Sequential scrub and resilvers

Currently, scrubs and resilvers can take an extremely
long time to complete. This is largely due to the fact
that zfs scans process pools in logical order, as
determined by each block's bookmark. This makes sense
from a simplicity perspective, but blocks in zfs are
often scattered randomly across disks, particularly
due to zfs's copy-on-write mechanisms.

This patch improves performance by splitting scrubs
and resilvers into a metadata scanning phase and an IO
issuing phase. The metadata scan reads through the
structure of the pool and gathers an in-memory queue
of I/Os, sorted by size and offset on disk. The issuing
phase will then issue the scrub I/Os as sequentially as
possible, greatly improving performance.

This patch also updates and cleans up some of the scan
code which has not been updated in several years.

Reviewed-by: Brian Behlendorf <[email protected]>
Authored-by: Saso Kiselkov <[email protected]>
Authored-by: Alek Pinchuk <[email protected]>
Authored-by: Tom Caputi <[email protected]>
Signed-off-by: Tom Caputi <[email protected]>
Closes #3625 
Closes #6256 

1 files changed, 2 insertions, 20 deletions

diff --git a/module/zfs/zio.c b/module/zfs/zio.c
index 4cfda7a9e..311f79e23 100644
--- a/module/zfs/zio.c
+++ b/module/zfs/zio.c
@@ -39,6 +39,7 @@
 #include <sys/ddt.h>
 #include <sys/blkptr.h>
 #include <sys/zfeature.h>
+#include <sys/dsl_scan.h>
 #include <sys/metaslab_impl.h>
 #include <sys/time.h>
 #include <sys/trace_zio.h>
@@ -1050,6 +1051,7 @@ zio_free_sync(zio_t *pio, spa_t *spa, uint64_t txg, const blkptr_t *bp,
 
 	metaslab_check_free(spa, bp);
 	arc_freed(spa, bp);
+	dsl_scan_freed(spa, bp);
 
 	/*
 	 * GANG and DEDUP blocks can induce a read (for the gang block header,
@@ -3333,26 +3335,6 @@ zio_vdev_io_start(zio_t *zio)
 
 	ASSERT3P(zio->io_logical, !=, zio);
 
-	/*
-	 * We keep track of time-sensitive I/Os so that the scan thread
-	 * can quickly react to certain workloads.  In particular, we care
-	 * about non-scrubbing, top-level reads and writes with the following
-	 * characteristics:
-	 *	- synchronous writes of user data to non-slog devices
-	 *	- any reads of user data
-	 * When these conditions are met, adjust the timestamp of spa_last_io
-	 * which allows the scan thread to adjust its workload accordingly.
-	 */
-	if (!(zio->io_flags & ZIO_FLAG_SCAN_THREAD) && zio->io_bp != NULL &&
-	    vd == vd->vdev_top && !vd->vdev_islog &&
-	    zio->io_bookmark.zb_objset != DMU_META_OBJSET &&
-	    zio->io_txg != spa_syncing_txg(spa)) {
-		uint64_t old = spa->spa_last_io;
-		uint64_t new = ddi_get_lbolt64();
-		if (old != new)
-			(void) atomic_cas_64(&spa->spa_last_io, old, new);
-	}
-
 	align = 1ULL << vd->vdev_top->vdev_ashift;
 
 	if (!(zio->io_flags & ZIO_FLAG_PHYSICAL) &&