Details
-
Bug
-
Resolution: Fixed
-
Major
-
2.0-beta
-
Security Level: Public
-
None
-
Centos 5.4 - 64bit
Description
Regression in toy build, unable to drain items
While executing mixed workload on the toy-build (399) which does Bunch of gets/ sets/updates/deletes, I see that disk_write_queue is stuck and is unable to drain items. Items are still sitting in the disk_write_queue.
This same workload passes against DP3 (388). Let me know if you need more info.
You can use QAkey.pem to access these instances.
Some stats from Node that still has pending items in disk_queue.
ep_diskqueue_drain: 50607333
ep_diskqueue_fill: 51951650
ep_diskqueue_items: 1344317
ep_diskqueue_memory: 150563504
ep_diskqueue_pending: 4727847122
ep_flusher_todo: 210
ep_queue_age_cap: 900
ep_queue_size: 1344108
ep_tap_bg_fetch_requeued: 0
ep_tap_requeue_sleep_time: 0.1
ep_tap_throttle_queue_cap: 1000000
ep_total_enqueued: 51951650
vb_active_queue_age: 5513468882000
vb_active_queue_drain: 25441744
vb_active_queue_fill: 26278279
vb_active_queue_memory: 93691920
vb_active_queue_pending: 3000439279
vb_active_queue_size: 836535
vb_pending_queue_age: 0
vb_pending_queue_drain: 0
vb_pending_queue_fill: 0
vb_pending_queue_memory: 0
vb_pending_queue_pending: 0
vb_pending_queue_size: 0
vb_replica_queue_age: 3820300325000
vb_replica_queue_drain: 25165589
vb_replica_queue_fill: 25673371
vb_replica_queue_memory: 56871584
vb_replica_queue_pending: 1727407843
vb_replica_queue_size: 507782