Thread: weaknesses and strenghts of PG
I am researching an advocation paper on Postgres. Basically I would like to make a list of the most important strengths and weaknesses of Postgres compared to a couple of other major databases. The aim is to have a factual technical background for when advocating Postgres to potential users. Specifically I would like to compare: - PG - MySql - Oracle - MS SQL server - Firebird the facts I am looking for is about the core functionality and application use aspects, while the answers I am looking for, is for example whether - one db more or less suited for GIS or warehouse or ... - one db is not well suited for performance operations - if there are types of operations a database is not good at, or conversely, is extremely good at. No DB can be the best on all application types, so my focus is on figuring out where PG is better than other DBs, where it is not and where it is evenly matched. regards thomas
Thomas PostgreSQL is the only database supported by once:radix (www.oncetechnologies.com). While it doesn't directly affect the performance of the database, it is certainly a consideration for people who want a rapid application development environment for building and deploying sophisticated web-based commercial applications without all the specialist skills that comparable applications need. We have systems that have replaced conventional client-server applications spanning continents with thousands of transactions being processed each day. Is that relevant to your comparison? Regards Rob Napier On 7/2/09 9:05 PM, "Thomas Finneid" <tfinneid@fcon.no> wrote: > > I am researching an advocation paper on Postgres. Basically I would like > to make a list of the most important strengths and weaknesses of > Postgres compared to a couple of other major databases. > > The aim is to have a factual technical background for when advocating > Postgres to potential users. > > Specifically I would like to compare: > > - PG > - MySql > - Oracle > - MS SQL server > - Firebird > > the facts I am looking for is about the core functionality and > application use aspects, while > > the answers I am looking for, is for example whether > - one db more or less suited for GIS or warehouse or ... > - one db is not well suited for performance operations > - if there are types of operations a database is not good at, > or conversely, is extremely good at. > > No DB can be the best on all application types, so my focus is on > figuring out where PG is better than other DBs, where it is not and > where it is evenly matched. > > regards > > thomas
Thomas Finneid wrote: > > I am researching an advocation paper on Postgres. Basically I would like > to make a list of the most important strengths and weaknesses of > Postgres compared to a couple of other major databases. > > The aim is to have a factual technical background for when advocating > Postgres to potential users. My Tech Talk at HP last year had a bunch of this: https://fossbazaar.org/content/josh-berkus-two-great-open-source-databases-comparison-2008-06-26 I will say overall, that a *real* comparison of appropriateness of various databases will have to be painfully detailed. I was asked at Sun to do a breakdown of "when should we recommend" just for Postgres, MySQL, Derby and Oracle; I wouldn't want to do the whole field. And don't forget that there's a whole new ball game of specialty databases these days, including (but not limited to): DW/BI databases: Greenplum Netezza Aster Paraccel LucidDB etc. Object/Multivalue DBs: DB4O Cache CouchDB Embedded DBs HSQLDB Derby SQLite Others Hypertable Memcached Also, when you compare "MySQL" you have to treat each storage engine really as a separate database for comparison purposes; they don't behave the same, and usually migration between table types is very difficult. I hope you have a year for this project! --Josh Berkus
Thomas Finneid wrote: > > I am researching an advocation paper on Postgres. Basically I would > like to make a list of the most important strengths and weaknesses of > Postgres compared to a couple of other major databases. > > The aim is to have a factual technical background for when advocating > Postgres to potential users. We've done comparisons against MySQL and SQL Server (both for regular use and spatial) http://www.postgresonline.com/journal/index.php?/archives/51-Cross-Compare-o f-SQL-Server,-MySQL,-and-PostgreSQL.html http://www.postgresonline.com/journal/index.php?/archives/51-Cross-Compare-o f-SQL-Server,-MySQL,-and-PostgreSQL.html Simon Greener has started one too -- he said he'd try to compensate for our lack of Oracle and DBII mention when he has time. His is more of a spatial compare though http://www.spatialdbadvisor.com/SpatialDatabaseComparison/99/spatial-databas e-comparison Hope that helps, Regina
