVs & 79 LVs.
# /sbin/vgs
VG #PV #LV #SN Attr VSize VFree
103 79 0 wz--n- 52t17t
Ok - so main issue could be too many PVs with relatively high latency of
mpath devices (which could be all actually simulated easily in lvm2 test suite)
The load is generated by multipath. lvmpolld does the IN_CLOSE_WRITE action
which is the trigger.
I'll check lvmpolld whether it's using correct locking while checking for the
operational state - you may possibly extend checking interval of polling
(although that's where the mentioned patchset has been enhancing couple things)
If you have too many disks in VG (again unclear how many there are paths
and how many distinct PVs) - user may *significantly* reduce burden
associated with metadata updating by reducing number of 'actively'
maintained metadata areas in VG - so i.e. if you have 100PVs in VG - you may
keep metadata only on 5-10 PVs to have 'enough' duplicate copies of lvm2
metadata within VG (vgchange --metadaatacopies X) - clearly it depends on
the use case and how many PVs are added/removed from a VG over the
lifetime
Thanks for the important info. I also found the related VG config from
/etc/lvm/backup/, this file shows 'metadata_copies = 0'.
This should be another solution. But why not lvm2 takes this behavior by
default, or give a notification when pv number beyond a threshold when user
executing pvs/vgs/lvs or pvmove.
There are too many magic switch, users don't know how to adjust them for
better performance.
Problem is always the same - selecting right 'default' :) what suites to user
A is sometimes 'no go' for user B. So ATM it's more 'secure/safe' to keep
metadata with each PV - so when a PV is discovered it's known how the VG using
such PV looks like. When only fraction of PV have the info - VG is way more
fragile on damage when disks are lost i.e. there is no 'smart' mechanism to
pick disks in different racks
So this option is there for administrators that are 'clever' enough to deal
with a new set of problems it may create for them.
Yes - lvm2 has lot of options - but that's what is usually necessary when we
want to be capable to provide optimal solution for really wide variety of
setups - so I think spending couple minutes on reading man pages pays off -
especially if you had to spend 'days' on build your disk racks ;)
And yes we may add few more hints - but then we are asked by 'second' group of
users ('skilled admins') - why do we print so many dumb messages every time
they do some simple operation :)
I'm busy with many bugs, still can't find a time slot to set up a env.
For this performance issue, it relates with mpath, I can't find a easy
way to set up a env. (I suspect it may trigger this issue by setting up
300 fake PVs without mpath, then do pvmove cmd.)
'Fragmented' LVs with small segment sizes my significantly raise the amount of
metadata updates needed during pvmove operation as each single LV segments
will be mirrored by individual mirror.
Zdenek
___
linux-lvm mailing list
linux-lvm@redhat.com
https://listman.redhat.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 10:43 Heming Zhao napsal(a):
On Wed, Aug 17, 2022 at 10:06:35AM +0200, Zdenek Kabelac wrote:
Dne 17. 08. 22 v 4:03 Heming Zhao napsal(a):
On Tue, Aug 16, 2022 at 12:26:51PM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 12:08 Heming Zhao napsal(a):
Ooh, very sorry, the subject is wrong, not IO performance but cpu high load
is triggered by pvmove.
On Tue, Aug 16, 2022 at 11:38:52AM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 11:28 Heming Zhao napsal(a):
Hello maintainers & list,
I bring a story:
One SUSE customer suffered lvmpolld issue, which cause IO performance dramatic
decrease.
How to trigger:
When machine connects large number of LUNs (eg 80~200), pvmove (eg, move a
single
disk to a new one, cmd like: pvmove disk1 disk2), the system will suffer high
cpu load. But when system connects ~10 LUNs, the performance is fine.
We found two work arounds:
1. set lvm.conf 'activation/polling_interval=120'.
2. write a speical udev rule, which make udev ignore the event for mpath
devices.
echo 'ENV{DM_UUID}=="mpath-*", OPTIONS+="nowatch"' >\
/etc/udev/rules.d/90-dm-watch.rules
Run above any one of two can make the performance issue disappear.
** the root cause **
lvmpolld will do interval requeset info job for updating the pvmove status
On every polling_interval time, lvm2 will update vg metadata. The update job
will
call sys_close, which will trigger systemd-udevd IN_CLOSE_WRITE event, eg:
2022--xxx systemd-udevd[pid]: dm-179: Inotify event: 8
for /dev/dm-179
(8 is IN_CLOSE_WRITE.)
These VGs underlying devices are multipath devices. So when lvm2 update
metatdata,
even if pvmove write a few data, the sys_close action trigger udev's "watch"
mechanism to gets notified frequently about a process that has written to the
device and closed it. This causes frequent, pointless re-evaluation of the udev
rules for these devices.
My question: Does LVM2 maintainers have any idea to fix this bug?
In my view, does lvm2 could drop VGs devices fds until pvmove finish?
Hi
Please provide more info about lvm2 metadata and also some 'lvs -av'
trace so we can get better picture about the layout - also version of
lvm2,systemd,kernel in use.
pvmove is progressing by mirroring each segment of an LV - so if there would
be a lot of segments - then each such update may trigger udev watch rule
event.
