diff mbox series

[13/36] xfs_scrub: use datadev parallelization estimates for thread count

Message ID 155259750953.31886.7244219282258067422.stgit@magnolia (mailing list archive)
State Accepted, archived
Headers show
Series xfsprogs-5.0: fix various problems | expand

Commit Message

Darrick J. Wong March 14, 2019, 9:05 p.m. UTC
From: Darrick J. Wong <darrick.wong@oracle.com>

During phases 2-5, xfs_scrub should estimate the level of
parallelization possible on the data device to determine the number of
threads spawned to scrub filesystem metadata, not just blindly using the
number of CPUs.  This avoids flooding non-rotational storage with random
reads, which totally destroys performance and makes scrub runtimes
higher.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 scrub/phase1.c |   14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

Comments

Eric Sandeen March 15, 2019, 5:36 p.m. UTC | #1
On 3/14/19 4:05 PM, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> During phases 2-5, xfs_scrub should estimate the level of
> parallelization possible on the data device to determine the number of
> threads spawned to scrub filesystem metadata, not just blindly using the
> number of CPUs.  This avoids flooding non-rotational storage with random
> reads, which totally destroys performance and makes scrub runtimes
> higher.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>

seems like when I first looked at this I had a concern but now I don't
remember, and today it seems fine? :)

Reviewed-by: Eric Sandeen <sandeen@redhat.com>

> ---
>  scrub/phase1.c |   14 +++++++-------
>  1 file changed, 7 insertions(+), 7 deletions(-)
> 
> 
> diff --git a/scrub/phase1.c b/scrub/phase1.c
> index 2113014b..6b472147 100644
> --- a/scrub/phase1.c
> +++ b/scrub/phase1.c
> @@ -109,13 +109,6 @@ _("Must be root to run scrub."));
>  		return false;
>  	}
>  
> -	ctx->nr_io_threads = nproc;
> -	if (verbose) {
> -		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
> -				ctx->mntpoint, scrub_nproc(ctx));
> -		fflush(stdout);
> -	}
> -
>  	if (!platform_test_xfs_fd(ctx->mnt_fd)) {
>  		str_info(ctx, ctx->mntpoint,
>  _("Does not appear to be an XFS filesystem!"));
> @@ -193,6 +186,13 @@ _("Unable to find realtime device path."));
>  		return false;
>  	}
>  
> +	ctx->nr_io_threads = disk_heads(ctx->datadev);
> +	if (verbose) {
> +		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
> +				ctx->mntpoint, scrub_nproc(ctx));
> +		fflush(stdout);
> +	}
> +
>  	if (ctx->fsinfo.fs_log) {
>  		ctx->logdev = disk_open(ctx->fsinfo.fs_log);
>  		if (error) {
>
Darrick J. Wong March 15, 2019, 5:41 p.m. UTC | #2
On Fri, Mar 15, 2019 at 12:36:34PM -0500, Eric Sandeen wrote:
> On 3/14/19 4:05 PM, Darrick J. Wong wrote:
> > From: Darrick J. Wong <darrick.wong@oracle.com>
> > 
> > During phases 2-5, xfs_scrub should estimate the level of
> > parallelization possible on the data device to determine the number of
> > threads spawned to scrub filesystem metadata, not just blindly using the
> > number of CPUs.  This avoids flooding non-rotational storage with random

s/non-//g in the line above, is what you complained about last time and
then I didn't bother to fix for subsequent repostings. :(

--D

> > reads, which totally destroys performance and makes scrub runtimes
> > higher.
> > 
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> 
> seems like when I first looked at this I had a concern but now I don't
> remember, and today it seems fine? :)
> 
> Reviewed-by: Eric Sandeen <sandeen@redhat.com>
> 
> > ---
> >  scrub/phase1.c |   14 +++++++-------
> >  1 file changed, 7 insertions(+), 7 deletions(-)
> > 
> > 
> > diff --git a/scrub/phase1.c b/scrub/phase1.c
> > index 2113014b..6b472147 100644
> > --- a/scrub/phase1.c
> > +++ b/scrub/phase1.c
> > @@ -109,13 +109,6 @@ _("Must be root to run scrub."));
> >  		return false;
> >  	}
> >  
> > -	ctx->nr_io_threads = nproc;
> > -	if (verbose) {
> > -		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
> > -				ctx->mntpoint, scrub_nproc(ctx));
> > -		fflush(stdout);
> > -	}
> > -
> >  	if (!platform_test_xfs_fd(ctx->mnt_fd)) {
> >  		str_info(ctx, ctx->mntpoint,
> >  _("Does not appear to be an XFS filesystem!"));
> > @@ -193,6 +186,13 @@ _("Unable to find realtime device path."));
> >  		return false;
> >  	}
> >  
> > +	ctx->nr_io_threads = disk_heads(ctx->datadev);
> > +	if (verbose) {
> > +		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
> > +				ctx->mntpoint, scrub_nproc(ctx));
> > +		fflush(stdout);
> > +	}
> > +
> >  	if (ctx->fsinfo.fs_log) {
> >  		ctx->logdev = disk_open(ctx->fsinfo.fs_log);
> >  		if (error) {
> >
diff mbox series

Patch

diff --git a/scrub/phase1.c b/scrub/phase1.c
index 2113014b..6b472147 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -109,13 +109,6 @@  _("Must be root to run scrub."));
 		return false;
 	}
 
-	ctx->nr_io_threads = nproc;
-	if (verbose) {
-		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
-				ctx->mntpoint, scrub_nproc(ctx));
-		fflush(stdout);
-	}
-
 	if (!platform_test_xfs_fd(ctx->mnt_fd)) {
 		str_info(ctx, ctx->mntpoint,
 _("Does not appear to be an XFS filesystem!"));
@@ -193,6 +186,13 @@  _("Unable to find realtime device path."));
 		return false;
 	}
 
+	ctx->nr_io_threads = disk_heads(ctx->datadev);
+	if (verbose) {
+		fprintf(stdout, _("%s: using %d threads to scrub.\n"),
+				ctx->mntpoint, scrub_nproc(ctx));
+		fflush(stdout);
+	}
+
 	if (ctx->fsinfo.fs_log) {
 		ctx->logdev = disk_open(ctx->fsinfo.fs_log);
 		if (error) {