Our customers or prospective customers often ask us how we do a performance audit (it’s our most popular service). I thought I should write a blog post that will both answer their question, so I can just reply “read all about it at this URL” and share our methodology with readers a little bit. This [...]
Thoughs on Innodb Incremental Backups
November 10, 2008 By 8 Comments
For normal Innodb “hot” backups we use LVM or other snapshot based technologies with pretty good success. However having incremental backups remain the problem. First why do you need incremental backups at all ? Why not just take the full backups daily. The answer is space – if you want to keep several generations to [...]
Recovery beyond data restore
August 2, 2008 By 11 Comments
Quite frequently I see customers looking at recovery as on ability to restore data from backup which can be far from being enough to restore the whole system to operating state, especially for complex systems. Instead of looking just at data restore process you better look at the whole process which is required to bring [...]
LVM Configuration mistake to Avoid
July 9, 2008 By 10 Comments
One of our customers was setting up box for MySQL LVM backups. The box had 2 RAID volumes (internal drives and enclosure) with database being stored on enclosure drives and internal drives used pretty much for OS and logs (database grew too large and so was moved to enclosure completely). As database was large and [...]
How SHOW SLAVE STATUS relates to CHANGE MASTER TO
July 7, 2008 By 21 Comments
As you probably know MySQL Replication (statement based) works by fetching statements from MASTERs binary log and executing them on the SLAVE. Since MySQL 4.0 this process is a bit more involved having events passing via relay logs on the Slave which also means there are two replication threads “IO Thread” and “SQL Thread” used [...]
Resyncing table on MySQL Slave
June 29, 2008 By 11 Comments
Sometimes MySQL Replication may run out of sync – because of its own buts or operational limitations or because of application mistake, such as writing to the slave when you should be only writing to the master. In any case you need slave to be synced with Master. To discover the difference between Master and [...]
Estimating Undo Space needed for LVM Snapshot
June 9, 2008 By 5 Comments
We know MySQL Backups using LVM are pretty cool (check out mylvmbackup) or MMM though it is quite typical LVM is not configurable properly to be usable for MySQL Backups. Quite frequently I find LVM installed on the system but no free space left to be used as snapshot undo space, which means LVM is [...]
MySQL Replication vs DRBD Battles
April 28, 2008 By 12 Comments
Well these days we see a lot of post for and against (more, more) using of MySQL and DRBD as a high availability practice. I personally think DRBD has its place but there are far more cases when other techniques would work much better for variety of reasons. First let me start with Florian’s comments [...]
Using MMM to ALTER huge tables
March 27, 2008 By 11 Comments
Few months ago, I wrote about a faster way to do certain table modifications online. It works well when all you want is to remove auto_increment or change ENUM values. When it comes to changes that really require table to be rebuilt – adding/dropping columns or indexes, changing data type, converting data to different character [...]
No more MySQL Crash Safe Replication in 5.0 ?
January 29, 2008 By 7 Comments
As you might know even if you’re only using Innodb tables your replication is not completely crash safe – if Slave MySQL Server crashes/power goes down it is likely for relay logs to run out of sync (they are not synced to the disk) plus position on the master which slave remembers becomes stale. During [...]
