[ADMIN] WAL files size
How can I limit the maximum size of WAL files? In my server there are now more than 80 WAL files and they use about 1.3 Gb Thanks, Jacopo ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://www.postgresql.org/search.mpl
[ADMIN] Re: Bugs in PostgreSQL v7.1 rpms for RedHat
- Original Message - From: "Lamar Owen" <[EMAIL PROTECTED]> To: "David Lizano" <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Thursday, April 19, 2001 7:29 PM Subject: Re: Bugs in PostgreSQL v7.1 rpms for RedHat > Works here, for some reason. But I'll check it. I run these very RPM's > on RH 6.2, 7.0, _and_ 7.1 and have not found this problem. But, I _will_ > double check, and will issue a patch that still uses pg_ctl accordingly. > (...) I agree, no problems on RH for me too. I agree with the default no networking conf and I think that with 7.1 there you do not need any more to set command line options ("-i"), now everything goes in the .conf file. I would rather set chkconfig to start a little bit later and to stop the service sooner in the shutdown, something like # chkconfig - 95 05 Then chkconfig --add could go in the rpm install script. Finally I did not find the new pg_restore tool for reloading databeses from new style dump archives. You can check with rpm -qil postgres-*|grep restore For this reason I had to recompile pgsql from scratch. I think it should go in the RPM packages. Thanks again to the RH packages maintainer. Jacopo ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
[ADMIN] Re: Bugs in PostgreSQL v7.1 rpms for RedHat
- Original Message - From: "Lamar Owen" <[EMAIL PROTECTED]> To: "David Lizano" <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Thursday, April 19, 2001 7:29 PM Subject: Re: Bugs in PostgreSQL v7.1 rpms for RedHat > Works here, for some reason. But I'll check it. I run these very RPM's > on RH 6.2, 7.0, _and_ 7.1 and have not found this problem. But, I _will_ > double check, and will issue a patch that still uses pg_ctl accordingly. > (...) I agree, no problems on RH for me too. I agree with the default no networking conf and I think that with 7.1 there you do not need any more to set command line options ("-i"), now everything goes in the .conf file. I would rather set chkconfig to start a little bit later and to stop the service sooner in the shutdown, something like # chkconfig - 95 05 Then chkconfig --add could go in the rpm install script. Finally I did not find the new pg_restore tool for reloading databeses from new style dump archives. You can check with rpm -qil postgres-*|grep restore For this reason I had to recompile pgsql from scratch. I think it should go in the RPM packages. Thanks again to the RH packages maintainer. Jacopo ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
[ADMIN] New release of Viruses...
We are even getting viruses from the ML... Is this a XMas present? Bye, and please, as long as you are using Windoze clients please do not open the attachment that our good friend [EMAIL PROTECTED] sent us :-) Jacopo
[ADMIN] PAM
Does anyone know of a PAM module that let you authenticate vs a pgsql table with e.g. username/password fields ? There is something similar for mysql but I would rather use my pgsql ... Jacopo S.
Re: [ADMIN] is a user in a group?
Thanks! For now I'll keep on doing searches on arrays, hoping for a new system table that manages in a better way the relation between user and groups. Bye, Jacopo - Original Message - From: "Peter Eisentraut" <[EMAIL PROTECTED]> To: "Jacopo Silva" <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Friday, November 10, 2000 5:29 PM Subject: Re: [ADMIN] is a user in a group? > Jacopo Silva writes: > > > There should be something more elegant than this!!! > > Correct. And soon we'll implement something. > > > There should be a system table which I can use to find > > the relation between user and groups. > > And that is how we'll do it. > > > I would be very grateful to anyone who can help me:-). > > Sorry, at the moment there's no bettter way. > > -- > Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/ >
[ADMIN] is a user in a group?
