MSDN UK recently posted a blog on the most common Windows Azure support issues. It’s a short 8m 30s video and is presented by one of Microsoft’s Azure support escalation engineers. As the blog suggests “don’t deploy any service in to production without watching this video first!”
The issues are grouped into three main topics, with examples provided for each:
1. The stateless nature of Windows Azure
- Azure responding to hardware issues
- The risks of saving data to the VMs local drive
- The relationship between virtual IP addresses and the load balancer
- Startup tasks
2. Testing the application’s load
- So bottlenecks are raised and addressed
- How SQL Azure scaling and throttling works
3. Disaster recovery
- Restoring accidentally deleted data
- Backups in triplicate
Further to this the engineer throws in a few tips for debugging and addressing the above areas, i.e. debugging tools and options, backup strategies, testing methodologies, etc.
A worthwhile watch.