I'm tired of the manual process of analyzing AWS usage patterns every quarter to make savings plan recommendations. It involves predicting future compute needs, calculating optimal coverage, getting approvals, and making purchases. By the time this process is done, my analysis is often outdated. The cost explorer's recommendations are a good starting point but they're based only on historical usage and don't consider upcoming projects, seasonal traffic, or architectural changes. Overcommitting can lead to wasted expenses while undercommitting leaves savings on the table. Is there an automated workflow that can handle this analysis, or do I just have to keep doing it manually?
8 Answers
There are companies that can manage this process for you, but generally, it's still quite manual unless you decide to automate it yourself. You've highlighted the risks associated with different options quite well.
If you pay for it, you can have someone do this for you, which we do. Ultimately, having someone who really pays attention to your spending and business goals is key to making the right financial choices. You can handle it in-house, hire out, or choose not to address it at all.
Deciding between compute savings plans versus EC2-specific ones is another critical point. Compute options are more flexible, whereas EC2-specific ones can save more money. It all depends on whether flexibility or maximum savings is more important to you.
Cost explorer's recommendations really miss the bigger picture. They don’t consider things like migrating to Lambda next quarter or the fact that traffic consistently doubles in Q4.
Yeah, I get the concern about three-year commitments, especially for startups. One-year terms are definitely safer since who knows how your infrastructure will change by 2027?
There's a secondary market for reserved instances that can help if you're worried about overcommitting. It can mitigate some of that risk.
While automation sounds appealing, relying entirely on a tool to buy commitments without human oversight is risky. I would definitely want to scrutinize the logic of any automated decision-making tool, like ProsperOps or Vantage’s autopilot feature, before letting it run on its own.
It’s frustrating when coverage drops unexpectedly. It requires digging into whether it’s due to a positive adjustment like moving to cheaper instances or a negative issue like losing previously secured coverage.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures