diff options
| author | Kotresh HR <khiremat@redhat.com> | 2016-04-29 17:45:31 +0530 |
|---|---|---|
| committer | Venky Shankar <vshankar@redhat.com> | 2016-05-19 23:18:25 -0700 |
| commit | e63049cba0f2b94b0b082cc8bfd5921b8443ef45 (patch) | |
| tree | 90037fc5942b63cddcfda8ecec6e7d6b8cd5b015 /xlators/features/bit-rot/src/bitd/bit-rot.c | |
| parent | 2e2ef4f833abe28f7895ea022654aac74e353702 (diff) | |
features/bitrot: Introduce scrubber monitor thread
The patch does following changes.
1. Introduce scrubber monitor thread.
2. Move scrub status related APIs to separate file
and make part of libbitrot library.
Problem:
Earlier, each child of the scrubber was maintaining
the state machine and hence there was no way to track
the start and end time of scrubbing as each brick has
it's own start and end time. Also each brick was maintaining
it's own timer wheel instance. It was also not possible
to get scrubbed files count per session as we could not
get last child which finishes scrubbing to reset it to
zero.
Solution:
Introduce scrubber monitor thread. It does following.
1. Maintains the scrubber state machine. Earlier each
child had it's own state machine. Now, only monitor
maintains on behalf of all it's children.
2. Maintains the timer wheel instance. Earlier each
child had it's own timer wheel instance. Now, only
monitor maintains on behalf of all it's children.
As a result, we can track the scrub statistics easily
and correctly.
Upstream master:
http://review.gluster.org/14044
http://review.gluster.org/#/c/14146
Upstream 3.7:
http://review.gluster.org/14140
NOTE: The patch #14146 is a compilation warning not detected
in master branch and detected only in 3.7 branch. Since the
compilation warning is introduced by patch #14044, the above
two backports are made into this single patch in 3.7 and the
same is backported to 3.8.
BUG: 1332839
Change-Id: I1cec0b229c851fec793a1288c23e43bcec755cba
Signed-off-by: Kotresh HR <khiremat@redhat.com>
Reviewed-on: http://review.gluster.org/14201
Smoke: Gluster Build System <jenkins@build.gluster.com>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Diffstat (limited to 'xlators/features/bit-rot/src/bitd/bit-rot.c')
| -rw-r--r-- | xlators/features/bit-rot/src/bitd/bit-rot.c | 178 |
1 files changed, 84 insertions, 94 deletions
diff --git a/xlators/features/bit-rot/src/bitd/bit-rot.c b/xlators/features/bit-rot/src/bitd/bit-rot.c index e4d9113bd25..ed436c34dc4 100644 --- a/xlators/features/bit-rot/src/bitd/bit-rot.c +++ b/xlators/features/bit-rot/src/bitd/bit-rot.c @@ -1092,21 +1092,11 @@ br_oneshot_signer (void *arg) static void br_set_child_state (br_child_t *child, br_child_state_t state) { - LOCK (&child->lock); + pthread_mutex_lock (&child->lock); { _br_set_child_state (child, state); } - UNLOCK (&child->lock); -} - -static void -br_set_scrub_state (br_child_t *child, br_scrub_state_t state) -{ - LOCK (&child->lock); - { - _br_child_set_scrub_state (child, state); - } - UNLOCK (&child->lock); + pthread_mutex_unlock (&child->lock); } /** @@ -1168,11 +1158,11 @@ br_launch_scrubber (xlator_t *this, br_child_t *child, { int32_t ret = -1; br_private_t *priv = NULL; + struct br_monitor *scrub_monitor = NULL; priv = this->private; - fsscan->kick = _gf_false; - fsscan->over = _gf_false; + scrub_monitor = &priv->scrub_monitor; ret = gf_thread_create (&child->thread, NULL, br_fsscanner, child); if (ret != 0) { gf_msg (this->name, GF_LOG_ALERT, 0, BRB_MSG_SPAWN_FAILED, @@ -1181,14 +1171,14 @@ br_launch_scrubber (xlator_t *this, br_child_t *child, goto error_return; } - /* this needs to be serialized with reconfigure() */ - pthread_mutex_lock (&priv->lock); + /* Signal monitor to kick off state machine*/ + pthread_mutex_lock (&scrub_monitor->mutex); { - ret = br_scrub_state_machine (this, child); + if (!scrub_monitor->inited) + pthread_cond_signal (&scrub_monitor->cond); + scrub_monitor->inited = _gf_true; } - pthread_mutex_unlock (&priv->lock); - if (ret) - goto cleanup_thread; + pthread_mutex_unlock (&scrub_monitor->mutex); /** * Everything has been setup.. add this subvolume to