Tagged with "databases"

2019-05-27 10:14 UTC

Fun MySQL fact of the day: the IO thread

On Friday, we looked at a pretty fun and very handy use of the MySQL binary log to perform point-in-time recovery during a make-believe disaster. We're going to circle back to point-in-time recovery a bit later this week, but I want to first start looking at MySQL's replication implementation. ... read more

2019-05-24 22:18 UTC

Fun MySQL fact of the day: `mysqlbinlog`

It's 3:37 A.M. when your phone blares out. Your hands are numbed from the early on-set of carpel tunnel and you can't hush the wailing siren that has become of your phone. With adrenaline coursing through your body, you shove your phone under your pillow mashing around in the dark ... read more

2019-05-23 20:20 UTC

Fun MySQL fact of the day: the binary log is binary

Yesterday, we started to form a (somewhat broken) mental model of MySQL's binary log. Today, we'll start taking a look at the on-disk representation before diving right in and using some of MySQL's tools to do fun (and useful!) things. ... read more

2019-05-22 23:55 UTC

Fun MySQL fact of the day: the binary log

The very first Fun MySQL fact of the day discussed a MySQL flag, sync_binlog, which is, indeed, a counter, and while we discussed that it can be used to tune a MySQL database's durability, we didn't really dig much deeper. Over the next little while, we'll do just that ... read more

2019-05-21 23:58 UTC

Fun MySQL fact of the day: duplicate weedout

Last week, we started looking at some of MySQL's semi-join optimisations for subqueries. We've already covered materialisiation, FirstMatch, LooseScan, and table pullout optimisations, but there's still one more to clover: the DuplicateWeedout optimisation, and then thistle be the last topic we discuss about about subqueries. ... read more

2019-05-17 23:06 UTC

Fun MySQL fact of the day: loose scan semi-join

Yesterday, we considered MySQL's FirstMatch semi-join optimisation, and while it's pretty simple and fairly straight-forward, some queries may just not benefit from it. ... read more

2019-05-16 23:45 UTC

Fun MySQL fact of the day: first match optimisation

If we consider back to Tuesday, one of MySQL's guarantees about subqueries is that it will not return the same row in the outer query twice. Compare this to a join, which will return a copy of the left-hand-table's row for each matching right-hand-table row. ... read more

2019-05-15 23:35 UTC

Fun MySQL fact of the day: subquery materialization

Yesterday, we started looking at the semi-join strategies MySQL provides, and today, we're going to continue looking into semi-joins by considering the "materialisation" strategy. ... read more

2019-05-14 23:30 UTC

Fun MySQL fact of the day: table pullouts

As noted yesterday, MySQL 5.6 was released with a large number of optimisations to the subquery engine. In particular, a number of changes were made to optimise for non-scalar, uncorrelated subqueries, allowing them to be run as... you guess it! Joins. ... read more

2019-05-13 23:05 UTC

Fun MySQL fact of the day: subqueries in MySQL5.5

The year 2013 can be remembered by 3 separate things: Justin Timberlake, the series finale of Breaking Bad, and the release of MySQL 5.6. Though, if you aren't a MySQL historian, the release of MySQL 5.6 may have passed you by and you may have missed the significance as it ... read more

« Previous 1 2 3 4 5 6 Next »