Обсуждение: PostgreSQL vs. MySQL vs. Oracle, 2005 report card
A recent Slashdot thread on MySQL performance (http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) contains a comment mentioning the following Fermilab report from May 2003: http://www-css.fnal.gov/dsg/external/freeware/mysql-vs-pgsql.html With the release of PostgreSQL 8.0, how does PostgreSQL currently stand? -- Marques Johansson marques@displague.com
On Friday 11 Feb 2005 6:07 pm, Marques Johansson wrote: > A recent Slashdot thread on MySQL performance > (http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) > contains a comment mentioning the following Fermilab report from May 2003: > > http://www-css.fnal.gov/dsg/external/freeware/mysql-vs-pgsql.html > > With the release of PostgreSQL 8.0, how does PostgreSQL currently stand? That seems to be fairly old and hence out-of-sync. Things have move a lot since, for all the products involved. But still I find it rather harsh on postgresql. 8K row limit is long gone and postgresql does support online backup, to mention a few. PostgreSQL now has native windows port, tablespaces and many many replication solutions, not to mention a list of companies that offer PostgreSQL support. Shridhar
Am Freitag, 11. Februar 2005 13:37 schrieb Marques Johansson: > A recent Slashdot thread on MySQL performance > (http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) > contains a comment mentioning the following Fermilab report from May 2003: They were apparently testing with PostgreSQL 6.3 at best, so I'd disregard this report. -- Peter Eisentraut http://developer.postgresql.org/~petere/
On Mon, 2005-02-14 at 09:28, Peter Eisentraut wrote: > Am Freitag, 11. Februar 2005 13:37 schrieb Marques Johansson: > > A recent Slashdot thread on MySQL performance > > (http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) > > contains a comment mentioning the following Fermilab report from May 2003: > > They were apparently testing with PostgreSQL 6.3 at best, so I'd disregard > this report. Coulda' been 6.4 even :-) I Wrote the contact person on the page a nice, well reasoned email explaining that many of the points were quite out of date, and giving him the skinny on how things are nowadays...
They also say that mysql is free, which isn't the case for commercial applications. On Mon, Feb 14, 2005 at 07:36:44PM +0530, Shridhar Daithankar wrote: > On Friday 11 Feb 2005 6:07 pm, Marques Johansson wrote: > > A recent Slashdot thread on MySQL performance > > (http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) > > contains a comment mentioning the following Fermilab report from May 2003: > > > > http://www-css.fnal.gov/dsg/external/freeware/mysql-vs-pgsql.html > > > > With the release of PostgreSQL 8.0, how does PostgreSQL currently stand? > > That seems to be fairly old and hence out-of-sync. Things have move a lot > since, for all the products involved. > > But still I find it rather harsh on postgresql. 8K row limit is long gone and > postgresql does support online backup, to mention a few. > > PostgreSQL now has native windows port, tablespaces and many many replication > solutions, not to mention a list of companies that offer PostgreSQL support. > > Shridhar > > ---------------------------(end of broadcast)--------------------------- > TIP 5: Have you checked our extensive FAQ? > > http://www.postgresql.org/docs/faq > -- Jim C. Nasby, Database Consultant decibel@decibel.org Give your computer some brain candy! www.distributed.net Team #1828 Windows: "Where do you want to go today?" Linux: "Where do you want to go tomorrow?" FreeBSD: "Are you guys coming, or what?"
Peter Eisentraut wrote: > Am Freitag, 11. Februar 2005 13:37 schrieb Marques Johansson: > >>A recent Slashdot thread on MySQL performance >>(http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) >>contains a comment mentioning the following Fermilab report from May 2003: > > > They were apparently testing with PostgreSQL 6.3 at best, so I'd disregard > this report. > Perhaps someone from advocacy could go through their list of issues and help them. They do seem to be trying to update it, and now have references to newer features: "Datafile location and a tablespace implementation as of version 8." However some of their criteria (savepoints:"no", incremental online backups:"no", and "access to multiple databases: no" (what about dblink)) still seem incorrect. Fermilab did have contact information ("mail comments to:") at the bottom of the page. Fermilab's a quite highly respected organization, so I think this page is probably trusted by many.
