zvol_write() can use dmu_tx_hold_write_by_dnode() #10184
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
We can improve the performance of writes to zvols by using dmu_tx_hold_write_by_dnode() instead of dmu_tx_hold_write(). This reduces lock contention on the first block of the dnode object, and also reduces the amount of CPU needed. The benefit will be highest with multi-threaded async writes (i.e. writes that don't call zil_commit()). Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
ahrens
added
Component: ZVOL
behlendorf
approved these changes
Apr 9, 2020
lundman
approved these changes
Apr 10, 2020
There was a problem hiding this comment.
I have tested this over at openzfsonwindows/ZFSin#240 where it does appear to make a difference.
behlendorf
added
Status: Accepted
tonynguien
approved these changes
Apr 10, 2020
12 tasks
tle211212
added a commit
to tle211212/zfs
that referenced
this pull request
May 19, 2020
Closed
jsai20
pushed a commit
to jsai20/zfs
that referenced
this pull request
Mar 30, 2021
We can improve the performance of writes to zvols by using dmu_tx_hold_write_by_dnode() instead of dmu_tx_hold_write(). This reduces lock contention on the first block of the dnode object, and also reduces the amount of CPU needed. The benefit will be highest with multi-threaded async writes (i.e. writes that don't call zil_commit()). Reviewed-by: Jorgen Lundman <lundman@lundman.net> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com> Signed-off-by: Matthew Ahrens <mahrens@delphix.com> Closes openzfs#10184
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation and Context
We can improve the performance of writes to zvols by using
dmu_tx_hold_write_by_dnode() instead of dmu_tx_hold_write(). This
reduces lock contention on the first block of the dnode object, and also
reduces the amount of CPU needed. The benefit will be highest with
multi-threaded async writes (i.e. writes that don't call zil_commit()).
See https://github.com/openzfsonwindows/ZFSin/issues/238
Description
Make zvol_write() call dmu_tx_hold_write_by_dnode() instead of dmu_tx_hold_write().
How Has This Been Tested?
Minimal testing so far. Would like to measure performance improvement with async writes.
Types of changes
Checklist:
Signed-off-by.