diff options
| author | Ashish Pandey <aspandey@redhat.com> | 2018-10-18 17:15:58 +0530 |
|---|---|---|
| committer | Shyamsundar Ranganathan <srangana@redhat.com> | 2018-11-08 19:05:46 +0000 |
| commit | d1b0de23c1fdc44a4ef661a8bb03c27228045634 (patch) | |
| tree | e89afd63d44d1a74ffbf1a1a7fbc67f24b97d90b /xlators | |
| parent | b651cb3e612214c8fdb434af2b95ca0c018b5c36 (diff) | |
cluster/afr : Check for UP bricks before starting heal
Problem:
Currently for replica volume, even if only one brick is UP
SHD will keep crawling index entries even if it can not
heal anything.
In thin-arbiter volume which is also a replica 2 volume,
this causes inode lock contention which in turn sends
upcall to all the clients to release notify locks, even
if it can not do anything for healing.
This will slow down the client performance and kills the
purpose of keeping in memory information about bad brick.
Solution: Before starting heal or even crawling, check if
sufficient number of children are UP and available to check
and heal entries.
(cherry picked from commit f73b4476b15f9d6d3dc3c8e20c9742aacd857f9f)
Change-Id: I011c9da3b37cae275f791affd56b8f1c1ac9255d
updates: bz#1644645
Signed-off-by: Ashish Pandey <aspandey@redhat.com>
Diffstat (limited to 'xlators')
| -rw-r--r-- | xlators/cluster/afr/src/afr-common.c | 2 | ||||
| -rw-r--r-- | xlators/cluster/afr/src/afr-self-heald.c | 15 | ||||
| -rw-r--r-- | xlators/cluster/afr/src/afr.h | 3 |
3 files changed, 19 insertions, 1 deletions
diff --git a/xlators/cluster/afr/src/afr-common.c b/xlators/cluster/afr/src/afr-common.c index 51d611d503a..0971104cc10 100644 --- a/xlators/cluster/afr/src/afr-common.c +++ b/xlators/cluster/afr/src/afr-common.c @@ -4919,7 +4919,7 @@ find_child_index(xlator_t *this, xlator_t *child) return i; } -static int +int __afr_get_up_children_count(afr_private_t *priv) { int up_children = 0; diff --git a/xlators/cluster/afr/src/afr-self-heald.c b/xlators/cluster/afr/src/afr-self-heald.c index 0cf01a041b4..72b7d0d56ba 100644 --- a/xlators/cluster/afr/src/afr-self-heald.c +++ b/xlators/cluster/afr/src/afr-self-heald.c @@ -716,6 +716,18 @@ out: return; } +gf_boolean_t +afr_bricks_available_for_heal(afr_private_t *priv) +{ + int up_children = 0; + + up_children = __afr_get_up_children_count(priv); + if (up_children < 2) { + return _gf_false; + } + return _gf_true; +} + void * afr_shd_index_healer(void *data) { @@ -735,6 +747,9 @@ afr_shd_index_healer(void *data) for (;;) { afr_shd_healer_wait(healer); + if (!afr_bricks_available_for_heal(priv)) + continue; + ASSERT_LOCAL(this, healer); priv->local[healer->subvol] = healer->local; diff --git a/xlators/cluster/afr/src/afr.h b/xlators/cluster/afr/src/afr.h index cc4bceef521..3d2c1950571 100644 --- a/xlators/cluster/afr/src/afr.h +++ b/xlators/cluster/afr/src/afr.h @@ -1284,6 +1284,9 @@ afr_ta_post_op_unlock(xlator_t *this, loc_t *loc); gf_boolean_t afr_is_pending_set(xlator_t *this, dict_t *xdata, int type); +int +__afr_get_up_children_count(afr_private_t *priv); + call_frame_t * afr_ta_frame_create(xlator_t *this); #endif /* __AFR_H__ */ |