I need to find out a quick way to check which groups a user is member of. In the pg_group table each group row has an array of userids of the members of the group. At the moment I am selecting all the rows than searching in each array for the id of the user I am looking for. There should be something more elegant than this!!! There should be a system table which I can use to find the relation between user and groups. I would be very grateful to anyone who can help me:-). Bye, Jacopo
Re: [ADMIN] disable auto-commit
If are deleting something, then you probably want delete it... I mean that the DELETE command phisically removes some rows from a table, and after the command they are gone forever. If you want to delete something that can be restored later, then you are not actually deleting anything, you are just "hiding" from the user but still keeping in the db. So what I usually do in this cases is using a boolean field named "hide" . If the user choose to delete a row I actually put the "true" value in that field, if I want to restore I can turn back the field value to "false". Then I select rows from the table with the "WHERE NOT hide" clause whenever I don't want the user to see what he "deleted" . I use to default the field value to "false" so I can forget it and use it only when I want to hide something. e.g.: CREATE TABLE foo ( ... ... ... hide bool DEFAULT false); Commit does not mean much in this contest. You commit when you force the operating system to write in the disk any changes it is storing in memory buffers, when you commit you flush the cache buffers to the disk. By default postgresql commit to the disk whenever there is any change, you can speed up the system by disabling this feature starting postmaster with "-o -F" . Anyway even if you do not commit the changes are stored in the operating system disk cache and you cannot undo them. Bye, Jacopo - Original Message - From: "John McKown" <[EMAIL PROTECTED]> To: "Joel Pang" <[EMAIL PROTECTED]> Cc: <[EMAIL PROTECTED]> Sent: Thursday, October 26, 2000 9:30 PM Subject: Re: [ADMIN] disable auto-commit > Well, you could start a transaction. But the record will be deleted once > you do a COMMIT. Is that what you want? If you want some way to > "delete" an entry, but then to later "undelete" it, there really isn't a > way. I have heard of other systems where things were "logically deleted", > but recoverable until a "purge" command was done. PostgreSQL does not have > this type of behaviour. > > John > > On Wed, 25 Oct 2000, Joel Pang wrote: > > > In psql, once I delete an entry, it's gone permanently. How do I disable > > such auto-commit behavior? > > >
[ADMIN] Re: MS Access out-performs PostgreSQL 7?
Ok. So if it is clear that Access is not a database server while postgres is, speed is not so much a problem. I mean that you are right. You should anyway try optimizing your pgsql configuration (e.g. -S , -B, -F options). But still many tests will show that Access is faster in performing some operations: insert or select, etc. I tried similar tests months ago and I had more or less the same results. You can find benchmarks between many different database systems at http://www.mysql.com/crashme/choose.php (they compare features and they have some graphics of performances) Time ago, in this list, someone said that you have on one side of the databases systems the dbm libs: fastest straight access to data, no concurrency, no nothing... At the other side you have pgsql or similar: sometimes slower access to data, but many complex features. I think you should try to understand very well what use of the database your application will do. If you think that you have just few insertions and update of data and many reads with select, and no transactions go to a faster and less complex sql server. Many sites with similar use of data are using mysql that is much better (and faster) than Access, and is still a server system. Tucows and many other software download sites are using mysql. Download sites have sporadic (if compared to other operations, of course) updates and thousands of reads. If you need a more complex data management, if you plan to implement transactions, if you have many users inserting, updating and deleting data, and not only reading data most of times, choose pgsql. Bye, Jac P.S.: > Ok, once again let me make this clear... > > We are not thinking of using access for our dataserver, your right, its NOT > a data SERVER. My question simply was, why does a program like access, which > is not a data server, outperform pgsql in our comparison?? You want a simple answer? Because it is simpler and it doesn't have many of the pgsql features...
R: [ADMIN] users and passwords problem
- Original Message - From: Anthony E. Greene <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, July 13, 2000 4:06 PM Subject: Re: [ADMIN] users and passwords problem > Denis Pugnere wrote: > > > > PG 7.0.2, RH Linux 6.2 > > > > I'm trying to secure access to pgsql databases. > > the politic I use is to only allow access databases with passwords. > > (...) > > Why this access is allowed ? > > How to secure accesses to databases ? > > Have that "unauthorized" user try a SELECT and see what happens. They > are allowed to connect, but not to retrieve any records. Yes but they still can create new objects, e.g. "CREATE TABLE foo (id serial, desc varchar(10));" even if they aren't database owners. I think this is a known problem and I've heard it will be soon fixed. Bye, Jacopo
R: [ADMIN] POSTMASTER
Why don't you start reading one of the many docs in the doc dir? They are very complete, and you'll find everything in there :-) Jacopo S. ( Some other might have answered RTFM ... :-) ) - Original Message - From: Gagan <[EMAIL PROTECTED]> To: postgresql admin <[EMAIL PROTECTED]> Sent: Tuesday, June 06, 2000 10:46 PM Subject: [ADMIN] POSTMASTER > Respected sir/madam, > (..) > I will be obliged to get by mind out of these confusions. > > regards > Gagan
[ADMIN] Authentication via PAM
Hi. Does anyone knows of an existing patch to authenticate users connecting to the database via pam modules? Thanks, Jacopo