EC2 are cloud based virtual machines, very easy to spin one up exactly as you need it to run whatever.
But they’re billed hourly and they get very expensive very very fast especially when you leave it on (the hourly billing is for a machine that’s on and running) and have shit tons of resources on it. You can have multiple terabyte RAM machines at the click of a button
The other thing they forgot to mention is that getting Amazon to actually tell you what anything will cost, or a breakdown of what your bill is, is like pulling teeth.
Amazon’s billing tool is the most useless thing I’ve ever seen lol. it’s like “$XYZ of your bill was server compute!”
like, there’s 40 ec2s, elastic beanstalk instances, fargate, kubernetes, like what the fuck is ‘server compute’ and what’s the breakdown of the actual individual parts?
It’s not even unhelpful, it’s literally worse than useless lmao
You can get a full itemized bill. The only thing that isn’t fully broken out are elastic ips. We found that out because we were tagging everything for billing and those weren’t showing up correctly.
Mind you, it’s likely a bit more itemized than you want. Like you’ll see a separate line item for each price tier you paid for something, and things like ebs disks are all split out. It can be a bit…much.
I mean, that’s kind of my point though. You either get worse than useless, over detailed to the point of useless, or you have to spend weeks pulling hacky crap with tags to try to jury-rig your own useful billing system out of their clusterfuck.
I think you’re over exaggerating the effort needed for tagging resources. Between terraform/pulumi/cdk and the tag tool, it’s relatively easy to make sure everything is tagged. Doubly so if you have a finance department who’s literal job is to go through and do that (or ask you for help with it)
Amazon charges per hour for renting a server. Some are very expensive. It’s a common mistake to forget to turn off a server after you’ve finished using it and end up paying more than you planned.
I saw this on all, and I don’t get the joke. Any programmers care to explain in lay terms?
EC2 are cloud based virtual machines, very easy to spin one up exactly as you need it to run whatever.
But they’re billed hourly and they get very expensive very very fast especially when you leave it on (the hourly billing is for a machine that’s on and running) and have shit tons of resources on it. You can have multiple terabyte RAM machines at the click of a button
Thanka for the detailed explanation!
The other thing they forgot to mention is that getting Amazon to actually tell you what anything will cost, or a breakdown of what your bill is, is like pulling teeth.
Oh good to know, I’m more of a GCP guy myself and it’s billing tool isn’t too shabby IMO lolol
Amazon’s billing tool is the most useless thing I’ve ever seen lol. it’s like “$XYZ of your bill was server compute!” like, there’s 40 ec2s, elastic beanstalk instances, fargate, kubernetes, like what the fuck is ‘server compute’ and what’s the breakdown of the actual individual parts?
It’s not even unhelpful, it’s literally worse than useless lmao
You can get a full itemized bill. The only thing that isn’t fully broken out are elastic ips. We found that out because we were tagging everything for billing and those weren’t showing up correctly.
Mind you, it’s likely a bit more itemized than you want. Like you’ll see a separate line item for each price tier you paid for something, and things like ebs disks are all split out. It can be a bit…much.
I mean, that’s kind of my point though. You either get worse than useless, over detailed to the point of useless, or you have to spend weeks pulling hacky crap with tags to try to jury-rig your own useful billing system out of their clusterfuck.
I think you’re over exaggerating the effort needed for tagging resources. Between terraform/pulumi/cdk and the tag tool, it’s relatively easy to make sure everything is tagged. Doubly so if you have a finance department who’s literal job is to go through and do that (or ask you for help with it)
You need to use tags on your resources if you want to get a better breakdown of what your more specific resources are costing you.
You can filter by tags in cost explorer.
I presume Google and Microsoft are similarly bad.
Oh is that why RAM is expensive AF anymore? Post quantum serverless demand for TB of RAMDisk?
What do you mean? You can just download more RAM!
https://www.DownloadMoreRAM.com
You can rent cloud machines from Amazon, they can get pretty expensive if you forget about them
Ahhh, thanks
Amazon charges per hour for renting a server. Some are very expensive. It’s a common mistake to forget to turn off a server after you’ve finished using it and end up paying more than you planned.