We have a lot of customers who do click analysis, site analytics, search engine marketing, online advertising, user behavior analysis, and many similar types of work. The first thing these have in common is that they’re generally some kind of loggable event. The next characteristic of a lot of these systems (real or planned) is [...]
Beware of MyISAM Key Cache mutex contention
August 12, 2008 By 10 Comments
Today I was working with the client loading data to MyISAM tables at very high rate. Hundreds of millions rows are loaded daily into single MySQL instance with bursts up to 100K of records/sec which need to be inserted (in the table with few indexes). It was good not all records had to go to [...]
Enum Fields VS Varchar VS Int + Joined table: What is Faster?
January 24, 2008 By 30 Comments
Really often in customers’ application we can see a huge tables with varchar/char fields, with small sets of possible values. These are “state”, “gender”, “status”, “weapon_type”, etc, etc. Frequently we suggest to change such fields to use ENUM column type, but is it really necessary (from performance standpoint)? In this post I’d like to present [...]
Innodb Performance Optimization Basics
November 1, 2007 By 75 Comments
Interviewing people for our Job Openings I like to ask them a basic question – if you have a server with 16GB of RAM which will be dedicated for MySQL with large Innodb database using typical Web workload what settings you would adjust and interestingly enough most people fail to come up with anything reasonable. [...]
MyISAM Scalability and Innodb, Falcon Benchmarks
October 12, 2007 By 10 Comments
We many times wrote about InnoDB scalability problems, this time We are faced with one for MyISAM tables. We saw that several times in synthetic benchmarks but never in production, that’s why we did not escalate MyISAM scalability question. This time working on the customer system we figured out that box with 1 CPU Core [...]
Redundant index is not always bad
August 28, 2007 By 7 Comments
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) [...]
How much overhead is caused by on disk temporary tables
August 16, 2007 By 26 Comments
As you might know while running GROUP BY and some other kinds of queries MySQL needs to create temporary tables, which can be created in memory, using MEMORY storage engine or can be created on disk as MYISAM tables. Which one will be used depends on the allowed tmp_table_size and also by the data which [...]
Using CHAR keys for joins, how much is the overhead ?
June 18, 2007 By 18 Comments
I prefer to use Integers for joins whenever possible and today I worked with client which used character keys, in my opinion without a big need. I told them this is suboptimal but was challenged with rightful question about the difference. I did not know so I decided to benchmark. The results below are for [...]
Predicting how long data load would take
May 24, 2007 By 14 Comments
I had this question asked many times during last week, and there is really no easy answer. There are just way too many variables to consider especially while loading large schemas with a lot of different table structures. So which variables affect the load speed: Table Structure This one is tricky. The shorter rows you [...]
Innodb Recovery Update – The tricks what failed.
May 24, 2007 By 3 Comments
As I wrote we had been recovering corrupted Innodb Tablespace and it is finally done now. As this was over than 1TB worth of data we really tried to avoid dumping the data and find some other way to recovery. Examining Innodb page content and crash information we figured out it should be page of [...]
