New answers tagged percona
With that kind of AWS instance, you must be planning heavy writes. I would add the following innodb_log_buffer_size = 32M innodb_read_io_threads = 16 innodb_write_io_threads = 16
If the file hasn't exceeded max_binlog_size or if a log flushing event hasn't occurred then it won't start a new file. expire_logs_days only removes non-active binary logs (i.e. files other than the one that shows up under a "show master status;").
I would suggest that even though your innodb_buffer_pool_size is undoubtedly too large for the available memory (192GB would be the theoretical max "sane" value) your thread-cache-size is likely to be what's occasionally pushing you over the edge. There's no rational justification for such a large value, particularly with max connections limited to 1200. ...
OMG 220G out of 240G ??? WAY TOO BIG !!! Take a look at this Pictorial Representation of InnoDB from Percona CTO Vadim Tkachenko Look at the Buffer Pool layout. You allocated 220G but it's all not available for data. There is a setting called innodb_change_buffer_max_size. The default is 25. This means that 25% of the Buffer Pool is dedicated to ...
Top 50 recent answers are included