How do people make sure they aren't wasting money like this or don't they?
Thanks!
I currently have a fair-amount-of-traffic website sitting on the 1 GB server, while I have several web apps, which aren't getting traffic at the moment, sitting on a smaller single server.
I don't have teams, but it is pretty easy to know what instances I spin up and which ones I'm using. Regardless of whether the server is being used or not, you still get charged. So to not get charged, I only use what I need.
I remember using AWS once, really just for testing, and somehow wracked up a bill for $90 or something like that, despite not getting much traffic. I think I set up a portfolio website, WordPress with like 6 pages, and forgot about it. I doubt it was getting a lot of traffic. Luckily, Amazon actually refunded me. So sometimes pay-for-usage is not always the best way to do things.
But as far as uptime for DreamCompute, so far, other than figuring out how much I had needed in the beginning.. because I had originally set my 1 GB server to 512 MB, the website kept crashing. Once I figured out I needed more RAM and put caching in place, the uptime has been 100%.
Stopping unused resources is half of it-- the overlooked other half is understanding what the bill shows you. Finance doesn't care about how much is compute vs storage, they care about COGS vs R&D spend.
One-off cost reductions are relatively easy; coming up with a longer term approach takes significantly more work.