Thread: Some basic info regarding postgresql?
Hi all, new to postgresql and WOW! IS it a great DB! I am blown away with it, it does far more than I thought it could.
I am interested in learning a bit more regarding a few areas of PostgreSQL. Either through replies, or links anyone can provide to more detailed information, I would be much appreciated.
I am looking to see how replication is done. I read that the eRServer or something that was once commercial was just recently open-sourced. Is this part of the 7.3.x or maybe the 7.4.x branch now? Or is it an add-on product, and how well does it handle the multiple ways of replication? In particular we would need data replication for a fail-over setup that would be done every 10 minutes or so, to ensure we have data that isn’t lost as much as it can be. What are some ways others are replicating data?
Clustering is of course another area we are concerned with. How well does PostgreSQL scale, how well does it handle heavy client load? We would probably either be using a dual Xeon 2.8Ghz system with 4GB ram, or an Itanium 2 system with 8GB ram as the DB server. We are looking to be able to handle a few 100 transactions per second in an “almost” 24/7 cycle… most of our clients don’t work between 10pm and 5am, but sometimes they do. How does it handle adding another DB to the cluster? How are some ways some of you are clustering and scaling with PostgreSQL?
Security is another biggie. I see that you can limit who can access it based on ip, user name, etc. What other levels of security are there? Can table permissions be set on an individual user, or a group of users? What about row permissions or column permissions? What other ways of security should we be concerned with in regards to the DB itself (deployed on linux of course)?
Lastly, performance. How well does it stack up for small and large/complex queries compared to Oracle, DB2 and MSSQL 2000? We are actually using all 3 of those DB’s at the request of some of our clients. How fast is it compared to mySQL these days? I read something from a year ago that said the performance gap is narrowing between pgsql and mysql, can anyone elaborate a bit? We are also looking at data warehousing possibilities. Is pgsql a good DB for these types of applications? Does it have any built-in support for it?
Would anyone be willing to provide a reference that our CIO/IT Manager may call to talk to if possible, to support the use of pgsql over the big boys?
Thanks.
*************************************************************
"The information contained in this e-mail message may be confidential and
protected from disclosure. If you are not the intended recipient, any
dissemination, distribution or copying is strictly prohibited. If you think
that you have received this e-mail message in error, please e-mail the
sender at exadmins@marketron.com."
*************************************************************
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.501 / Virus Database: 299 - Release Date: 7/14/2003
On Thu, Sep 11, 2003 at 16:18:04 -0700, "Duffey, Kevin" <KDuffey@marketron.com> wrote: > > Security is another biggie. I see that you can limit who can access it based on ip, user name, etc. What other levels ofsecurity are there? Can table permissions be set on an individual user, or a group of users? What about row permissionsor column permissions? What other ways of security should we be concerned with in regards to the DB itself (deployedon linux of course)? You can limit access to individual databases in a cluster by ip address or username (both postgres username and host username). This is done in pg_hba.conf. You can also limit access to objects by postgres username and group. This includes databases, schemas, tables, views, functions, sequences and probably a few other things. You can't directly restrict access to rows or columns of a table. You can create views that can only be used by some users or groups that restrict access to rows and or columns of a table.