d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add separate aggregation limit for non-rotating media.

Browse Source 
Before sequential scrub patches ZFS never aggregated I/Os above 128KB.
Sequential scrub bumped that to 1MB, which motivation I understand for
spinning disks, since it should reduce number of head seeks.  But for
SSDs it makes much less sense to me, especially on FreeBSD, where due
to MAXPHYS limitation device will likely still see bunch of 128KB I/Os
instead of one large.  Having more strict aggregation limit allows to
avoid allocation of large memory buffer and memcpy to/from it, that is
a serious problem when bandwidth reaches few GB/s.

MFC after:	1 month
Sponsored by:	iXsystems, Inc.
This commit is contained in:
Alexander Motin 2019-03-08 19:38:52 +00:00
parent 849bfb6d3d
commit 673544c3dd
Notes: svn2git 2020-12-20 02:59:44 +00:00 
svn path=/head/; revision=344934
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_queue.c
View File

@ -178,6 +178,7 @@ int zfs_vdev_async_write_active_max_dirty_percent = 60;
* they aren't able to help us aggregate at this level.
*/
int zfs_vdev_aggregation_limit = 1 << 20;
int zfs_vdev_aggregation_limit_non_rotating = SPA_OLD_MAXBLOCKSIZE;
int zfs_vdev_read_gap_limit = 32 << 10;
int zfs_vdev_write_gap_limit = 4 << 10;
@ -262,6 +263,9 @@ ZFS_VDEV_QUEUE_KNOB_MAX(initializing);
SYSCTL_INT(_vfs_zfs_vdev, OID_AUTO, aggregation_limit, CTLFLAG_RWTUN,
&zfs_vdev_aggregation_limit, 0,
"I/O requests are aggregated up to this size");
SYSCTL_INT(_vfs_zfs_vdev, OID_AUTO, aggregation_limit_non_rotating, CTLFLAG_RWTUN,
&zfs_vdev_aggregation_limit_non_rotating, 0,
"I/O requests are aggregated up to this size for non-rotating media");
SYSCTL_INT(_vfs_zfs_vdev, OID_AUTO, read_gap_limit, CTLFLAG_RWTUN,
&zfs_vdev_read_gap_limit, 0,
"Acceptable gap between two reads being aggregated");
@ -682,9 +686,13 @@ vdev_queue_aggregate(vdev_queue_t *vq, zio_t *zio)
ASSERT(MUTEX_HELD(&vq->vq_lock));
maxblocksize = spa_maxblocksize(vq->vq_vdev->vdev_spa);
limit = MAX(MIN(zfs_vdev_aggregation_limit, maxblocksize), 0);
if (vq->vq_vdev->vdev_rotation_rate == VDEV_RATE_NON_ROTATING)
limit = zfs_vdev_aggregation_limit_non_rotating;
else
limit = zfs_vdev_aggregation_limit;
limit = MAX(MIN(limit, maxblocksize), 0);
if (zio->io_flags & ZIO_FLAG_DONT_AGGREGATE || limit == 0)
if (zio->io_flags & ZIO_FLAG_DONT_AGGREGATE || zio->io_size >= limit)
return (NULL);
first = last = zio;