Need "DB_PORT" in Makefile and support throughout build

We have our PostgreSQL database running on a non-default port.

Is there a syntax for DB_HOST in the Makefile that will allow me to
specify a different port? I tried localhost:OURPORT and that did not
work.

Please add support for setting and using a DB_PORT in the Makefile

We have our PostgreSQL database running on a non-default port.

This might seem a silly question, but why?

Alan Charlton

We have our PostgreSQL database running on a non-default port.

This might seem a silly question, but why?

Alan Charlton

Not so much silly as irrelevant and somewhat indicative of the thought
that having a database server on a non-default port is “Wrong.”

If it were wrong, there wouldn’t be an option to run the database
server on another port.

It’s on another port because we’re running an older version of PostgreSQL
on the default port.

Defualt port: 7.0.2, been working fine for our other existing
databases, don’t TOUCH IT.

Non-default port: 7.1.2, requirement of installing RT 2.x, in testing.

“JB” == Jeff Blaine jblaine@mitre.org writes:

JB> It’s on another port because we’re running an older version of PostgreSQL
JB> on the default port.

|+ This might seem a silly question, but why?

Some of us run multiple database servers on the same machine for various
reasons. Testing a new version of a database, testing software (you
wouldn’t want to be doing testing on the production database), etc.

While many of you may say ‘why not put it on another machine’. This is
not always possible.

Sounds like a perfectly reasonable request to me :wink:
There’s at least one other postgres-related option I need to roll in.
This should happen for 2.0.6 or 2.0.7.

    -jOn Wed, Aug 15, 2001 at 02:39:47PM -0400, Vivek Khera wrote:

“JB” == Jeff Blaine jblaine@mitre.org writes:

JB> It’s on another port because we’re running an older version of PostgreSQL
JB> on the default port.

From a cursory scan of the initdb script, my guess is that if you
specify DB_HOST=localhost;port=NNNN in the Makefile, it should trick
it into using the correct DBI incantation to connect to port NNNN
instead of the default. You will of course have to try it to find
out, and double check the results after config.pm is generated.


rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

jesse reed vincent – root@eruditorum.orgjesse@fsck.com
70EBAC90: 2A07 FC22 7DB4 42C1 9D71 0108 41A3 3FB3 70EB AC90

that’s security the same way that asking for directions to topeka and
being told that a seal is a mammal is informative
-robin@apocalypse.org

FWIW, the ‘DB_HOST = localhost;port=5433’ trick does not work.

% make dropdb

Dropping Pg database rt2.
DBI->connect(dbname=template1;host=localhost;port=5433) failed:
connectDBStart() – connect() failed: Connection refused at
/rcf/rt/lib/perl5/site_perl/5.6.1/DBIx/DataSource/Driver.pm line 74
connectDBStart() – connect() failed: Connection refused at
/rcf/rt/lib/perl5/site_perl/5.6.1/DBIx/DataSource/Driver.pm line 74,
line 2.
% psql -p 5433 -l
List of databases
Database | Owner
template0 | postgres
template1 | postgres
(2 rows)

That’s exactly what I would say.
If a test situation is considered viral enough to affect a working software
config, the assumption should be made, in order to protect the production
config, that the test may in fact affect the hardware as well, therefore, in
our shop, all tests get their own box.-----Original Message-----
From: rt-users-admin@lists.fsck.com
[mailto:rt-users-admin@lists.fsck.com]On Behalf Of Darrin Walton
Sent: Wednesday, August 15, 2001 3:00 PM
To: Alan Charlton
Cc: jblaine@mitre.org; rt-users@lists.fsck.com
Subject: Re: [rt-users] Need “DB_PORT” in Makefile and support
throughout build

|+ This might seem a silly question, but why?

Some of us run multiple database servers on the same machine for various
reasons. Testing a new version of a database, testing software (you
wouldn’t want to be doing testing on the production database), etc.

While many of you may say ‘why not put it on another machine’. This is
not always possible.

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

It’s on another port because we’re running an older version of
PostgreSQL on the default port.

Fair enough…

Cheers,
Alan

|+ That’s exactly what I would say.
|+ If a test situation is considered viral enough to affect a working software
|+ config, the assumption should be made, in order to protect the production
|+ config, that the test may in fact affect the hardware as well, therefore, in
|+ our shop, all tests get their own box.

Let me present this situation to you.

You have to bring up a circuit (say OC48). Bringing up the circuit to
test could, in some way crash the router. (router vendor not important,
it could be either cisco or juniper).

Now, bringing the circuit up on a ‘test’ router, if you have one available
for both sides of the connections (which let me tell you, gets quite
expensive, usually test equipment is kept in a lab somewhere). You test
the circuit, it is clean. Now, you still have to put it on the production
router. Putting the circuit on the production router could crash the
router. What do you do?

Your solution is not feesable for every situation.

I’d like to point out that this discussion is getting needlessly off topic
for the rt-users list.

Trevor Sky Garside
trevor@trevorsky.com

PS – Please don’t flame me… ;)From: “Darrin Walton” darrinw@nixc.net
To: “Technical Support” service@lobosoft.com
Cc: rt-users@lists.fsck.com
Sent: Wednesday, August 15, 2001 5:11 PM
Subject: Re: [rt-users] Need “DB_PORT” in Makefile and support throughout
build

Let me present this situation to you.

You have to bring up a circuit (say OC48). Bringing up the circuit to
test could, in some way crash the router. (router vendor not important,
it could be either cisco or juniper).

Now, bringing the circuit up on a ‘test’ router, if you have one available
for both sides of the connections (which let me tell you, gets quite
expensive, usually test equipment is kept in a lab somewhere). You test
the circuit, it is clean. Now, you still have to put it on the production
router. Putting the circuit on the production router could crash the
router. What do you do?

Your solution is not feesable for every situation.

I’m going to have to agree with trevor here. Besides, I implemented
it tonight. It’ll be in the next prerelease of 2.0.6. I look forward to
hearing about whether it works right from folks who need it ;)On Wed, Aug 15, 2001 at 06:00:48PM -0700, Trevor Sky Garside wrote:

I’d like to point out that this discussion is getting needlessly off topic
for the rt-users list.

Trevor Sky Garside
trevor@trevorsky.com

PS – Please don’t flame me… :wink:

----- Original Message -----
From: “Darrin Walton” darrinw@nixc.net
To: “Technical Support” service@lobosoft.com
Cc: rt-users@lists.fsck.com
Sent: Wednesday, August 15, 2001 5:11 PM
Subject: Re: [rt-users] Need “DB_PORT” in Makefile and support throughout
build

Let me present this situation to you.

You have to bring up a circuit (say OC48). Bringing up the circuit to
test could, in some way crash the router. (router vendor not important,
it could be either cisco or juniper).

Now, bringing the circuit up on a ‘test’ router, if you have one available
for both sides of the connections (which let me tell you, gets quite
expensive, usually test equipment is kept in a lab somewhere). You test
the circuit, it is clean. Now, you still have to put it on the production
router. Putting the circuit on the production router could crash the
router. What do you do?

Your solution is not feesable for every situation.


rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

jesse reed vincent – root@eruditorum.orgjesse@fsck.com
70EBAC90: 2A07 FC22 7DB4 42C1 9D71 0108 41A3 3FB3 70EB AC90

I admit that X is the second worst windowing system in the world, but all the
others I’ve used are tied for first.