c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2026-05-25 03:37:45 +03:00
Code Issues Packages Projects Releases Wiki Activity

Parallelize vdev_load

Browse Source 
metaslab_init is the slowest part of importing a mature pool, and it
must be repeated hundreds of times for each top-level vdev.  But its
speed is dominated by a few serialized disk accesses.  That can lead to
import times of > 1 hour for pools with many top-level vdevs on spinny
disks.

Speed up the import by using a taskqueue to parallelize vdev_load across
all top-level vdevs.

This also requires adding mutex protection to
metaslab_class_t.mc_historgram.  The mc_histogram fields were
unprotected when that code was first written in "Illumos 4976-4984 -
metaslab improvements" (OpenZFS
f3a7f6610f).  The lock wasn't added until
3dfb57a35e, though it's unclear exactly
which fields it's supposed to protect.  In any case, it wasn't until
vdev_load was parallelized that any code attempted concurrent access to
those fields.

Sponsored by: Axcient
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alan Somers <asomers@gmail.com>
Closes #11470
This commit is contained in:
Alan Somers
2021-01-11 17:00:19 -07:00
committed by Brian Behlendorf
parent dfb44c500e
commit a0e01997ec
3 changed files with 46 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
module/zfs/metaslab.c
+6
View File
@@ -522,6 +522,7 @@ metaslab_class_histogram_verify(metaslab_class_t *mc)
mc_hist = kmem_zalloc(sizeof (uint64_t) * RANGE_TREE_HISTOGRAM_SIZE,
KM_SLEEP);
mutex_enter(&mc->mc_lock);
for (int c = 0; c < rvd->vdev_children; c++) {
vdev_t *tvd = rvd->vdev_child[c];
metaslab_group_t *mg = vdev_get_mg(tvd, mc);
@@ -546,6 +547,7 @@ metaslab_class_histogram_verify(metaslab_class_t *mc)
VERIFY3U(mc_hist[i], ==, mc->mc_histogram[i]);
}
mutex_exit(&mc->mc_lock);
kmem_free(mc_hist, sizeof (uint64_t) * RANGE_TREE_HISTOGRAM_SIZE);
}
@@ -1067,6 +1069,7 @@ metaslab_group_histogram_add(metaslab_group_t *mg, metaslab_t *msp)
return;
mutex_enter(&mg->mg_lock);
mutex_enter(&mc->mc_lock);
for (int i = 0; i < SPACE_MAP_HISTOGRAM_SIZE; i++) {
IMPLY(mg == mg->mg_vd->vdev_log_mg,
mc == spa_embedded_log_class(mg->mg_vd->vdev_spa));
@@ -1075,6 +1078,7 @@ metaslab_group_histogram_add(metaslab_group_t *mg, metaslab_t *msp)
mc->mc_histogram[i + ashift] +=
msp->ms_sm->sm_phys->smp_histogram[i];
}
mutex_exit(&mc->mc_lock);
mutex_exit(&mg->mg_lock);
}
@@ -1089,6 +1093,7 @@ metaslab_group_histogram_remove(metaslab_group_t *mg, metaslab_t *msp)
return;
mutex_enter(&mg->mg_lock);
mutex_enter(&mc->mc_lock);
for (int i = 0; i < SPACE_MAP_HISTOGRAM_SIZE; i++) {
ASSERT3U(mg->mg_histogram[i + ashift], >=,
msp->ms_sm->sm_phys->smp_histogram[i]);
@@ -1102,6 +1107,7 @@ metaslab_group_histogram_remove(metaslab_group_t *mg, metaslab_t *msp)
mc->mc_histogram[i + ashift] -=
msp->ms_sm->sm_phys->smp_histogram[i];
}
mutex_exit(&mc->mc_lock);
mutex_exit(&mg->mg_lock);
}