Skip to content

Commit

Permalink
Reduce ZIO io_lock contention on sorted scrub
Browse files Browse the repository at this point in the history
During sorted scrub multiple threads (one per vdev) are issuing many
ZIOs same time, all using the same scn->scn_zio_root ZIO as parent.
It causes huge lock contention on the single global lock on that ZIO.
Improve it by introducing per-queue null ZIOs, children to that one,
and using them instead as proxy.

For 12 SSD pool storing 1.5TB of 4KB blocks on 80-core system this
dramatically reduces lock contention and reduces scrub time from 21
minutes down to 12.5, while actual read stages (not scan) are about
3x faster, reaching 100K blocks per second per vdev.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Sponsored-By: iXsystems, Inc.
Closes openzfs#13553
  • Loading branch information
amotin authored and andrewc12 committed Sep 23, 2022
1 parent e779a3b commit 48c56c0
Showing 1 changed file with 15 additions and 4 deletions.
19 changes: 15 additions & 4 deletions module/zfs/dsl_scan.c
Original file line number Diff line number Diff line change
Expand Up @@ -280,6 +280,7 @@ typedef struct scan_io {
struct dsl_scan_io_queue {
dsl_scan_t *q_scn; /* associated dsl_scan_t */
vdev_t *q_vd; /* top-level vdev that this queue represents */
zio_t *q_zio; /* scn_zio_root child for waiting on IO */

/* trees used for sorting I/Os and extents of I/Os */
range_tree_t *q_exts_by_addr;
Expand Down Expand Up @@ -3036,15 +3037,19 @@ scan_io_queues_run_one(void *arg)
dsl_scan_io_queue_t *queue = arg;
kmutex_t *q_lock = &queue->q_vd->vdev_scan_io_queue_lock;
boolean_t suspended = B_FALSE;
range_seg_t *rs = NULL;
scan_io_t *sio = NULL;
range_seg_t *rs;
scan_io_t *sio;
zio_t *zio;
list_t sio_list;

ASSERT(queue->q_scn->scn_is_sorted);

list_create(&sio_list, sizeof (scan_io_t),
offsetof(scan_io_t, sio_nodes.sio_list_node));
zio = zio_null(queue->q_scn->scn_zio_root, queue->q_scn->scn_dp->dp_spa,
NULL, NULL, NULL, ZIO_FLAG_CANFAIL);
mutex_enter(q_lock);
queue->q_zio = zio;

/* Calculate maximum in-flight bytes for this vdev. */
queue->q_maxinflight_bytes = MAX(1, zfs_scan_vdev_limit *
Expand Down Expand Up @@ -3111,7 +3116,9 @@ scan_io_queues_run_one(void *arg)
scan_io_queue_insert_impl(queue, sio);
}

queue->q_zio = NULL;
mutex_exit(q_lock);
zio_nowait(zio);
list_destroy(&sio_list);
}

Expand Down Expand Up @@ -4076,6 +4083,7 @@ scan_exec_io(dsl_pool_t *dp, const blkptr_t *bp, int zio_flags,
dsl_scan_t *scn = dp->dp_scan;
size_t size = BP_GET_PSIZE(bp);
abd_t *data = abd_alloc_for_io(size, B_FALSE);
zio_t *pio;

if (queue == NULL) {
ASSERT3U(scn->scn_maxinflight_bytes, >, 0);
Expand All @@ -4084,6 +4092,7 @@ scan_exec_io(dsl_pool_t *dp, const blkptr_t *bp, int zio_flags,
cv_wait(&spa->spa_scrub_io_cv, &spa->spa_scrub_lock);
spa->spa_scrub_inflight += BP_GET_PSIZE(bp);
mutex_exit(&spa->spa_scrub_lock);
pio = scn->scn_zio_root;
} else {
kmutex_t *q_lock = &queue->q_vd->vdev_scan_io_queue_lock;

Expand All @@ -4092,12 +4101,14 @@ scan_exec_io(dsl_pool_t *dp, const blkptr_t *bp, int zio_flags,
while (queue->q_inflight_bytes >= queue->q_maxinflight_bytes)
cv_wait(&queue->q_zio_cv, q_lock);
queue->q_inflight_bytes += BP_GET_PSIZE(bp);
pio = queue->q_zio;
mutex_exit(q_lock);
}

ASSERT(pio != NULL);
count_block(scn, dp->dp_blkstats, bp);
zio_nowait(zio_read(scn->scn_zio_root, spa, bp, data, size,
dsl_scan_scrub_done, queue, ZIO_PRIORITY_SCRUB, zio_flags, zb));
zio_nowait(zio_read(pio, spa, bp, data, size, dsl_scan_scrub_done,
queue, ZIO_PRIORITY_SCRUB, zio_flags, zb));
}

/*
Expand Down

0 comments on commit 48c56c0

Please sign in to comment.