Обсуждение: Oracle vs. PostgreSQL web speed
Are there any benchmarks for Oracle vs. PostgreSQL primarily for SELECTs.
I'm trying to figure out which is faster for web page display (ASP, JSP, etc).
thank you!
Troy Campano
On Sat, May 11, 2002 at 12:21:04PM -0400, Troy.Campano@LibertyMutual.com wrote: > Are there any benchmarks for Oracle vs. PostgreSQL primarily for SELECTs. Oracle don't allow benchmarks to be published without their consent, it's in their EULA. > I'm trying to figure out which is faster for web page display (ASP, JSP, > etc). Since a relational database does not "display" web pages, I doubt you're going get very far. Cheers, Neil -- Neil Conway <neilconway@rogers.com> PGP Key ID: DB3C29FC
Sorry, I meant how fast the data comes back from the SELECT statement, not display in a page. -----Original Message----- From: nconway@klamath.dyndns.org [mailto:nconway@klamath.dyndns.org] Sent: Saturday, May 11, 2002 1:14 PM To: Troy.Campano@LibertyMutual.com Cc: pgsql-general@postgresql.org Subject: Re: [GENERAL] Oracle vs. PostgreSQL web speed On Sat, May 11, 2002 at 12:21:04PM -0400, Troy.Campano@LibertyMutual.com wrote: > Are there any benchmarks for Oracle vs. PostgreSQL primarily for > SELECTs. Oracle don't allow benchmarks to be published without their consent, it's in their EULA. > I'm trying to figure out which is faster for web page display (ASP, > JSP, etc). Since a relational database does not "display" web pages, I doubt you're going get very far. Cheers, Neil -- Neil Conway <neilconway@rogers.com> PGP Key ID: DB3C29FC
On Sat, 11 May 2002 Troy.Campano@LibertyMutual.com wrote: > Sorry, I meant how fast the data comes back from the SELECT statement, not > display in a page. As much as we all would want meaninfull comparisons, they are very difficult to produce. In particular you would need to have simmilar hardware to compare. Based on a wildly inaccurate/impressice mini test I did, it seems Oracle was faster with complex queries involving agregrates. The machines did not have the same amount of memory, nor the same type of disks, but the query was the same and the amount of rows and row sizes were very close. A better approach to what you want may be just to try postgreSQL with the type of queries and if it is fast enough use it. At least that test is a lot easier and much cheaper to do than to try with Oracle. In particular consider the cost of an individual doing the test. This is often overlooked. Even though I don't do much with Oracle based on the little I have done, I find PostgreSQL a lot easier to learn, use, configure and optimize. We use it were I work as a reporting server and it works very well.