Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Improve scrub maxinflight_bytes math. #12046

Merged
merged 1 commit into from
May 27, 2021
Merged

Improve scrub maxinflight_bytes math. #12046

merged 1 commit into from
May 27, 2021

Conversation

amotin
Copy link
Member

@amotin amotin commented May 14, 2021

Previous code scaled maxinflight_bytes based on total number of disks
in the pool. It means 3-wide mirror was receiving queue depth of 3
disks, which it should not, since it reads from all the disks inside.
For wide raidz the situation was slightly better, but still 3-wide
raidz1 received depth of 3 disks instead of 2.

The new code counts only unique data disks, i.e. 1 disk for mirrors
and non-parity disks for raidz/draid. For draid the math is still
imperfect, since vdev_get_nparity() returns number of parity disks
per group, not per vdev, but still some better than it was.

This should slightly reduce scrub influence on payload for some pool
topologies by avoiding excessive queuing.

How Has This Been Tested?

I've manually verified return value of dsl_scan_count_data_disks() for some mirrors, raidzs and draids for matching expectations. Run scrubs without surprises.

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Performance enhancement (non-breaking change which improves efficiency)
  • Code cleanup (non-breaking change which makes code smaller or more readable)
  • Breaking change (fix or feature that would cause existing functionality to change)
  • Library ABI change (libzfs, libzfs_core, libnvpair, libuutil and libzfsbootenv)
  • Documentation (a change to man pages or other documentation)

Checklist:

@amotin
Improve scrub maxinflight_bytes math.
a61012d 
Previous code scaled maxinflight_bytes based on total number of disks
in the pool.  It means 3-wide mirror was receiving queue depth of 3
disks, which it should not, since it reads from all the disks inside.
For wide raidz the situation was slightly better, but still 3-wide
raidz1 received depth of 3 disks instead of 2.

The new code counts only unique data disks, i.e. 1 disk for mirrors
and non-parity disks for raidz/draid.  For draid the math is still
imperfect, since vdev_get_nparity() returns number of parity disks
per group, not per vdev, but still some better than it was.

This should slightly reduce scrub influence on payload for some pool
topologies by avoiding excessive queuing.

Signed-off-by:	Alexander Motin <mav@FreeBSD.org>
Sponsored-By:	iXsystems, Inc.
@amotin amotin requested a review from behlendorf May 14, 2021 21:30
@amotin amotin self-assigned this May 14, 2021
@amotin amotin added the Status: Code Review Needed Ready for review and testing label May 14, 2021
@ahrens ahrens assigned mmaybee and unassigned amotin May 20, 2021
@amotin amotin requested a review from a user May 21, 2021 18:56
behlendorf
behlendorf approved these changes May 26, 2021
View reviewed changes
Copy link
Contributor

@behlendorf behlendorf left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good and makes sense. I hadn't really considered the N-way mirror case which effectively inflates the zfs_scan_vdev_limit. As for dRAID it doesn't quite fit in the model here, but you're right vdev_get_ndisks(vd) - vdev_get_nparity(vd) is a perfectly good approximation.

@mmaybee mmaybee added Status: Accepted Ready to integrate (reviewed, tested) and removed Status: Code Review Needed Ready for review and testing labels May 27, 2021
@mmaybee mmaybee merged commit 2041d6e into openzfs:master May 27, 2021
behlendorf pushed a commit to behlendorf/zfs that referenced this pull request Jun 8, 2021
@amotin @behlendorf
Improve scrub maxinflight_bytes math.
40a03c8 
Previously, ZFS scaled maxinflight_bytes based on total number of
disks in the pool.  A 3-wide mirror was receiving a queue depth of 3
disks, which it should not, since it reads from all the disks inside.
For wide raidz the situation was slightly better, but still a 3-wide
raidz1 received a depth of 3 disks instead of 2.

The new code counts only unique data disks, i.e. 1 disk for mirrors
and non-parity disks for raidz/draid.  For draid the math is still
imperfect, since vdev_get_nparity() returns number of parity disks
per group, not per vdev, but still some better than it was.