Thanks for the tip. thomas Rob Napier wrote: > Thomas > > PostgreSQL is the only database supported by once:radix > (www.oncetechnologies.com). > > While it doesn't directly affect the performance of the database, it is > certainly a consideration for people who want a rapid application > development environment for building and deploying sophisticated web-based > commercial applications without all the specialist skills that comparable > applications need. > > We have systems that have replaced conventional client-server applications > spanning continents with thousands of transactions being processed each day. > > Is that relevant to your comparison? > > Regards > > Rob Napier > > On 7/2/09 9:05 PM, "Thomas Finneid" <tfinneid@fcon.no> wrote: > >> I am researching an advocation paper on Postgres. Basically I would like >> to make a list of the most important strengths and weaknesses of >> Postgres compared to a couple of other major databases. >> >> The aim is to have a factual technical background for when advocating >> Postgres to potential users. >> >> Specifically I would like to compare: >> >> - PG >> - MySql >> - Oracle >> - MS SQL server >> - Firebird >> >> the facts I am looking for is about the core functionality and >> application use aspects, while >> >> the answers I am looking for, is for example whether >> - one db more or less suited for GIS or warehouse or ... >> - one db is not well suited for performance operations >> - if there are types of operations a database is not good at, >> or conversely, is extremely good at. >> >> No DB can be the best on all application types, so my focus is on >> figuring out where PG is better than other DBs, where it is not and >> where it is evenly matched. >> >> regards >> >> thomas > > >
Thanks for the info, I will have a look at it. I dont have a specified project plan or date this paper. I just wont to gather some knowledge on the subject and write down what I find so I am prepared whenever I end up in a discussion. Regarding all the specialty DB's, some are just novelty dbs, sort of like MS Access, some could be useful as a small application internal transactional storage and some are too specialised to be of interest (ref BI db's) The problem with many of them are that they are written in Java, java isn't suitable for the kind of performance, stability and resource control a proper database server needs. For that one needs to look at real DB's, such as PG, MySql, Oracle etc. There is this guy at work who made a system which uses Derby, basically it has problems. thomas Josh Berkus wrote: > Thomas Finneid wrote: >> >> I am researching an advocation paper on Postgres. Basically I would >> like to make a list of the most important strengths and weaknesses of >> Postgres compared to a couple of other major databases. >> >> The aim is to have a factual technical background for when advocating >> Postgres to potential users. > > My Tech Talk at HP last year had a bunch of this: > > https://fossbazaar.org/content/josh-berkus-two-great-open-source-databases-comparison-2008-06-26 > > > I will say overall, that a *real* comparison of appropriateness of > various databases will have to be painfully detailed. I was asked at > Sun to do a breakdown of "when should we recommend" just for Postgres, > MySQL, Derby and Oracle; I wouldn't want to do the whole field. And > don't forget that there's a whole new ball game of specialty databases > these days, including (but not limited to): > > DW/BI databases: > Greenplum > Netezza > Aster > Paraccel > LucidDB > etc. > > Object/Multivalue DBs: > DB4O > Cache > CouchDB > > Embedded DBs > HSQLDB > Derby > SQLite > > Others > Hypertable > Memcached > > Also, when you compare "MySQL" you have to treat each storage engine > really as a separate database for comparison purposes; they don't behave > the same, and usually migration between table types is very difficult. > > I hope you have a year for this project! > > --Josh Berkus >
Thanks, will have a look at it. thomas Paragon Corporation wrote: > > Thomas Finneid wrote: >> I am researching an advocation paper on Postgres. Basically I would >> like to make a list of the most important strengths and weaknesses of >> Postgres compared to a couple of other major databases. >> >> The aim is to have a factual technical background for when advocating >> Postgres to potential users. > > We've done comparisons against MySQL and SQL Server (both for regular use > and spatial) > > http://www.postgresonline.com/journal/index.php?/archives/51-Cross-Compare-o > f-SQL-Server,-MySQL,-and-PostgreSQL.html > > http://www.postgresonline.com/journal/index.php?/archives/51-Cross-Compare-o > f-SQL-Server,-MySQL,-and-PostgreSQL.html > > Simon Greener has started one too -- he said he'd try to compensate for our > lack of Oracle and DBII mention when he has time. His is more of a spatial > compare though > > http://www.spatialdbadvisor.com/SpatialDatabaseComparison/99/spatial-databas > e-comparison > > Hope that helps, > Regina > > >