Re: [WebDNA] To Commitdatabase or Not to Commitdatabase
This WebDNA talk-list message is from 2012
It keeps the original formatting.
numero = 108440
interpreted = N
texte = We did a test that i published no so long ago: If you commit databases to disk after modification, it will entirely = depend on the size of your databases and the frequency you write it to = disk: a standard 7200rpm drive is usually able to write 70MB/s. So, if = you get one new [append] every second and your database is a 100MB one, = then you might expect serious I/O disk activity and probably some = latency. The speed difference with "Only commit databases to disk when = instructed" will be a huge one. If the database is 2MB only, the = difference will be smaller. We did some tests that i published not long ago: Automatically commit databases to disk after modification = --------------------------------------------------------------------------= ------- WebDNA 7.0.628 iMac 2.4GHz, 3GB RAM Open a database (2 fields), append 10000 records.......................: = 5 s Only commit databases to disk when instructed = --------------------------------------------------------------------------= ------- WebDNA 7.0.628 iMac 2.4GHz, 3GB RAM Open a database (2 fields), append 10000 records.......................: = 0 s Note that this does not measure one single database write, but rather = 10,000 successive writes - chris On Feb 2, 2012, at 16:48, Govinda wrote: >> The reason for my question above was to figure out if my next move = should be enabling
or just taking on the task of doing a = after (more or less) all of my [append] or [replace] tags. >>=20 >> If I was to gain speed / performance by choosing one over the other = then that is what I am looking for. >=20 > I don't know. Maybe someone (WSC?) has clocked the difference. =20 > You might able to make the speed tests yourself, if you wanted. > I am guessing it is the same internal mechanism, and so no speed = difference.. (assuming you put a [commitdatabase...] after *every* = write operation). >=20 > Personally, if I wanted to be sure every write operation resulted in = an updated disk copy, I would not bother any more more than just turning = on the auto-commit feature. >=20 > -G >=20 > --------------------------------------------------------- > This message is sent to you because you are subscribed to > the mailing list . > To unsubscribe, E-mail to: > archives: http://mail.webdna.us/list/talk@webdna.us > Bug Reporting: support@webdna.us
Associated Messages, from the most recent to the oldest:
We did a test that i published no so long ago: If you commit databases to disk after modification, it will entirely = depend on the size of your databases and the frequency you write it to = disk: a standard 7200rpm drive is usually able to write 70MB/s. So, if = you get one new [append] every second and your database is a 100MB one, = then you might expect serious I/O disk activity and probably some = latency. The speed difference with "Only commit databases to disk when = instructed" will be a huge one. If the database is 2MB only, the = difference will be smaller. We did some tests that i published not long ago: Automatically commit databases to disk after modification = --------------------------------------------------------------------------= ------- WebDNA 7.0.628 iMac 2.4GHz, 3GB RAM Open a database (2 fields), append 10000 records.......................: = 5 s Only commit databases to disk when instructed = --------------------------------------------------------------------------= ------- WebDNA 7.0.628 iMac 2.4GHz, 3GB RAM Open a database (2 fields), append 10000 records.......................: = 0 s Note that this does not measure one single database write, but rather = 10,000 successive writes - chris On Feb 2, 2012, at 16:48, Govinda wrote: >> The reason for my question above was to figure out if my next move = should be enabling or just taking on the task of doing a = after (more or less) all of my [append] or [replace] tags. >>=20 >> If I was to gain speed / performance by choosing one over the other = then that is what I am looking for. >=20 > I don't know. Maybe someone (WSC?) has clocked the difference. =20 > You might able to make the speed tests yourself, if you wanted. > I am guessing it is the same internal mechanism, and so no speed = difference.. (assuming you put a [commitdatabase...] after *every* = write operation). >=20 > Personally, if I wanted to be sure every write operation resulted in = an updated disk copy, I would not bother any more more than just turning = on the auto-commit feature. >=20 > -G >=20 > --------------------------------------------------------- > This message is sent to you because you are subscribed to > the mailing list . > To unsubscribe, E-mail to: > archives: http://mail.webdna.us/list/talk@webdna.us > Bug Reporting: support@webdna.us
christophe.billiottet@webdna.us
DOWNLOAD WEBDNA NOW!
Top Articles:
Talk List
The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...
Related Readings:
Lastautonumber (2005)
Shownext never shows next...still (1997)
Running a store on BOTH http and https (1998)
RE: Supressing Error Messages (1998)
WebCat.acgi from /cgi-bin/? (1997)
Using Protection =) (2003)
Looking for PHP programmer (2007)
WCS Newbie question (1997)
2nd WebCatalog2 Feature Request (1996)
includes and cart numbers (1997)
WebCat2b12--[searchstring] bug (1997)
different show next (1997)
OFF-TOPIC...list serve software for newsletter? (1998)
Bug or syntax error on my part? (1997)
WebCat2b15MacPlugIn - [authenticate] not [protect] (1997)
converting dates in database (1998)
A Toughie (1998)
WCS Newbie question (1997)
Strange intermittent WebDNA problems (2008)
Re[2]: Editor? (2000)