This should slightly reduce scrub influence on payload for some pool
topologies by avoiding excessive queuing.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by:	Alexander Motin <mav@FreeBSD.org>
Sponsored-By:	iXsystems, Inc.
Closing openzfs#12046
behlendorf pushed a commit to behlendorf/zfs that referenced this pull request Jun 8, 2021
@amotin @behlendorf
Improve scrub maxinflight_bytes math.
09c0a8f 
Previously, ZFS scaled maxinflight_bytes based on total number of
disks in the pool.  A 3-wide mirror was receiving a queue depth of 3
disks, which it should not, since it reads from all the disks inside.
For wide raidz the situation was slightly better, but still a 3-wide
raidz1 received a depth of 3 disks instead of 2.

The new code counts only unique data disks, i.e. 1 disk for mirrors
and non-parity disks for raidz/draid.  For draid the math is still
imperfect, since vdev_get_nparity() returns number of parity disks
per group, not per vdev, but still some better than it was.

This should slightly reduce scrub influence on payload for some pool
topologies by avoiding excessive queuing.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by:	Alexander Motin <mav@FreeBSD.org>
Sponsored-By:	iXsystems, Inc.
Closing openzfs#12046
@amotin amotin deleted the scrub branch August 24, 2021 20:18
amotin added a commit to amotin/zfs that referenced this pull request Apr 29, 2022
@amotin
Improve mg_aliquot math.
e60c77d 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 1.5MB.
After this change in all the cases each disk should get 1MB.

Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
@amotin amotin mentioned this pull request Apr 29, 2022
Merged
13 tasks
amotin added a commit to amotin/zfs that referenced this pull request Apr 29, 2022
@amotin
Improve mg_aliquot math.
7a2c291 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
behlendorf pushed a commit that referenced this pull request May 4, 2022
@amotin
Improve mg_aliquot math
c55b293 
When calculating mg_aliquot alike to #12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes #13388
amotin added a commit to amotin/zfs that referenced this pull request Jul 20, 2022
@amotin
Improve mg_aliquot math
c1c90b6 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
behlendorf pushed a commit that referenced this pull request Jul 26, 2022
@amotin @behlendorf
Improve mg_aliquot math
6e1e90d 
When calculating mg_aliquot alike to #12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes #13388
nicman23 pushed a commit to nicman23/zfs that referenced this pull request Aug 22, 2022
@amotin @nicman23
Improve mg_aliquot math
bc3d87d 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
nicman23 pushed a commit to nicman23/zfs that referenced this pull request Aug 22, 2022
@amotin @nicman23
Improve mg_aliquot math
eb00326 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
lundman pushed a commit to openzfsonwindows/openzfs that referenced this pull request Sep 6, 2022
@amotin @lundman
Improve mg_aliquot mah
ce94cca 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
beren12 pushed a commit to beren12/zfs that referenced this pull request Sep 19, 2022
@amotin @beren12
Improve mg_aliquot math
b6c16f9 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
9419563 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
c74a7f3 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
4b6c457 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
51eae90 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
52b289f 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
a1a20b3 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
3675865 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
0a5045b 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
1b04fab 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
065f3f5 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
429869d 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
0203ce1 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
28d74d3 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
f1330a2 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
1f97269 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
3c4a94c 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
941fa04 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
andrewc12 pushed a commit to andrewc12/openzfs that referenced this pull request Sep 23, 2022
@amotin @andrewc12
Improve mg_aliquot mah
60f19c2 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
snajpa pushed a commit to vpsfreecz/zfs that referenced this pull request Oct 22, 2022
@amotin @snajpa
Improve mg_aliquot math
5a2a619 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
snajpa pushed a commit to vpsfreecz/zfs that referenced this pull request Oct 22, 2022
@amotin @snajpa
Improve mg_aliquot math
63a09f7 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
snajpa pushed a commit to vpsfreecz/zfs that referenced this pull request Oct 23, 2022
@amotin @snajpa
Improve mg_aliquot math
271e900 
When calculating mg_aliquot alike to openzfs#12046 use number of unique data
disks in the vdev, not the total number of children vdev.  Increase
default value of the tunable from 512KB to 1MB to compensate.

Before this change each disk in striped pool was getting 512KB of
sequential data, in 2-wide mirror -- 1MB, in 3-wide RAIDZ1 -- 768KB.
After this change in all the cases each disk should get 1MB.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13388
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@behlendorf behlendorf behlendorf approved these changes

Assignees

@mmaybee mmaybee

Labels
Status: Accepted Ready to integrate (reviewed, tested)
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@amotin @behlendorf @mmaybee