Category Archives: mysql

on durability

MySQL did not start as a durable data store and had lots of mockery for that – (ISAM? no replication?). Eventually InnoDB took over, and it brought at least parts of MySQL into a reliable storage world. Checksummed pages, decent … Continue reading

Posted in facebook, mysql | 3 Comments

Hash of shame

I may retract this post, I may have been way too incorrect here, not sure yet. Hash maps were invented sixty years ago, apparently. MySQL reinvented them. Original idea was too fancy and too good, I guess. It allowed very … Continue reading

Posted in mysql | 9 Comments

MySQL 5.6 @ Facebook development tree Steaphan is a hero (well, everyone else on database engineering team are too) and he is driving efforts to publish MySQL 5.6 changes we’re making to the open. Now they’re on the github (yet … Continue reading

Link | Posted on by | Tagged ,

On performance schemas

You should probably read Marc Alff’s post about configuring Performance Schema in MySQL 5.6. We wrote another guide, how to start using user/table/index statistics first introduced in Google patch, now part of various MySQL branches and forks: Start MySQL

Posted in mysql | 3 Comments

The saddest bug of them all (SQL is dead?)

From time to time I will observe servers wasting lots of CPU when doing batch row operations. In perf top it will look like this: 8.24% mysqld [.] Arg_comparator::compare_int_unsigned() 7.17% mysqld [.] Item_cond_and::val_int() 4.37% mysqld [.] Item_field::val_int() 4.37% mysqld [.] … Continue reading

Posted in mysql | Tagged | 16 Comments

on wikipedia and mariadb

There’s some media coverage about Wikipedia switching to MariaDB, I just wanted to point out that performance figures cited are somewhat incorrect and don’t attribute gains to correct authors. Proper performance evaluation should include not just MariaDB 5.5 but Oracle’s … Continue reading

Posted in mysql, wikitech | 20 Comments

on seconds_behind_master sleuthing

With large enough infrastructure it gets a bit more and more complicated to detect whether an incident or a problem is real systems problem or a monitoring glitch. This is a story of one such investigation. With a sufficiently large … Continue reading

Posted in mysql | 1 Comment