Thursday, November 14, 2013

Will AWS plans for PostgreSQL RDS help it finally pick up?

"Amazon to add Postgres to its most-favored database list" says GigaOM:
"To many this is no-brainer. Amazon wants to support the databases that its developer audiences want to use. This is simply a  case of Amazon responding to user demand and oh-by-the-way making its cloud infrastructure more attractive to a specific target audience. Some say Postgres has gained traction since Oracle’s acquisition of MySQL via its Sun buyout a few years back."

Some people I know said "yea, the writing was on the wall...". Well, was it?? Really? 

AWS finally got the time to "plan" for supporting Postgres now? After supporting MySQL, Oracle and SQL Servers for almost 3 years?! Writing was on the wall? Where can I find a wall this old?

PostgreSQL has not picked up. 

This is why it is a far 4th on Amazon's list. The writer of the text above also makes clear efforts not to pick a side here... "to many this is a no-brainer" or "some say Postgres has gained traction". 

It has been around for ages, thru many "oh! it's now happening!" events, such as the acquisition by of MySQL by Sun, then by Oracle... 

Technically, PostgreSQL's few superior capabilities, especially around schema online modifications (which gets more important these days!), probably could not change its fate, and it's still being held back by too many inferior capabilities, around performance, robustness, ecosystem... 

So - with plans for RDS, will Postgres now pick up? 

Feel free to Share your thoughts... 


  1. "held back by too many inferior capabilities, around performance, robustness, ecosystem... "

    Wow, did you ever work woth PG ? I managed cluster of MySQL and PostgreSQL and those assumptions were valid 7 years ago, not anymore.

    1. I appreciate your comment and my answer will "yes!", I have worked with both, I totally agree that pg is in constant improvement, but I still can't shake this feeling that... things ran better, more stable and faster and constant, for me, with MySQL, it was expected, no surprises, indexes, execution plans, just works. With pg, I found myself scratching my head a little more often... #sorry. Saying that, I see pg is gaining and improving faster than MySQL is.

      Here's an interesting question, you might be able to answer with your experience: it's about ecosystem and tools around it... how come pg, BSD open source free as a bird, and GPL and Oracle based MySQL has all the community?
      MySQL has branches, Maria, Percona, Drizzle... So many tools... pg is still pushed a lot by EnterpriseDB, less by community.
      Weird... I can't explain. Part of it I guess it's the "storage engine" effect, but I don't know.

      If you have any idea - I'd love to hear!

  2. People fail to realize that AWS has offered PostgreSQL for some time. The data warehousing RedShift is actually PostgresSQL.

    And for the record, I still consider Ingres, which is open source, more fully featured and previously had huge user bases as a more superior product then MySQL and PostgreSQL by a long margin. PostgreSQL is actually being a fork from Ingres. So why is Ingres now a dead product?

    1. Ronald thank you for you comment.
      You're making a valuable point for our readers:

      Throughout the years, BSD license made PostgreSQL "embedded"-worthy, and it is in fact embedded in many products including many analytics MPP databases, I can mention a few, such as GreenPlum, Netezza, Aster.
      ParAccel (the foundation under AWS Redshift) also embeds PostgreSQL code, yet not to a large extent, according to Mr. Monash on the subject:

      IMHO Ingres is a fine choice for a database!

      Thanks for sharing.