Showing entries 11 to 14
« 10 Newer Entries
Displaying posts with tag: iibench (reset)
OpenSQLCamp Lightning Talk Videos

OpenSQLCamp was a huge success! Not many folks have blogged about what they learned there….if you missed it, all is not lost. We did take videos of most of the sessions (we only had 3 video cameras, and 4 rooms, and 2 sessions were not recorded).

All the videos have been processed, and I am working on uploading them to YouTube and filling in details for the video descriptions. Not all the videos are up right now….right now all the lightning talks are up.

[Read more]
InnoDB purge - another potential performance problem

If you have delete-intensive workloads on InnoDB, then you need to understand how purge works. Dimitri has an interesting post on this. And I wrote about measuring purge lag.

I haven't had to deal with this problem yet, but the insert benchmark has a new option to make it delete intensive. So, I think I can reproduce workloads that generate a lot of purge lag.

Performance impact of prefetching in InnoDB

InnoDB prefetches blocks when it detects multiple accesses to blocks within an extent. Unfortunately, there are no metrics in the server to determine whether it is effective. There are also weak metrics in the server to determine how frequently it is done -- counters incremented each time the readahead code prefetches one or more blocks rather than once per prefetch request.

There are cases where prefetch improves performance. A query that does a full table scan was run with prefetch enabled and disabled. It was 35% slower with prefetch disabled.

Percona and Matt have written about potential performance problems from this feature. There isn't …

[Read more]
InnoDB checksum performance

Once again Domas is unhappy with some aspect of Innodb performance and doing crazy things with gdb to tune it. I made it faster by changing the checksum code to process one 32-bit word at a time rather than one byte at a time. This will be in a future Google patch and is enabled with the parameter innodb_fast_checksum. This is not compatible with the old checksum so you must dump and reload the database to use it.

I measured the benefit using the insert benchmark from Tokutek on a server that can do a lot of IO. CPU overheads are measured using oprofile. The data below lists the percentage of time for the top 4 functions in mysqld. The checksum is computed in buf_calc_page_new_checksum. By using the fast checksum, the checksum overhead drops from 33.6% to 22.1% for gcc -O2 and …

[Read more]
Showing entries 11 to 14
« 10 Newer Entries