Inside Out

There has been quite a big discussion in the blogosphere about the recent downtime of Amazon S3. As we are going to use S3 and EC2 in some of our upcoming projects we ‘re paying close attention to what’s going on with Amazon Webservices (AWS). Of course it also helps to keep an eye on the overall development in the web service/cloud computing space. Some thoughts.

Instant return on (no) investment
AWS does not require any upfront payments. This allows us to develop our project while exploring Amazon’s systems without any investment in hardware, traffic packages etc. Everything is pay per use so there is no financial overhead from servers standing around and waiting to be used.

Out of the box scalability
AWS works by exposing Amazons infrastructure trough web services and thus making it easily accessible for 3rd party use. Due to Amazons massive architecture and computing power we don’t need to worry about storage space and scaling issues. This was one of the key points for us when we decided to go with AWS.

Uptime vs. Price
AWS is pretty cheap. Prices for storage, traffic and computing time are reasonable and seem to be a good deal overall. However, when you are looking for high availability, you might want to look somewhere else. By the time we chose to go with Amazon, none of their web services did have an SLA, which basically means that they could be down anytime leaving us with no chance for compensation. To our pleasant surprise, Amazon introduced an SLA. As we already decided to use AWS without SLA, the 99.9 percent uptime was/is good enough for us (some might argue that it’s not enough).
There are several other providers of similar services and some actually offer 100 percent uptime (e.g. content delivery networks) but the bill and the end of the month might look a little bit different.

The recent AWS downtime reminded us that even Amazon is not impeccable.

Therefore, our basic rules for AWS usage are:

• Do not use AWS as core of your application so that your base services still run if AWS is down.
• If your system relies on AWS, always have a backup solution in place so that you can keep your basic service up.