-
Type:
Bug
-
Status: Done
-
Priority:
Medium
-
Resolution: Fixed
-
Affects Version/s: 5.7.26-29, 8.0.x
-
Component/s: None
-
Labels:None
-
Upstream Bug URL:
-
Needs Review:Yes
Analysis done so far based on core from 5.7.26 (see attached) goes like:
All Threads that are in __lll_lock_wait wait for Thread 44
Thread 44 waits on srv_innodb_monitor_mutex
Thread 45 has srv_innodb_monitor_mutex at srv_printf_innodb_monitor - srv0srv.cc:1315
Thread 45 waits on ibuf_mutex at ibuf_print - ibuf0ibuf.cc:4924
Thread 57 has ibuf_mutex at ibuf_insert_low - ibuf0ibuf.cc:3448
Thread 57 waits on rw_lock_s_lock(&block->lock) - buf0buf.cc:4078
Attached is pmp for 5.7.26 and also for 5.7.21 which was suffering similar issues
List of queries extracted from core dump
[kolita@lilith ded30_us4]$ cat gdb.txt |egrep '^\$' $1 = 0x7ff07c5f6030 "show engine innodb status" $2 = 0x7ff1379f0030 "show full processlist" $3 = 0x7ff0838f4aa0 "FLUSH LOGS" $5 = 0x7ff23b3fd030 "SELECT * FROM information_schema.processlist WHERE STATE = 'Killing slave'" $8 = 0x7ff11f024030 "select count(*) from information_schema.processlist where User='editedvalue'" $9 = 0x7ff11dc24030 "SHOW GLOBAL STATUS" $10 = 0x7ff13fee8030 "SELECT * FROM information_schema.processlist WHERE STATE = 'Killing slave'" $11 = 0x7ff13e935030 "SHOW GLOBAL STATUS" $12 = 0x7ff1db824030 "SHOW ENGINE INNODB STATUS" $13 = 0x7ff1dbc24ab0 "show status like 'Uptime'" $14 = 0x7ff1ddc24030 "SHOW GLOBAL STATUS" $15 = 0x7ff23c024030 "select floor((count(*)/@@max_connections)*100) from information_schema.processlist"