April 20, 2014

Extending Index for Innodb tables can hurt performance in a surprising way

One schema optimization we often do is extending index when there are queries which can use more key part. Typically this is safe operation, unless index length increases dramatically queries which can use index can also use prefix of the new index are they ? It turns there are special cases when this is not […]

Recovering Innodb table Corruption

Assume you’re running MySQL with Innodb tables and you’ve got crappy hardware, driver bug, kernel bug, unlucky power failure or some rare MySQL bug and some pages in Innodb tablespace got corrupted. In such cases Innodb will typically print something like this: InnoDB: Database page corruption on disk or a failed InnoDB: file read of […]

Should you name indexes while doing ALTER TABLE ?

MySQL Server does not require you to specify name of the index if you’re running ALTER TABLE statement – it is optional. Though what might be good practical reasons to specify the key name or omit ? Things what you should be looking at is how MySQL names indexes automatically as well as what maintaining […]

Hacking to make ALTER TABLE online for certain changes

Suppose you want to remove auto_increment from 100G table. No matter if it’s InnoDB or MyISAM, you’d usually ALTER TABLE huge_table CHANGE id id int(6) NOT NULL and then wait hours for table rebuild to complete. If you’re unlucky i.e. you have a lot of indexes and not too much RAM – you could end […]

Redundant index is not always bad

About year ago Peter wrote about redundant indexes and mentioned sometimes it is good to leave two indexes, even one is first part of another. I’m speaking about BTREE indexes, for example, KEY (A), and KEY (A,B). From SQL point of view KEY(A) is not needed, as for queries like WHERE A=5 the index (A,B) […]

The ARCHIVE Storage Engine – does it do what you expect?

Sometimes there is a need for keeping large amounts of old, rarely used data without investing too much on expensive storage. Very often such data doesn’t need to be updated anymore, or the intent is to leave it untouched. I sometimes wonder what I should really suggest to our Support customers. For this purpose, the […]

How InnoDB promotes UNIQUE constraints

The other day I was running pt-duplicate-key-checker on behalf of a customer and noticed some peculiar recommendations on an InnoDB table with an odd structure (no PRIMARY key, but multiple UNIQUE constraints). This got me thinking about how InnoDB promotes UNIQUE constraints to the role of PRIMARY KEYs. The documentation is pretty clear: [DOCS] When […]

Benchmarking Percona Server TokuDB vs InnoDB

After compiling Percona Server with TokuDB, of course I wanted to compare InnoDB performance vs TokuDB. I have a particular workload I’m interested in testing – it is an insert-intensive workload (which is TokuDB’s strong suit) with some roll-up aggregation, which should produce updates in-place (I will use INSERT .. ON DUPLICATE KEY UPDATE statements […]

Concatenating MyISAM files

Recently, I found myself involved in the migration of a large read-only InnoDB database to MyISAM (eventually packed). The only issue was that for one of the table, we were talking of 5 TB of data, 23B rows. Not small… I calculated that with something like insert into MyISAM_table… select * from Innodb_table… would take […]

InnoDB’s gap locks

One of the most important features of InnoDB is the row level locking. This feature provides better concurrency under heavy write load but needs additional precautions to avoid phantom reads and to get a consistent Statement based replication. To accomplish that, row level locking databases also acquire gap locks. What is a Phantom Read A […]