Bug #64187 | Unclear Documentation for some mysql variables | ||
---|---|---|---|
Submitted: | 1 Feb 2012 0:00 | Modified: | 9 Mar 2013 1:44 |
Reporter: | Simon Mudd (OCA) | Email Updates: | |
Status: | Closed | Impact on me: | |
Category: | MySQL Server: Documentation | Severity: | S3 (Non-critical) |
Version: | All | OS: | Any |
Assigned to: | John Russell | CPU Architecture: | Any |
Tags: | windmill |
[1 Feb 2012 0:00]
Simon Mudd
[1 Feb 2012 14:46]
Valeriy Kravchuk
Looks like you are right. At least my code review attempts led me to the same conclusions: values of read_buffer_size and read_rnd_buffer_size hardly matter at all for InnoDB tables (non-partitioned ones at least).
[9 Feb 2012 4:05]
Jon Stephens
It's true that different engines employ this differently (or don't, in some cases), and that it has little or no bearing on InnoDB performance. We're still looking at how best to present this in the documentation.
[1 Feb 2013 20:02]
John Russell
I'm revisiting some doc bugs that got pushed down in priority due to 5.6 development work. Still thinking how best to accomplish this goal. It seems like having a big list of server configuration options, then elsewhere a separate list of InnoDB configuration options, gives the impression that whatever in the first list is of primary importance and the InnoDB-specific ones are less relevant. Which is not really true with InnoDB being the default table type. Would it make sense to fold the InnoDB options into the main list on http://dev.mysql.com/doc/refman/5.5/en/server-system-variables.html ? And/or to separate out MyISAM-specific options into a separate list?
[9 Mar 2013 1:44]
John Russell
Added details from Sinisa in support about the other engine-independent ways this variable is used. I'm closing the bug now because the rest of the suggestion, to clarify how each variable is or isn't intertwined with InnoDB, is an ongoing work item.