Обсуждение: Pg database, need a solution to a problem

I'm running Postgres on a Redhat Linux 9 server to
keep track of testcases we run in our lab.  The
interface to add new testcases to the DB is cumbersome
for adding large groups, well actually it isn't
possible really.  So on the development Linux machine,
when someone writes a new test case and adds it, it is
a very manual process.  When someone creates a suite
of testcases to run, again that process is manual in
that they have to select all the testcases to be in
the suite.

We have a "production" Linux server in our lab that
runs the completed test cases and suites.  It has an
entirely separate Postgres database from the
development server, but both have the same schema,
etc.  Obviously they have very different data sets in
the results tables because people are running tests at
different times, on different test cases and suites in
"production" vs the "development" test box.  But since
adding testcases is such a manual process, it really
sucks whenever new testcases are written to have to
get them into the "production" tables that have the
test case names, id's, etc.  What I have been doing is
doing a pg_dump of all the data in the dev database
and simply reloading it to the production Pg database.
 This sucks though in that all the results for the
production box are lost and so are any configuration
settings we have.  Essentially it makes a dump copy of
everything at its current state in development.

Does anyone have any suggestions for how to get the
testcase id's and suite's replicated to the production
database with out touching the other tables in there?
I would just do a pg_dump on selected tables, but
there are sequences in the DB for adding new suites,
and testcases and I don't want these getting out of
sync.  Or can I copy the sequence id's as well so they
are updated on the production Pg DB?

Thanks

> Does anyone have any suggestions for how to get the
> testcase id's and suite's replicated to the production
> database with out touching the other tables in there?
> I would just do a pg_dump on selected tables, but
> there are sequences in the DB for adding new suites,
> and testcases and I don't want these getting out of
> sync.  Or can I copy the sequence id's as well so they
> are updated on the production Pg DB?

You could run replication for just the tables you need.
Check out www.slony.info or www.commandprompt.com

Sincerely,

Joshua D. Drake



>
> Thanks
>
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 1: subscribe and unsubscribe commands go to majordomo@postgresql.org


--
Command Prompt, Inc., your source for PostgreSQL replication,
professional support, programming, managed services, shared
and dedicated hosting. Home of the Open Source Projects plPHP,
plPerlNG, pgManage,  and pgPHPtoolkit.
Contact us now at: +1-503-667-4564 - http://www.commandprompt.com

On Friday 28 January 2005 15:42, Joshua D. Drake wrote:
> > Does anyone have any suggestions for how to get the
> > testcase id's and suite's replicated to the production
> > database with out touching the other tables in there?
> > I would just do a pg_dump on selected tables, but
> > there are sequences in the DB for adding new suites,
> > and testcases and I don't want these getting out of
> > sync.  Or can I copy the sequence id's as well so they
> > are updated on the production Pg DB?
>
> You could run replication for just the tables you need.
> Check out www.slony.info or www.commandprompt.com
>

Since you may not want to have real time replication going (don't push up new
test cases until they make it out of test), I wonder if something like Robert
Kaye's lazy replication system he uses for music brainz might be more
applicable.  I can't recall if his OSCon talk is linked from techdocs, but it
should be available on the OSCon 2004 site.  Google on some of the above
words to find out more info.

--
Robert Treat
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL

On Sun, 30 Jan 2005 11:56:04 -0500, Robert Treat
<xzilla@users.sourceforge.net> wrote:
> On Friday 28 January 2005 15:42, Joshua D. Drake wrote:
> > > Does anyone have any suggestions for how to get the
> > > testcase id's and suite's replicated to the production
> > > database with out touching the other tables in there?
> > > I would just do a pg_dump on selected tables, but
> > > there are sequences in the DB for adding new suites,
> > > and testcases and I don't want these getting out of
> > > sync.  Or can I copy the sequence id's as well so they
> > > are updated on the production Pg DB?
> >
> > You could run replication for just the tables you need.
> > Check out www.slony.info or www.commandprompt.com
> >
>
> Since you may not want to have real time replication going (don't push up new
> test cases until they make it out of test), I wonder if something like Robert
> Kaye's lazy replication system he uses for music brainz might be more
> applicable.  I can't recall if his OSCon talk is linked from techdocs, but it
> should be available on the OSCon 2004 site.  Google on some of the above
> words to find out more info.

http://conferences.oreillynet.com/cs/os2004/view/e_sess/5480

Unfortunately, the actual


--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
L. Friedman                                    netllama@gmail.com
LlamaLand                       http://netllama.linux-sxs.org