Repeatedly on my slave server, I am getting assertion failures after 7-8 hrs for semaphore wait and has lasted for more than 600 seconds.
General load on server - jobs run on this server creating temporary tables
MySQL Version - 5.6.21-70.1
OS - Linux
Hi MySQL-user;
This is a tricky one, as you could be running into legitimate contention issues that are causing this, or it could be a bug. One thing you could try is setting Innodb_adaptive_hash_index=0 to disable the adaptive hash index. This appears to fix similar issues others are having, so it is worth a try.
If that does not work, post the rest of the monitor output so we can see what InnoDB is up to.
[url]MySQL :: MySQL 5.6 Reference Manual :: 14.14 InnoDB Startup Options and System Variables
[url]http://dev.mysql.com/doc/refman/5.6/en/innodb-adaptive-hash.html[/url]
-Scott
Thanks Scott !!
I disabled the adaptive hash index for now. After that I have not seen any crashes, which I suppose that activity on database has also gone down (from application side). I will be monitoring the same for few days.
As you identified that a possible reason could be legitimate contention issues, what would be your suggestion to resolve or lower the degree of contention happening in MySQL.
Thanks!!
Will post you entire innodb status observed during a crash.