[draft] dirty buffer pool pages at the end of shutdown
Affects | Status | Importance | Assigned to | Milestone | ||
---|---|---|---|---|---|---|
Percona Server moved to https://jira.percona.com/projects/PS | Status tracked in 5.7 | |||||
5.5 |
New
|
Undecided
|
Unassigned | |||
5.6 |
Incomplete
|
High
|
Unassigned | |||
5.7 |
New
|
Undecided
|
Unassigned |
Bug Description
On 5.6 trunk:
innodb_
Test ended at 2017-03-25 10:10:25
line
Attempting backtrace. You can use the following information to find out
^ Found warnings in /mnt/workspace/
ok
...
Program terminated with signal SIGABRT, Aborted.
#0 0x00007f4a79e810c1 in __pthread_kill (threadid=
61 ../nptl/
#0 0x00007f4a79e810c1 in __pthread_kill (threadid=
#1 0x0000000000ac9dea in my_write_core (sig=6) at /mnt/workspace/
#2 0x000000000073c192 in handle_fatal_signal (sig=6) at /mnt/workspace/
#3 <signal handler called>
#4 0x00007f4a78014067 in __GI_raise (sig=sig@entry=6) at ../nptl/
#5 0x00007f4a78015448 in __GI_abort () at abort.c:89
#6 0x0000000000d2ceab in buf_flush_
#7 0x00007f4a79e7c064 in start_thread (arg=0x7f4a633f
#8 0x00007f4a780c762d in clone () at ../sysdeps/
...
Thread 12 (Thread 0x7f4a4ffff700 (LWP 4035)):
#0 pthread_
#1 0x0000000000bd41d3 in os_cond_wait (cond=0x1ceba58, fast_mutex=
#2 0x0000000000bd4506 in os_event_wait_low (event=0x1ceba20, reset_sig_count=42) at /mnt/workspace/
#3 0x0000000000d2acf2 in buf_flush_
#4 0x0000000000d2b6c5 in buf_flush_LRU_tail () at /mnt/workspace/
#5 0x0000000000d2cfdd in buf_flush_
#6 0x00007f4a79e7c064 in start_thread (arg=0x7f4a4fff
#7 0x00007f4a780c762d in clone () at ../sysdeps/
I/O handler threads are idle.
It is not clear why the LRU flusher is stuck on waiting for a batch to complete.
This could be related to https:/
See bug 1689552