Comments on: Abusing MySQL http://www.bofh-hunter.com/2008/05/08/abusing-mysql/ Admin Spotting for Fun and Profit Fri, 16 Jul 2010 09:41:46 +0000 hourly 1 http://wordpress.org/?v=3.0 By: Jim Perrin http://www.bofh-hunter.com/2008/05/08/abusing-mysql/comment-page-1/#comment-55 Jim Perrin Sun, 18 May 2008 12:31:04 +0000 http://www.bofh-hunter.com/?p=25#comment-55 No apology needed, I was simply curious and wondering if you'd follow up. I know that I can come across as hostile to folks who mean well, and so the apology should be mine. I would very much be interested in seeing ZRM support maatkit's parallel backups, or even improving upon them if possible. Thank you for your follow up, and I'll try to not retaliate against folks who actually stop by here. :-) No apology needed, I was simply curious and wondering if you’d follow up. I know that I can come across as hostile to folks who mean well, and so the apology should be mine. I would very much be interested in seeing ZRM support maatkit’s parallel backups, or even improving upon them if possible. Thank you for your follow up, and I’ll try to not retaliate against folks who actually stop by here. :-)

]]> By: Ken http://www.bofh-hunter.com/2008/05/08/abusing-mysql/comment-page-1/#comment-54 Ken Sun, 18 May 2008 07:14:00 +0000 http://www.bofh-hunter.com/?p=25#comment-54 Until I read your blog post, I was not aware of Maatkit (and mk-find is most definitely cool!) ZRM, at this time, does not support using mk-parallel-dump to perform its dumps. That's not to say that it can't or won't. We just released a new version that uses InnoHotCopy so we're not opposed to adding support for more backup methods. The focus of the ZRM tool itself is the manageability of MySQL backups, independent of the underlying backup mechanism(s). Using mk-parallel-dump seems to be one of the most performant backup solutions out there, and so support for it would definitely be a plus. I've posed your question to the rest of our engineers to see what their thoughts are. I will follow-up when I hear more from them. If you're not interested, I apologize if I've wasted your time. (FWIW, no astroturfing was intended, which is why I linked to the community version, versus the paid version.) Until I read your blog post, I was not aware of Maatkit (and mk-find is most definitely cool!)

ZRM, at this time, does not support using mk-parallel-dump to perform its dumps. That’s not to say that it can’t or won’t. We just released a new version that uses InnoHotCopy so we’re not opposed to adding support for more backup methods. The focus of the ZRM tool itself is the manageability of MySQL backups, independent of the underlying backup mechanism(s). Using mk-parallel-dump seems to be one of the most performant backup solutions out there, and so support for it would definitely be a plus. I’ve posed your question to the rest of our engineers to see what their thoughts are. I will follow-up when I hear more from them.

If you’re not interested, I apologize if I’ve wasted your time. (FWIW, no astroturfing was intended, which is why I linked to the community version, versus the paid version.)

]]> By: Jim Perrin http://www.bofh-hunter.com/2008/05/08/abusing-mysql/comment-page-1/#comment-53 Jim Perrin Thu, 15 May 2008 00:49:08 +0000 http://www.bofh-hunter.com/?p=25#comment-53 Initially I considered this comment an astroturfing incident, and was going to block it, but instead I want to give this post the attention it deserves. Zmanda does do good things for mysql backups, there's no question about that; however from their own documentation, it uses mysqldump, hotcopy, or snapshotting depending on configuration. That it uses mysqldump is the part I want to focus on here, and invite Ken to reply. Will it support maatkit's mk-parallel-dump, or will it be restricted to mysqldump proper? Using mk-parallel-dump is certainly beneficial for faster backups as well as table level archives instead of the more common 'one file for all' approach. Ken, can Zmanda use mk-parallel-dump (since I spent this post expounding upon its virtues) or were you simply driving by and slapping up a sticker for your project? Initially I considered this comment an astroturfing incident, and was going to block it, but instead I want to give this post the attention it deserves. Zmanda does do good things for mysql backups, there’s no question about that; however from their own documentation, it uses mysqldump, hotcopy, or snapshotting depending on configuration.

That it uses mysqldump is the part I want to focus on here, and invite Ken to reply. Will it support maatkit’s mk-parallel-dump, or will it be restricted to mysqldump proper?

Using mk-parallel-dump is certainly beneficial for faster backups as well as table level archives instead of the more common ‘one file for all’ approach.

Ken, can Zmanda use mk-parallel-dump (since I spent this post expounding upon its virtues) or were you simply driving by and slapping up a sticker for your project?

]]> By: Ken Crandall http://www.bofh-hunter.com/2008/05/08/abusing-mysql/comment-page-1/#comment-52 Ken Crandall Wed, 14 May 2008 23:42:29 +0000 http://www.bofh-hunter.com/?p=25#comment-52 If you're interested in good, fast MySQL backup, check-out the Zmanda Recovery Manager (ZRM) at http://www.zmanda.com/backup-mysql.html If you’re interested in good, fast MySQL backup, check-out the Zmanda Recovery Manager (ZRM) at http://www.zmanda.com/backup-mysql.html

]]>