Comments on: SSD, XFS, LVM, fsync, write cache, barrier and lost transactions http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/ Everything about MySQL Performance Sat, 21 Nov 2009 05:23:57 -0800 http://wordpress.org/?v=2.8.4 hourly 1 By: Baron Schwartz http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-611795 Baron Schwartz Fri, 10 Jul 2009 01:41:23 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-611795 Peter [Eisentraut], I've been following your benchmark blog posts with a lot of interest, too :-) Peter [Eisentraut], I’ve been following your benchmark blog posts with a lot of interest, too :-)

]]> By: http://petereisentraut.blogspot.com/ http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-611212 http://petereisentraut.blogspot.com/ Thu, 09 Jul 2009 14:57:06 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-611212 I meant http://petereisentraut.blogspot.com/2009/07/solid-state-drive-benchmarks-and-write.html ... funny markup ... I meant http://petereisentraut.blogspot.com/2009/07/solid-state-drive-benchmarks-and-write.html … funny markup …

]]> By: http://petereisentraut.blogspot.com/ http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-611211 http://petereisentraut.blogspot.com/ Thu, 09 Jul 2009 14:55:52 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-611211 Somewhat after the fact I realized that this information relates closely with the SSD tests I did with PostgreSQL, reported here: . With a plain dd test I get about 50% performance loss when turning the write cache off on the X25-E. That's much better than the more than fourfold loss you are observing (if I parse your numbers right). I'm planning to do a pgbench test later, which might provide more insight. Somewhat after the fact I realized that this information relates closely with the SSD tests I did with PostgreSQL, reported here: . With a plain dd test I get about 50% performance loss when turning the write cache off on the X25-E. That’s much better than the more than fourfold loss you are observing (if I parse your numbers right). I’m planning to do a pgbench test later, which might provide more insight.

]]> By: Baron Schwartz http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-530886 Baron Schwartz Mon, 06 Apr 2009 12:21:55 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-530886 Oh, and the broad consensus in the room was "these things are worth a lot less if they lie to the operating system about durability, and a UPS is not an acceptable workaround" :-) Oh, and the broad consensus in the room was “these things are worth a lot less if they lie to the operating system about durability, and a UPS is not an acceptable workaround” :-)

]]> By: Baron Schwartz http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-530885 Baron Schwartz Mon, 06 Apr 2009 12:20:31 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-530885 Jignesh presented on a similar topic at the PostgreSQL conference this weekend. Slides are at http://blogs.sun.com/jkshah/entry/effects_of_flash_ssd_on Jignesh presented on a similar topic at the PostgreSQL conference this weekend. Slides are at http://blogs.sun.com/jkshah/entry/effects_of_flash_ssd_on

]]> By: B http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-523661 B Sun, 29 Mar 2009 23:37:42 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-523661 True, but even lower cost servers come with redundant power supplies as an option. I certainly wouldn't specify a server for mission critical database data without redundant PSU's. B True, but even lower cost servers come with redundant power supplies as an option. I certainly wouldn’t specify a server for mission critical database data without redundant PSU’s.

B

]]> By: Baron Schwartz http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-523261 Baron Schwartz Sun, 29 Mar 2009 13:35:01 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-523261 A UPS isn't the whole solution. What happens when your power supply (the one inside the server) fails, for example? "Keep the power from going off" and "keep the server from crashing" are not the same thing as "I want this device not to lie when I ask for this data to be written to durable storage." A UPS isn’t the whole solution. What happens when your power supply (the one inside the server) fails, for example? “Keep the power from going off” and “keep the server from crashing” are not the same thing as “I want this device not to lie when I ask for this data to be written to durable storage.”

]]> By: Mark Callaghan http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-522493 Mark Callaghan Sun, 29 Mar 2009 01:22:48 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-522493 When will Amazon provide a UPS on EC2 servers? When will Amazon provide a UPS on EC2 servers?

]]> By: B http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-522416 B Sat, 28 Mar 2009 22:09:45 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-522416 @ Justmy2cents, I totally agree. Who isn't running UPS's on their DB servers? Even an old one that could hold the system on for 30 seconds would be enough, as long as further transactions didn't keep coming in during that time. B @ Justmy2cents,

I totally agree. Who isn’t running UPS’s on their DB servers? Even an old one that could hold the system on for 30 seconds would be enough, as long as further transactions didn’t keep coming in during that time.

B

]]> By: justmy2cents http://www.mysqlperformanceblog.com/2009/03/02/ssd-xfs-lvm-fsync-write-cache-barrier-and-lost-transactions/comment-page-1/#comment-514266 justmy2cents Sat, 21 Mar 2009 10:33:53 +0000 http://www.mysqlperformanceblog.com/?p=631#comment-514266 Why not just adding an UPS to ur system. Why not just adding an UPS to ur system.

]]>