Prevent unnecessary resilver restarts

If a device is participating in an active resilver, then it will have a
non-empty DTL. Operations like vdev_{open,reopen,probe}() can cause the
resilver to be restarted (or deferred to be restarted later), which is
unnecessary if the DTL is still covered by the current scan range. This
is similar to the logic in vdev_dtl_should_excise() where the DTL can
only be excised if it's max txg is in the resilvered range.

Reviewed-by: Brian Behlendorf <[email protected]>
Reviewed-by: John Gallagher <[email protected]>
Reviewed-by: Kjeld Schouten <[email protected]>
Signed-off-by: John Poduska <[email protected]>
Issue #840 
Closes #9155
Closes #9378
Closes #9551
Closes #9588 

1 files changed, 48 insertions, 28 deletions

diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c
index 3a120b001..30f0515cf 100644
--- a/module/zfs/vdev.c
+++ b/module/zfs/vdev.c
@@ -27,6 +27,7 @@
  * Copyright 2016 Toomas Soome <[email protected]>
  * Copyright 2017 Joyent, Inc.
  * Copyright (c) 2017, Intel Corporation.
+ * Copyright (c) 2019, Datto Inc. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -835,7 +836,7 @@ vdev_alloc(spa_t *spa, vdev_t **vdp, nvlist_t *nv, vdev_t *parent, uint_t id,
 		    &vd->vdev_resilver_txg);
 
 		if (nvlist_exists(nv, ZPOOL_CONFIG_RESILVER_DEFER))
-			vdev_set_deferred_resilver(spa, vd);
+			vdev_defer_resilver(vd);
 
 		/*
 		 * In general, when importing a pool we want to ignore the
@@ -1873,18 +1874,12 @@ vdev_open(vdev_t *vd)
 	}
 
 	/*
-	 * If a leaf vdev has a DTL, and seems healthy, then kick off a
-	 * resilver.  But don't do this if we are doing a reopen for a scrub,
-	 * since this would just restart the scrub we are already doing.
+	 * If this is a leaf vdev, assess whether a resilver is needed.
+	 * But don't do this if we are doing a reopen for a scrub, since
+	 * this would just restart the scrub we are already doing.
 	 */
-	if (vd->vdev_ops->vdev_op_leaf && !spa->spa_scrub_reopen &&
-	    vdev_resilver_needed(vd, NULL, NULL)) {
-		if (dsl_scan_resilvering(spa->spa_dsl_pool) &&
-		    spa_feature_is_enabled(spa, SPA_FEATURE_RESILVER_DEFER))
-			vdev_set_deferred_resilver(spa, vd);
-		else
-			spa_async_request(spa, SPA_ASYNC_RESILVER);
-	}
+	if (vd->vdev_ops->vdev_op_leaf && !spa->spa_scrub_reopen)
+		dsl_scan_assess_vdev(spa->spa_dsl_pool, vd);
 
 	return (0);
 }
@@ -3703,14 +3698,11 @@ vdev_clear(spa_t *spa, vdev_t *vd)
 		if (vd != rvd && vdev_writeable(vd->vdev_top))
 			vdev_state_dirty(vd->vdev_top);
 
-		if (vd->vdev_aux == NULL && !vdev_is_dead(vd)) {
-			if (dsl_scan_resilvering(spa->spa_dsl_pool) &&
-			    spa_feature_is_enabled(spa,
-			    SPA_FEATURE_RESILVER_DEFER))
-				vdev_set_deferred_resilver(spa, vd);
-			else
-				spa_async_request(spa, SPA_ASYNC_RESILVER);
-		}
+		/* If a resilver isn't required, check if vdevs can be culled */
+		if (vd->vdev_aux == NULL && !vdev_is_dead(vd) &&
+		    !dsl_scan_resilvering(spa->spa_dsl_pool) &&
+		    !dsl_scan_resilver_scheduled(spa->spa_dsl_pool))
+			spa_async_request(spa, SPA_ASYNC_RESILVER_DONE);
 
 		spa_event_notify(spa, vd, NULL, ESC_ZFS_VDEV_CLEAR);
 	}
@@ -4703,18 +4695,46 @@ vdev_deadman(vdev_t *vd, char *tag)
 }
 
 void
-vdev_set_deferred_resilver(spa_t *spa, vdev_t *vd)
+vdev_defer_resilver(vdev_t *vd)
 {
-	for (uint64_t i = 0; i < vd->vdev_children; i++)
-		vdev_set_deferred_resilver(spa, vd->vdev_child[i]);
+	ASSERT(vd->vdev_ops->vdev_op_leaf);
 
-	if (!vd->vdev_ops->vdev_op_leaf || !vdev_writeable(vd) ||
-	    range_tree_is_empty(vd->vdev_dtl[DTL_MISSING])) {
-		return;
+	vd->vdev_resilver_deferred = B_TRUE;
+	vd->vdev_spa->spa_resilver_deferred = B_TRUE;
+}
+
+/*
+ * Clears the resilver deferred flag on all leaf devs under vd. Returns
+ * B_TRUE if we have devices that need to be resilvered and are available to
+ * accept resilver I/Os.
+ */
+boolean_t
+vdev_clear_resilver_deferred(vdev_t *vd, dmu_tx_t *tx)
+{
+	boolean_t resilver_needed = B_FALSE;
+	spa_t *spa = vd->vdev_spa;
+
+	for (int c = 0; c < vd->vdev_children; c++) {
+		vdev_t *cvd = vd->vdev_child[c];
+		resilver_needed |= vdev_clear_resilver_deferred(cvd, tx);
 	}
 
-	vd->vdev_resilver_deferred = B_TRUE;
-	spa->spa_resilver_deferred = B_TRUE;
+	if (vd == spa->spa_root_vdev &&
+	    spa_feature_is_active(spa, SPA_FEATURE_RESILVER_DEFER)) {
+		spa_feature_decr(spa, SPA_FEATURE_RESILVER_DEFER, tx);
+		vdev_config_dirty(vd);
+		spa->spa_resilver_deferred = B_FALSE;
+		return (resilver_needed);
+	}
+
+	if (!vdev_is_concrete(vd) || vd->vdev_aux ||
+	    !vd->vdev_ops->vdev_op_leaf)
+		return (resilver_needed);
+
+	vd->vdev_resilver_deferred = B_FALSE;
+
+	return (!vdev_is_dead(vd) && !vd->vdev_offline &&
+	    vdev_resilver_needed(vd, NULL, NULL));
 }
 
 /*