But ATM I could hardly imagine how this could cause some 'dramatic'
performance decrease - maybe there is something wrong with udev rules on
the system ?
What is the actual impact ?
Note - pvmove was never designed as a high performance operation (in fact it
tries to not eat all the disk bandwidth as such)
Regards
Zdenek
My mistake, I write here again:
The subject is wrong, not IO performance but cpu high load is triggered by
pvmove.
There is no IO performance issue.
When system is connecting 80~200, the cpu load increase by 15~20, the
cpu usage by ~20%, which corresponds to about ~5,6 cores and led at
times to the cores fully utilized.
In another word: a single pvmove process cost 5-6 (sometime 10) cores
utilization. It's abnormal & unaccepted.
The lvm2 is 2.03.05, kernel is 5.3. systemd is v246.
BTW:
I change this mail subject from: lvmpolld causes IO performance issue
to: lvmpolld causes high cpu load issue
Please use this mail for later discussing.
Hi
Could you please retest with recent version of lvm2. There have been
certainly some improvements in scanning - which might have caused in the
older releases some higher CPU usage with longer set of devices.
Regards
Zdenek
The highest lvm2 version in SUSE products is lvm2-2.03.15, does this
version include the improvements change?
Could you mind to point out which commits related with the improvements?
I don't have the reproducible env, I need to get a little detail before
asking customer to try new version.
Please try to reproduce your customer's problem and see if the newer version
solves the issue. Otherwise we could waste hours on theoretical
discussions what might or might not have helped with this problem. Having a
reproducer is a starting point for fixing it, if the problem is still there.
Here is one commit that may possibly affect CPU load:
d2522f4a05aa027bcc911ecb832450bc19b7fb57
Regards
Zdenek
I gave a little bit explain for the root cause in previous mail, And the
work around <2> also matchs my analysis.
The machine connects lots of LUNs. pvmove one disk will trigger lvm2
update all underlying mpath devices (80~200). I guess the update job is
vg_commit() which updates latest metadata info, and the metadata locates in
all PVs. The update job finished with close(2) which trigger hundreds
devices udevd IN_CLOSE_WRITE event. every IN_CLOSE_WRITE will trigger
mpathd udev rules (11-dm-mpath.rules) to start scanning devices. So the
real world will flooding hundreds of multipath processes, the cpus load
become high.
Your 'guess explanation' is not as useful as you might think - as
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 17. 08. 22 v 4:03 Heming Zhao napsal(a):
On Tue, Aug 16, 2022 at 12:26:51PM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 12:08 Heming Zhao napsal(a):
Ooh, very sorry, the subject is wrong, not IO performance but cpu high load
is triggered by pvmove.
On Tue, Aug 16, 2022 at 11:38:52AM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 11:28 Heming Zhao napsal(a):
Hello maintainers & list,
I bring a story:
One SUSE customer suffered lvmpolld issue, which cause IO performance dramatic
decrease.
How to trigger:
When machine connects large number of LUNs (eg 80~200), pvmove (eg, move a
single
disk to a new one, cmd like: pvmove disk1 disk2), the system will suffer high
cpu load. But when system connects ~10 LUNs, the performance is fine.
We found two work arounds:
1. set lvm.conf 'activation/polling_interval=120'.
2. write a speical udev rule, which make udev ignore the event for mpath
devices.
echo 'ENV{DM_UUID}=="mpath-*", OPTIONS+="nowatch"' >\
/etc/udev/rules.d/90-dm-watch.rules
Run above any one of two can make the performance issue disappear.
** the root cause **
lvmpolld will do interval requeset info job for updating the pvmove status
On every polling_interval time, lvm2 will update vg metadata. The update job
will
call sys_close, which will trigger systemd-udevd IN_CLOSE_WRITE event, eg:
2022--xxx systemd-udevd[pid]: dm-179: Inotify event: 8
for /dev/dm-179
(8 is IN_CLOSE_WRITE.)
These VGs underlying devices are multipath devices. So when lvm2 update
metatdata,
even if pvmove write a few data, the sys_close action trigger udev's "watch"
mechanism to gets notified frequently about a process that has written to the
device and closed it. This causes frequent, pointless re-evaluation of the udev
rules for these devices.
My question: Does LVM2 maintainers have any idea to fix this bug?
In my view, does lvm2 could drop VGs devices fds until pvmove finish?
Hi
Please provide more info about lvm2 metadata and also some 'lvs -av'
trace so we can get better picture about the layout - also version of
lvm2,systemd,kernel in use.
pvmove is progressing by mirroring each segment of an LV - so if there would
be a lot of segments - then each such update may trigger udev watch rule
event.
But ATM I could hardly imagine how this could cause some 'dramatic'
performance decrease - maybe there is something wrong with udev rules on
the system ?
What is the actual impact ?
Note - pvmove was never designed as a high performance operation (in fact it
tries to not eat all the disk bandwidth as such)
Regards
Zdenek
My mistake, I write here again:
The subject is wrong, not IO performance but cpu high load is triggered by
pvmove.
