On Wed, May 14, 2008 at 3:40 PM, Jay A. Kreibich <[EMAIL PROTECTED]> wrote:
>
>
> WORKAROUNDS:
>
> Set SQLites page size to be much larger (PRAGMA page_size).
> Makes SQLite deal with bigger chunks of data, reducing the overhead
> percentage. You'll very likely need to turn this up pretty high
> to see significant changes.
>
> Set SQLites page cache to be much larger (PRAGMA cache_size)
> Reduces the number of I/O operations. Great for lookups and sorts.
> Not that useful for writes. Depends a lot on how you use the DB.
>
> Live dangerously and turn down/off disk syncing (PRAGMA synchronous).
> Reduces the delay for writes. Dangerous.
>
> Or, brute force: Copy the file locally, do your stuff, copy it back.
>
Thank you all for your suggestions and explanations. I now understand
better the complexity underlying networked volumes.
I tried:
PRAGMA page_size = SQLITE_MAX_PAGE_SIZE
PRAGMA cache_size = 1000000
PRAGMA synchronous = OFF
(all executed before creating any tables)
There was no improvement in first read/write performance at all.
Looks like the brute force solution is the only answer here.
Cheers,
Peter.
--
---------------------------------------------
Peter K. Stys, MD
Dept. of Clinical Neurosciences
Hotchkiss Brain Institute
University of Calgary
tel (403) 210-8646
---------------------------------------------
_______________________________________________
sqlite-users mailing list
[email protected]
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