scrubbers @@ -1203,8 +1193,6 @@ br_launch_scrubber (xlator_t *this, br_child_t *child, return 0; - cleanup_thread: - (void) gf_thread_cleanup_xint (child->thread); error_return: return -1; } @@ -1237,10 +1225,6 @@ br_enact_scrubber (xlator_t *this, br_child_t *child) INIT_LIST_HEAD (&fsscan->queued); INIT_LIST_HEAD (&fsscan->ready); - /* init scheduler related variables */ - pthread_mutex_init (&fsscan->wakelock, NULL); - pthread_cond_init (&fsscan->wakecond, NULL); - ret = br_launch_scrubber (this, child, fsscan, fsscrub); if (ret) goto error_return; @@ -1261,7 +1245,7 @@ br_child_enaction (xlator_t *this, br_child_t *child, br_stub_init_t *stub) int32_t ret = -1; br_private_t *priv = this->private; - LOCK (&child->lock); + pthread_mutex_lock (&child->lock); { if (priv->iamscrubber) ret = br_enact_scrubber (this, child); @@ -1276,7 +1260,7 @@ br_child_enaction (xlator_t *this, br_child_t *child, br_stub_init_t *stub) "Connected to brick %s..", child->brick_path); } } - UNLOCK (&child->lock); + pthread_mutex_unlock (&child->lock); return ret; } @@ -1303,6 +1287,7 @@ br_brick_connect (xlator_t *this, br_child_t *child) GF_VALIDATE_OR_GOTO (this->name, child, out); GF_VALIDATE_OR_GOTO (this->name, this->private, out); + br_child_set_scrub_state (child, _gf_false); br_set_child_state (child, BR_CHILD_STATE_INITIALIZING); loc.inode = inode_ref (child->table->root); @@ -1364,12 +1349,17 @@ br_cleanup_scrubber (xlator_t *this, br_child_t *child) { int32_t ret = 0; br_private_t *priv = NULL; - struct br_scanfs *fsscan = NULL; struct br_scrubber *fsscrub = NULL; + struct br_monitor *scrub_monitor = NULL; priv = this->private; - fsscan = &child->fsscan; fsscrub = &priv->fsscrub; + scrub_monitor = &priv->scrub_monitor; + + if (_br_is_child_scrub_active (child)) { + scrub_monitor->active_child_count--; + br_child_set_scrub_state (child, _gf_false); + } /** * 0x0: child (brick) goes out of rotation @@ -1401,21 +1391,6 @@ br_cleanup_scrubber (xlator_t *this, br_child_t *child) 0, BRB_MSG_SCRUB_THREAD_CLEANUP, "Error cleaning up scanner thread"); - /** - * 0x2: free()up resources - */ - if (fsscan->timer) { - (void) gf_tw_del_timer (priv->timer_wheel, fsscan->timer); - - GF_FREE (fsscan->timer); - fsscan->timer = NULL; - } - - /** - * 0x3: reset scrubber state - */ - _br_child_set_scrub_state (child, BR_SCRUB_STATE_INACTIVE); - gf_msg (this->name, GF_LOG_INFO, 0, BRB_MSG_SCRUBBER_CLEANED, "Cleaned up scrubber for brick [%s]", child->brick_path); @@ -1432,23 +1407,33 @@ int32_t br_brick_disconnect (xlator_t *this, br_child_t *child) { int32_t ret = 0; + struct br_monitor *scrub_monitor = NULL; br_private_t *priv = this->private; - LOCK (&child->lock); + scrub_monitor = &priv->scrub_monitor; + + /* Lock order should be wakelock and then child lock to + * dead locks. + */ + pthread_mutex_lock (&scrub_monitor->wakelock); { - if (!_br_is_child_connected (child)) - goto unblock; + pthread_mutex_lock (&child->lock); + { + if (!_br_is_child_connected (child)) + goto unblock; - /* child is on death row.. */ - _br_set_child_state (child, BR_CHILD_STATE_DISCONNECTED); + /* child is on death row.. */ + _br_set_child_state (child, BR_CHILD_STATE_DISCONNECTED); - if (priv->iamscrubber) - ret = br_cleanup_scrubber (this, child); - else - ret = br_cleanup_signer (this, child); - } + if (priv->iamscrubber) + ret = br_cleanup_scrubber (this, child); + else + ret = br_cleanup_signer (this, child); + } unblock: - UNLOCK (&child->lock); + pthread_mutex_unlock (&child->lock); + } + pthread_mutex_unlock (&scrub_monitor->wakelock); return ret; } @@ -1569,7 +1554,7 @@ br_scrubber_status_get (xlator_t *this, dict_t **dict) memset (key, 0, 256); snprintf (key, 256, "scrubbed-files"); - ret = dict_set_uint32 (*dict, key, scrub_stats->scrubbed_files); + ret = dict_set_uint64 (*dict, key, scrub_stats->scrubbed_files); if (ret) { gf_msg_debug (this->name, 0, "Failed to setting scrubbed file " "entry to the dictionary"); @@ -1577,7 +1562,7 @@ br_scrubber_status_get (xlator_t *this, dict_t **dict) memset (key, 0, 256); snprintf (key, 256, "unsigned-files"); - ret = dict_set_uint32 (*dict, key, scrub_stats->unsigned_files); + ret = dict_set_uint64 (*dict, key, scrub_stats->unsigned_files); if (ret) { gf_msg_debug (this->name, 0, "Failed to set unsigned file count" " entry to the dictionary"); @@ -1585,7 +1570,7 @@ br_scrubber_status_get (xlator_t *this, dict_t **dict) memset (key, 0, 256); snprintf (key, 256, "scrub-duration"); - ret = dict_set_uint32 (*dict, key, scrub_stats->scrub_duration); + ret = dict_set_uint64 (*dict, key, scrub_stats->scrub_duration); if (ret) { gf_msg_debug (this->name, 0, "Failed to set scrub duration" " entry to the dictionary"); @@ -1844,6 +1829,33 @@ br_signer_init (xlator_t *this, br_private_t *priv) } static void +br_free_scrubber_monitor (xlator_t *this, br_private_t *priv) +{ + struct br_monitor *scrub_monitor = &priv->scrub_monitor; + + if (scrub_monitor->timer) { + (void) gf_tw_del_timer (priv->timer_wheel, scrub_monitor->timer); + + GF_FREE (scrub_monitor->timer); + scrub_monitor->timer = NULL; + } + + (void) gf_thread_cleanup_xint (scrub_monitor->thread); + + /* Clean up cond and mutex variables */ + pthread_mutex_destroy (&scrub_monitor->mutex); + pthread_cond_destroy (&scrub_monitor->cond); + + pthread_mutex_destroy (&scrub_monitor->wakelock); + pthread_cond_destroy (&scrub_monitor->wakecond); + + pthread_mutex_destroy (&scrub_monitor->donelock); + pthread_cond_destroy (&scrub_monitor->donecond); + + LOCK_DESTROY (&scrub_monitor->lock); +} + +static void br_free_children (xlator_t *this, br_private_t *priv, int count) { br_child_t *child = NULL; @@ -1851,7 +1863,7 @@ br_free_children (xlator_t *this, br_private_t *priv, int count) for (--count; count >= 0; count--) { child = &priv->children[count]; mem_pool_destroy (child->timer_pool); - LOCK_DESTROY (&child->lock); + pthread_mutex_destroy (&child->lock); } GF_FREE (priv->children); @@ -1875,10 +1887,9 @@ br_init_children (xlator_t *this, br_private_t *priv) while (trav) { child = &priv->children[i]; - LOCK_INIT (&child->lock); + pthread_mutex_init (&child->lock, NULL); child->witnessed = 0; - br_set_scrub_state (child, BR_SCRUB_STATE_INACTIVE); br_set_child_state (child, BR_CHILD_STATE_DISCONNECTED); child->this = this; @@ -1999,6 +2010,9 @@ fini (xlator_t *this) if (!priv->iamscrubber) br_fini_signer (this, priv); + else + (void) br_free_scrubber_monitor (this, priv); + br_free_children (this, priv, priv->child_count); this->private = NULL; @@ -2008,26 +2022,23 @@ fini (xlator_t *this) } static void -br_reconfigure_child (xlator_t *this, br_child_t *child) +br_reconfigure_monitor (xlator_t *this) { int32_t ret = 0; - ret = br_scrub_state_machine (this, child); + ret = br_scrub_state_machine (this); if (ret) { gf_msg (this->name, GF_LOG_ERROR, 0, BRB_MSG_RESCHEDULE_SCRUBBER_FAILED, - "Could not reschedule scrubber for brick: %s. Scubbing " - "will continue according to old frequency.", - child->brick_path); + "Could not reschedule scrubber for the volume. Scrubbing " + "will continue according to old frequency."); } } static int br_reconfigure_scrubber (xlator_t *this, dict_t *options) { - int i = 0; int32_t ret = -1; - br_child_t *child = NULL; br_private_t *priv = NULL; priv = this->private; @@ -2042,32 +2053,11 @@ br_reconfigure_scrubber (xlator_t *this, dict_t *options) goto err; /* change state for all _up_ subvolume(s) */ - for (; i < priv->child_count; i++) { - child = &priv->children[i]; - - LOCK (&child->lock); - { - if (_br_child_failed_conn (child)) { - gf_msg (this->name, GF_LOG_INFO, - 0, BRB_MSG_BRICK_INFO, - "Scrubber for brick [%s] failed " - "initialization, rescheduling is " - "skipped", child->brick_path); - goto unblock; - } - - if (_br_is_child_connected (child)) - br_reconfigure_child (this, child); - - /** - * for the rest.. either the child is in initialization - * phase or is disconnected. either way, updated values - * would be reflected on successful connection. - */ - } - unblock: - UNLOCK (&child->lock); + pthread_mutex_lock (&priv->lock); + { + br_reconfigure_monitor (this); } + pthread_mutex_unlock (&priv->lock); err: return ret; |
