Back to the month index |
Back to the list index
|
Gary Bickford (garyb@outlawnet.com)
Tue, 7 Jan 1997 12:37:30 -0800
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
- Next message: Gary Bickford: "[mSQL] Searching TEXT fields"
- Previous message: Thomas Baetzler: "Re: [mSQL] Automatic sequence?"
- In reply to: Mike Black: "[mSQL] Automatic sequence?"
- Next in thread: Gary Bickford: "Re: [mSQL] 2.0 - can it cache results?"
Message-Id: <v01540b07aef8583dce30@[208.192.38.36]> Date: Tue, 7 Jan 1997 12:37:30 -0800 From: garyb@outlawnet.com (Gary Bickford) Subject: Re: [mSQL] 2.0 - can it cache results?Gee, I'm glad my suggestion on table caching has invoked some heavy thinking!
Obviously both bambi & rasmus have a much better handle on this than I. I
agree with Rasmus' analysis, especially to the extent that the impact on
Bambi's effort is minimized.
Here's another thought that might be useful.
The temporary tables are already being created. This could be implemented
as a separate database, which contains all the temporary tables, plus an
extra "table of tables". This table would maintain the mapping between the
name passed and the actual table, and the scheduling per Rasmus, etc.
This could be handled by a separate process from msqldb, removing
responsiblility from msqldb entirely for persistent cached tables. msqldb
could create the result tables, hand them to this new process, which would
itself use msql to maintain the table of tables and delete orphans.
Then the primary (only?) changes to msqldb would be the additional syntax,
the passing of persistent tables to the new process. Access to the tables
would be two-step - one to acquire the actual table name from the new
process' table of tables, and then the actual access.
Of course the first access would also serve as a test of whether a table
exists - for instance if I always name a certain search result
'tbl-ore-cat', then I could search the table of tables for a matching
string. If it exists, I get back the real name of the table, and I use it.
When I do an update to a particular primary table, I can also delete the
cache table so it doesn't retain an incorrect cached result. In this case
I might look for all tables with the name 'tbl-ore%' so I can delete them.
gb
end
=======
Gary Bickford, FXT Corporation http://www.fxt.com
System integration, active web site design, intranets.
garyb@fxt.com 541-923-3060
--------------------------------------------------------------------------
To remove yourself from the Mini SQL mailing list send a message containing
"unsubscribe" to "unsubscribe" to msql-list-request@bunyip.com. Send a message containing
"info msql-list" to majordomo@bunyip.com for info on monthly archives of
the list. For more help, mail owner-msql-list@bunyip.com NOT the msql-list!
- Next message: Gary Bickford: "[mSQL] Searching TEXT fields"
- Previous message: Thomas Baetzler: "Re: [mSQL] Automatic sequence?"
- In reply to: Mike Black: "[mSQL] Automatic sequence?"
- Next in thread: Gary Bickford: "Re: [mSQL] 2.0 - can it cache results?"