There is no IO performance issue.
When system is connecting 80~200, the cpu load increase by 15~20, the
cpu usage by ~20%, which corresponds to about ~5,6 cores and led at
times to the cores fully utilized.
In another word: a single pvmove process cost 5-6 (sometime 10) cores
utilization. It's abnormal & unaccepted.
The lvm2 is 2.03.05, kernel is 5.3. systemd is v246.
BTW:
I change this mail subject from: lvmpolld causes IO performance issue
to: lvmpolld causes high cpu load issue
Please use this mail for later discussing.
Hi
Could you please retest with recent version of lvm2. There have been
certainly some improvements in scanning - which might have caused in the
older releases some higher CPU usage with longer set of devices.
Regards
Zdenek
The highest lvm2 version in SUSE products is lvm2-2.03.15, does this
version include the improvements change?
Could you mind to point out which commits related with the improvements?
I don't have the reproducible env, I need to get a little detail before
asking customer to try new version.
Please try to reproduce your customer's problem and see if the newer version
solves the issue. Otherwise we could waste hours on theoretical discussions
what might or might not have helped with this problem. Having a reproducer is
a starting point for fixing it, if the problem is still there.
Here is one commit that may possibly affect CPU load:
d2522f4a05aa027bcc911ecb832450bc19b7fb57
Regards
Zdenek
___
linux-lvm mailing list
linux-lvm@redhat.com
https://listman.redhat.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
Re: [linux-lvm] lvmpolld causes high cpu load issue
Dne 16. 08. 22 v 12:08 Heming Zhao napsal(a):
Ooh, very sorry, the subject is wrong, not IO performance but cpu high load
is triggered by pvmove.
On Tue, Aug 16, 2022 at 11:38:52AM +0200, Zdenek Kabelac wrote:
Dne 16. 08. 22 v 11:28 Heming Zhao napsal(a):
Hello maintainers & list,
I bring a story:
One SUSE customer suffered lvmpolld issue, which cause IO performance dramatic
decrease.
How to trigger:
When machine connects large number of LUNs (eg 80~200), pvmove (eg, move a
single
disk to a new one, cmd like: pvmove disk1 disk2), the system will suffer high
cpu load. But when system connects ~10 LUNs, the performance is fine.
We found two work arounds:
1. set lvm.conf 'activation/polling_interval=120'.
2. write a speical udev rule, which make udev ignore the event for mpath
devices.
echo 'ENV{DM_UUID}=="mpath-*", OPTIONS+="nowatch"' >\
/etc/udev/rules.d/90-dm-watch.rules
Run above any one of two can make the performance issue disappear.
** the root cause **
lvmpolld will do interval requeset info job for updating the pvmove status
On every polling_interval time, lvm2 will update vg metadata. The update job
will
call sys_close, which will trigger systemd-udevd IN_CLOSE_WRITE event, eg:
2022--xxx systemd-udevd[pid]: dm-179: Inotify event: 8 for
/dev/dm-179
(8 is IN_CLOSE_WRITE.)
These VGs underlying devices are multipath devices. So when lvm2 update
metatdata,
even if pvmove write a few data, the sys_close action trigger udev's "watch"
mechanism to gets notified frequently about a process that has written to the
device and closed it. This causes frequent, pointless re-evaluation of the udev
rules for these devices.
My question: Does LVM2 maintainers have any idea to fix this bug?
In my view, does lvm2 could drop VGs devices fds until pvmove finish?
Hi
Please provide more info about lvm2 metadata and also some 'lvs -av'
trace so we can get better picture about the layout - also version of
lvm2,systemd,kernel in use.
pvmove is progressing by mirroring each segment of an LV - so if there would
be a lot of segments - then each such update may trigger udev watch rule
event.
But ATM I could hardly imagine how this could cause some 'dramatic'
performance decrease - maybe there is something wrong with udev rules on
the system ?
What is the actual impact ?
Note - pvmove was never designed as a high performance operation (in fact it
tries to not eat all the disk bandwidth as such)
Regards
Zdenek
My mistake, I write here again:
The subject is wrong, not IO performance but cpu high load is triggered by
pvmove.
There is no IO performance issue.
When system is connecting 80~200, the cpu load increase by 15~20, the
cpu usage by ~20%, which corresponds to about ~5,6 cores and led at
times to the cores fully utilized.
In another word: a single pvmove process cost 5-6 (sometime 10) cores
utilization. It's abnormal & unaccepted.
The lvm2 is 2.03.05, kernel is 5.3. systemd is v246.
BTW:
I change this mail subject from: lvmpolld causes IO performance issue
to: lvmpolld causes high cpu load issue
Please use this mail for later discussing.
Hi
Could you please retest with recent version of lvm2. There have been certainly
some improvements in scanning - which might have caused in the older releases
some higher CPU usage with longer set of devices.
Regards
Zdenek
___
linux-lvm mailing list
linux-lvm@redhat.com
https://listman.redhat.com/mailman/listinfo/linux-lvm
read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
