Comments on: Read Buffers, mmap, malloc and MySQL Performance http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/ Percona's MySQL & InnoDB performance and scalability blog Sat, 11 Feb 2012 16:45:54 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Project 2061 Techlog » Optimizing MySQL Server Runtime Parameters http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-296422 Project 2061 Techlog » Optimizing MySQL Server Runtime Parameters Wed, 07 May 2008 18:55:41 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-296422 [...] MySQL Performance Blog: Read Buffers, mmap, malloc and MySQL Performance [...] [...] MySQL Performance Blog: Read Buffers, mmap, malloc and MySQL Performance [...]

]]> By: 记录/分享/…… » Blog Archive » A Few Common Performance Mistakes http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-249464 记录/分享/…… » Blog Archive » A Few Common Performance Mistakes Fri, 07 Mar 2008 01:35:45 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-249464 [...] http://mysql-ha.com/2007/09/06/read-buffer-performance-hit/ http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/ [...] [...] http://mysql-ha.com/2007/09/06/read-buffer-performance-hit/ http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/ [...]

]]> By: Big DBA Head! - Database Brain Power! » A Few Common Performance Mistakes http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-249315 Big DBA Head! - Database Brain Power! » A Few Common Performance Mistakes Thu, 06 Mar 2008 16:19:51 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-249315 [...] http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/ [...] [...] http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/ [...]

]]> By: Apachez http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-170463 Apachez Fri, 21 Sep 2007 05:36:05 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-170463 By the way, how does other DBMS handle memory allocation? For example postgre, do they even have this many memory options available to screw things up? By the way, how does other DBMS handle memory allocation?

For example postgre, do they even have this many memory options available to screw things up?

]]> By: peter http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-167071 peter Thu, 13 Sep 2007 11:01:19 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-167071 Interesting. So if we're having multi value inserts (not insert select) - which buffer is allocated for writes ? Or are writes to MYD file performed row by row buffered by OS only ? In any case I do not insist on something and I would not be surprised if there is something unusual in Monty's case and it would be great to get the repeatable bug for it Interesting. So if we’re having multi value inserts (not insert select) – which buffer is allocated for writes ?
Or are writes to MYD file performed row by row buffered by OS only ?

In any case I do not insist on something and I would not be surprised if there is something unusual in Monty’s case
and it would be great to get the repeatable bug for it

]]> By: Sergei Golubchik http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-167062 Sergei Golubchik Thu, 13 Sep 2007 10:38:31 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-167062 Insert - yes, the buffer of that size is used for bulk inserts. INSERT ... SELECT for example would probably do it. But not simple single-value inserts. In my tests multi-value inserts didn't trigger such an allocation either, but I could believe they would under different circumstances. Although I didn't try blobs, I don't think read_buffer_size would be allocated for single value inserts with blobs. Insert – yes, the buffer of that size is used for bulk inserts. INSERT … SELECT for example would probably do it. But not simple single-value inserts. In my tests multi-value inserts didn’t trigger such an allocation either, but I could believe they would under different circumstances.

Although I didn’t try blobs, I don’t think read_buffer_size would be allocated for single value inserts with blobs.

]]> By: peter http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-167044 peter Thu, 13 Sep 2007 09:46:37 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-167044 Thanks Sergey, I would expect this not to be that simple. What is about inserts ? Can it get write buffers allocated in some conditions, like inserting blobs etc. Thanks Sergey,

I would expect this not to be that simple.

What is about inserts ? Can it get write buffers allocated in some conditions, like inserting blobs etc.

]]> By: Sergei Golubchik http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/comment-page-1/#comment-167015 Sergei Golubchik Thu, 13 Sep 2007 07:47:47 +0000 http://www.mysqlperformanceblog.com/2007/09/12/read-buffers-mmap-malloc-and-mysql-performance/#comment-167015 I didn't see any memory allocations of read_buffer_size happening on connect time (with breakpoints on all these places and connecting/disconnecting a few times) I didn’t see any memory allocations of read_buffer_size happening on connect time
(with breakpoints on all these places and connecting/disconnecting a few times)

]]>