aboutsummaryrefslogtreecommitdiffstats
path: root/module/zfs/spa.c
diff options
context:
space:
mode:
authorTom Caputi <[email protected]>2018-10-19 00:06:18 -0400
committerBrian Behlendorf <[email protected]>2018-10-18 21:06:18 -0700
commit80a91e7469669e2a5da5873b8f09a752f7869062 (patch)
treeef5a4462892becccb939b2cd42a54ed580f5894f /module/zfs/spa.c
parent9f438c5f948c0072f16431407a373ead34fabf6e (diff)
Defer new resilvers until the current one ends
Currently, if a resilver is triggered for any reason while an existing one is running, zfs will immediately restart the existing resilver from the beginning to include the new drive. This causes problems for system administrators when a drive fails while another is already resilvering. In this case, the optimal thing to do to reduce risk of data loss is to wait for the current resilver to end before immediately replacing the second failed drive, which allows the system to operate with two incomplete drives for the minimum amount of time. This patch introduces the resilver_defer feature that essentially does this for the admin without forcing them to wait and monitor the resilver manually. The change requires an on-disk feature since we must mark drives that are part of a deferred resilver in the vdev config to ensure that we do not assume they are done resilvering when an existing resilver completes. Reviewed-by: Matthew Ahrens <[email protected]> Reviewed-by: John Kennedy <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: @mmaybee Signed-off-by: Tom Caputi <[email protected]> Closes #7732
Diffstat (limited to 'module/zfs/spa.c')
-rw-r--r--module/zfs/spa.c16
1 files changed, 12 insertions, 4 deletions
diff --git a/module/zfs/spa.c b/module/zfs/spa.c
index fdce49c40..3785981b7 100644
--- a/module/zfs/spa.c
+++ b/module/zfs/spa.c
@@ -6059,9 +6059,14 @@ spa_vdev_attach(spa_t *spa, uint64_t guid, nvlist_t *nvroot, int replacing)
/*
* Schedule the resilver to restart in the future. We do this to
* ensure that dmu_sync-ed blocks have been stitched into the
- * respective datasets.
+ * respective datasets. We do not do this if resilvers have been
+ * deferred.
*/
- dsl_resilver_restart(spa->spa_dsl_pool, dtl_max_txg);
+ if (dsl_scan_resilvering(spa_get_dsl(spa)) &&
+ spa_feature_is_enabled(spa, SPA_FEATURE_RESILVER_DEFER))
+ vdev_set_deferred_resilver(spa, newvd);
+ else
+ dsl_resilver_restart(spa->spa_dsl_pool, dtl_max_txg);
if (spa->spa_bootfs)
spa_event_notify(spa, newvd, NULL, ESC_ZFS_BOOTFS_VDEV_ATTACH);
@@ -6933,6 +6938,7 @@ static void
spa_async_thread(void *arg)
{
spa_t *spa = (spa_t *)arg;
+ dsl_pool_t *dp = spa->spa_dsl_pool;
int tasks;
ASSERT(spa->spa_sync_on);
@@ -7008,8 +7014,10 @@ spa_async_thread(void *arg)
/*
* Kick off a resilver.
*/
- if (tasks & SPA_ASYNC_RESILVER)
- dsl_resilver_restart(spa->spa_dsl_pool, 0);
+ if (tasks & SPA_ASYNC_RESILVER &&
+ (!dsl_scan_resilvering(dp) ||
+ !spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_RESILVER_DEFER)))
+ dsl_resilver_restart(dp, 0);
/*
* Let the world know that we're done.