The Google Disease Afflicting AWS

I hope this makes the front page. I&#x27;ve personally been complaining about this problem for a couple years now and it&#x27;s only getting worse.<p>Amazon is incentivized to release 100&#x27;s of half baked services because an org only needs to use a couple of them and suddenly they find themselves unable to leave AWS. It&#x27;s like a drug dealer offering free hits to get you hooked. There&#x27;s little incentive to build new feature rich and comprehensive services or continue to improve the existing ones.<p>It also feels like Amazon&#x27;s organization has structural problems. Some services are managed well, some are not. Some have a comprehensive vision, others are a hodge-podge of amazon services YOU have to glue together and manage yourself (see setting up Guard Duty email notifications as an example). Good luck if it doesn&#x27;t work and the documentation is frequently outdated or incorrect. And then as the article points out there&#x27;s the variations upon variations of an existing service and no clear reason why the functionality justifies separate products.<p>Don&#x27;t even get me started on the console. At least it&#x27;s not as bad as Google or Azure in my experience.<p>It&#x27;s a mess.

I just ran into this. What is the difference between EC2 autoscaling groups, Spot Fleets, and EC2 fleets? At one point they had pretty different features which was a problem, because you might very well need a combination of features no single service had. AWS has largely fixed that by adding featurea to each service, but now it isn&#x27;t clear why you should choose one over the others. At the very least I wish they would have documentation on how to pick which one to use, and maybe focus on one and deprecate the others.