On Wed, 2005-02-16 at 11:13, Ron Mayer wrote: > Peter Eisentraut wrote: > > Am Freitag, 11. Februar 2005 13:37 schrieb Marques Johansson: > > > >>A recent Slashdot thread on MySQL performance > >>(http://developers.slashdot.org/article.pl?sid=05/02/11/038232&from=rss) > >>contains a comment mentioning the following Fermilab report from May 2003: > > > > > > They were apparently testing with PostgreSQL 6.3 at best, so I'd disregard > > this report. > > > > Perhaps someone from advocacy could go through their list of issues > and help them. They do seem to be trying to update it, and now > have references to newer features: "Datafile location and a > tablespace implementation as of version 8." > > > However some of their criteria (savepoints:"no", incremental > online backups:"no", and "access to multiple databases: no" > (what about dblink)) still seem incorrect. > > > Fermilab did have contact information ("mail comments to:") at the > bottom of the page. Fermilab's a quite highly respected organization, > so I think this page is probably trusted by many. Wandering about that page a bit myself, it seems most of this was written between the 6.5 and 7.2 versions of PostgreSQL. I don't think there's any active tendency towards misinformation, mostly it's just out of date.
Scott Marlowe <smarlowe@g2switchworks.com> writes:
> On Wed, 2005-02-16 at 11:13, Ron Mayer wrote:
>> Fermilab did have contact information ("mail comments to:") at the
>> bottom of the page. Fermilab's a quite highly respected organization,
>> so I think this page is probably trusted by many.
> Wandering about that page a bit myself, it seems most of this was
> written between the 6.5 and 7.2 versions of PostgreSQL. I don't think
> there's any active tendency towards misinformation, mostly it's just out
> of date.
Right. The question is whether we can get them to update it. The
odds are that the MySQL and Oracle information is just as dated; so
really the only fair way would be for them to redo the whole evaluation
from scratch, which I could well believe they don't have the time for.
It would be good to at least try to get them to label the page with
the versions that they tested, and note that the info is now quite
out of date.
regards, tom lane
Tom Lane wrote:
> Scott Marlowe <smarlowe@g2switchworks.com> writes:
>>On Wed, 2005-02-16 at 11:13, Ron Mayer wrote:
>>
>>>Fermilab did have contact information ("mail comments to:") at the
>...
> Right. The question is whether we can get them to update it.
I think so... as mentioned earlier in the thread, they did update
some of the answers saying "postgresql version 8 has this" already.
I just sent them an email with links to the postgresql docs for
each of the answers I know have been already addressed.
The email I sent them is attached below. Feel free to send them
corrections to my corrections. :)
-------- Original Message --------
Subject: Your database comparison page
Date: Wed, 16 Feb 2005 14:35:02 -0800
From: Ron Mayer [XXX]
To: css-dsg@fnal.gov
On your database comparison page here:
http://www-css.fnal.gov/dsg/external/freeware/mysql-vs-pgsql.html
where you write "Mail comments to:css-dsg@fnal.gov"
I think there are a few new updates in postgresql that you
might not have been aware of...
* Partial rollback of transaction:
...
Oracle8: Rollback to savepoint.
PostgreSQL:No.
Savepoints and Rollback to Savepoint have been added
to postgresql in version 8.
http://www.postgresql.org/docs/current/static/sql-savepoint.html
* Incremental and on-line backups:
...
Oracle8: Yes.
PostgreSQL: No.
Postgresql now supports a form of on-line backups by
archiving it's logs that contain only incremental changes
from the last full backup.
http://www.postgresql.org/docs/current/static/backup-online.html
* Access to multiple databases in one session:
....
Oracle8: Each session can use data from many instances, using
transparent distributed SQL access.
PostgreSQL: Only switching between databases. No support
for selecting data from different databases.
The option "dblink" module provides a mechanism for
making a query that spans databases in a single session.
An example is hidden in here.
http://www.postgresql.org/docs/current/static/queries-table-expressions.html
It's not as transparent as Oracle's, but it's quite
useful in some cases. Note that it can even link
to Microsoft SQL databases, and (i think) oracle.
* Subqueries in SQL query:
...
PostgreSQL:Sql compliance with Oracle.
Actually, I think PostgreSQL deviates from Oracle in places
where Oracle deviates from the SQL spec; so PostgreSQL
follows DB2 (and the SQL standard)'s way of handling
subqueries instead. An example, is that Oracle
doesn't like "AS ..." after a subquery.