Tuesday, July 14, 2009

Spot the bottleneck

This article (http://www.internetretailing.net/news/waitrose-picks-precise-to-manage-website-performance-and-availability) is quite interesting for a couple of reasons:

  1. Did they really need a contract with a third party to work out that some SQL statement were causing performance issues?
  2. The fact that improving SQL performance is their primary mechanism for performance gains.
  3. IBM provided the platform.

It would be very interesting to get a view into their existing architecture – to see how much tuning they have done so far.

(PS – I know performance is difficult – you only need to look at the website I currently work with/for/on to see that I don’t have all the answers; my argument is that if they have the money to employ someone specifically for the task that suggests they have exhausted all possible internal solutions. There is surely nothing more dispiriting for a development team than to be held to account for a performance issue and then denied the resources to solve it, only to see a third party parachuted in and given all the assistance they require.)

Update:

The other thing that I wanted to mention was the issue of scalability – we’re being asked to scale 1,000% in three months – should I be scared? An increase of 35% seems trivial – why can’t they scale out to cover that increase?

Friday, July 03, 2009

Tipping Point?

The mercury has now popped out of the top - when Computerworld starts picking up on these things we can now assume they have gone mainstream: http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9135086

To paraphrase ((c) Computerworld.com)
The movement's chief champions [...] learned to get by at their cash-strapped startups without Oracle by building their own data storage solutions, emulating those being built by Google and Amazon.

Now that their open source data stores manage hundreds of terabytes or even petabytes of data for thriving Web 2.0 and cloud computing vendors, switching back is neither technically, economically or even ideologically feasible.