We are having issues with MySQL stalling. Overall speed of our system is quite good (when not stalled 
), but we encounter several times a day where the system will stall for 1 to 5 minutes. It seems after a large period of write activity the system is bottle necked writing to disk.
This has become a problem after we recently upgraded an onsite instance from 10.x to 12.x and migrated from a windows to linux stack.
The error below is seen in the mysql log:
[Note] InnoDB: page_cleaner: 1000ms intended loop took 180537ms. The settings might not be optimal. (flushed=200 and evicted=0, during the time.)
I have tried adjusting various mysql variables as suggested here: www.percona.com/.../
ex:
innodb_io_capacity_max=1500
innodb_io_capacity=200
innodb_lru_scan_depth=256
innodb_flush_neighbors=0
However, this has not helped. Has anyone had similar issues in the past? If so, any suggestions to help troubleshoot? Any input is appreciated.
Thanks,
Lee
