I've done the Advanced Operations course of AWS in CAPSiDE, the company where I'm working. We are the only official AWS formation center in Spain so if you want to get some classes just contact us. If you are in UK maybe you want to come to Spain to receive the course because in London there are a waiting list of months, depending on the course.

It's a four day course and I think is the most advanced course of AWS in this subject. I've really enjoyed it, because is the perfect place to make a lot of questions to the teachers, and we have done labs where you can spawn all the instances you want and test new things, apart from the labs of the course itself that are very complete and useful to understand how everything works.

The course is focused mainly in the automation of the construction of a production/devel environment using CloudFormation, knowing how to build high available stacks, how to use the CloudWatch to retrieve different information of your platform and to trigger alarms if required, and it touches other services such as SQS, a deep dive S3, SES, Elasitic MapReduce, Redshift, and all the other AWS services.

What I've found more interesting have been CloudFormation (in a relation of love/hate) and how it can be combined with Chef. The feeling about all the products (I should say services?) is that are quite easy to use, but you must fully understand how they work for being able to use all the products alltogether and configure the platform you need taking advantage of all the capabilities of AWS.

Overall, I think the most important thing is that they do not guarantee you anything (as any other provider) so you must build a fault tolerant stack for your critical services. This is not something you must do just with AWS, it's something general that applies to any provider or self hosted solution. You cannot relay on your hardware, network, etc.. and expecting that because you have a dedicated instance you will not have problems is like not making backups because you have Raid (and one day the rack burns into flames and you are fucked up). The instances and services in general in AWS are reliable enough as a unit, but AWS provides you a suite of tools and services that you can combine to build a High Available and really easy to manage platform.

And this is kinda cool.

Another cool thing is that you can check here (https://awsu.s3.amazonaws.com/manual/operations/index.html) the content of the course, but you will not be able to do the labs, thought.