Details
-
Bug
-
Status: Resolved
-
High
-
Resolution: Done
-
Oxygen-SR4, Neon, Sodium, Magnesium
-
None
Description
We experience frequent safepoint pauses seen in this log:
// code
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2653.486: FindDeadlocks [ 870 1 1 ] [ 1 0 2 1 0 ] 1
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2653.986: FindDeadlocks [ 870 0 0 ] [ 0 0 0 1 0 ] 0
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2654.486: FindDeadlocks [ 870 0 3 ] [ 0 0 0 1 0 ] 0
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2654.986: FindDeadlocks [ 870 0 2 ] [ 0 0 0 1 0 ] 0
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2655.486: FindDeadlocks [ 871 0 1 ] [ 0 0 0 1 0 ] 0
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2655.986: FindDeadlocks [ 871 0 0 ] [ 0 0 0 1 0 ] 0
vmop [threads: total initially_running wait_to_block] [time: spin block sync cleanup vmop] page_trap_count
2656.486: FindDeadlocks [ 870 0 0 ] [ 0 0 0 1 0 ] 0
Which was traced to ThreadsWatcher calling FindDeadlocks every 500ms.
Attachments
| # | Subject | Branch | Project | Status | CR | V |
|---|---|---|---|---|---|---|
| 86025,2 | Disable ThreadsWatcher by default | master | infrautils | Status: MERGED | +2 | +1 |
| 86027,1 | Disable ThreadsWatcher by default | stable/neon | infrautils | Status: MERGED | +2 | +1 |
| 86029,1 | Disable ThreadsWatcher by default | stable/sodium | infrautils | Status: MERGED | +2 | +1 |
