>>>>> "Scott" == Scott Hess <scott@stripped> writes:
Scott> On Thu, Mar 09, 2000 at 03:17:06PM +0200, Michael Widenius wrote:
>> Monty wrote:
>> >> How big tables and keys do you have; isamchk is now that slow and the
>> >> OPTIMIZE TABLE in next MySQL version will be as fast as isamchk.
>> >>>>> "Scott" == Scott Hess <scott@stripped> writes:
Scott> [In case anyone is paying attention, this is the table I mentioned
Scott> a month or two ago, where it had three integer columns, a timestamp
Scott> column, and a blob column, where the blob columns was _always_ null,
Scott> and no other column was ever null. I had expected that the table
Scott> would never fragment in that case, but it still seemed to.]
>> Why do you have the blob column?
Scott> It's a holdover from a time when the blob column was stored in the
Scott> database. As usage rose, it became clear that the various tables with
Scott> blob columns were causing issues. We developed a system of storing
Scott> blobs in filesystem files which worked transparently under the
Scott> higher-level parts of the system. Unfortunately, since we support
Scott> sites with this custom change and sites without, removing the column
Scott> was considered somewhat of a risk.
Can't you just cange the BLOB column to 'CHAR(0)' ?
Scott> In any case, I was quite surprised to find out that the table was
Scott> still fragmenting. At that point, it was decided to move the entire
Scott> table out of MYSQL, which provided a significant performance
Scott> [Lest anyone take this wrong - we're pushing a lot of data through
Scott> the system, and while we're looking at other database servers, we
Scott> have strong reservations about whether any of them can satisfactorily
Scott> handle our needs. We're very happy with MYSQL in general, there
Scott> are just relatively tiny edge cases that I'm sniping at, here :-).]
>> By the way, is the table of type ISAM or MyISAM; You should get
>> much less fragmentation with MyISAM...
Scott> ISAM. Once 3.23 graduates to a beta release, we'll be testing it on
Scott> our staging machines... but probably not before.