Command Prompt, Inc.

Postgres Centered full stack support, consulting & development.

Blog

Log level correlations

There was a post recently on using Syslog with PostgreSQL on pgsql-general. As you can see, Tom Lane kindly replied in the thread. I decided that it might be a good idea to submit a DOC patch. I only received on comment on the patch, so I don't know if it will be applied or not. The comment, caused me to consider creating a table for the docs that …

PostgreSQL Conference Fall 07

When sitting in the booth with the Army of Smurfs (tm) at OSCON 2007, I was struck by the idea that having PostgreSQL Conference(s) (yes plural) could be a great way to insure the growth of our community. Check -- duh! You would think that this is obvious. In fact I would think this is obvious. This is why the community had the Anniversary two years ago and Dan Langille …

PostgreSQL 8.3, it's faster, really it is!

I downloaded the latest check out of PostgreSQL 8.3 (not even beta yet) just to see how things are progressing. Of particular interest to me was a recent conversation about sequential scans I had with Jeff Davis. In short, in 8.3 sequential scans should be faster. I decided to test the theory. Using the exact same machine, and an optimized 8.1.8 installation, I compared 8.1 …

How many rows do I have anyway?

Have you ever wondered how many tuples you have in a relation? Normally to find out how many tuples you have you would do something like:
SELECT count(1) FROM rows;
 count 
-------
 10002
This will return the exact number, per your snapshot of committed tuples within a relation. It is also a guaranteed sequence scan on the relation and a performance nightmare on large tables. So how do you get …

Surrogate versus Natural Primary Keys

This is a constant source of argument, flame and general discomfort with any database design. On the PostgreSQL lists it comes up occasionally and it is always a long drawn out thread with people arguing on each side about which one is correct. Before I go on about the good and bad of both, let me define a couple of things. 1. Surrogate keys (also known as artificial keys) are …

LewisC and the PostgreSQL vs EnterpriseDB: Hypocritical crap?

Yet another round of EDB PR came out today and of course, some in the community disagreed with it. This time I was not a big objector although I do think their PR was a little misleading, I think it was generally a reasonable release and that the community needs to relax a bit. You can see that thread here. Toward the end of the thread LewisC of An Expert …

OSCON 2007, PostgreSQL army of smurfs!

OSCON 2007 was a huge success for the PostgreSQL community. Of course, OSCON is usually a good conference for PostgreSQL but this year was different. So what was different about this year than other years? How did we stand out from other community booths? What about MySQL? Professional presence. The community moved to a professional presence of advocating only PostgreSQL at the community booths. Flexing our community muscle to provide …

PgDay Portland, A huge success!

On July 22nd, PostgreSQL.Org held a single day conference in Portland Oregon preceding OSCON 2007. This conference, although short notice was a huge success. We had solid attendance from new and old community members. Notable talks for me was Theo Schlossnagle's talk on Solaris and PostgreSQL. It was enlightening to see where PostgreSQL is lacking, (places I didn't realize) and how Theo has worked around the problems …

For the record, EnterpriseDB and Command Prompt, Inc.

I recently made a post on the pgsql-advocacy list about a press release that EnterpriseDB put out that was less than flattering about PostgreSQL [1]. This thread was long and a little tiring. To make matters worse an Oracle blogger [2] picked up the thread and blogged an incorrect assessment of what happen.

I would like to take a moment and set the record straight. First there were some in …

From the field: On Josh's Rules (of Database Contracting)

Josh Berkus wrote an excellent bullet point list of things to do and not do when doing Database Contracting. I would like to expand on that list and add some comments to a couple of his points:
  1. Data Reflects the Business: show me a client with a chronic database problem, and I'll show you a client with a chronic management problem. Generally I would agree with this statement but …