Comment by pid-1 | Hacker News Reader

Comment by pid-1 | Hacker News Reader

Parent Full thread

> where the money is going

AWS in particular is handsomely engineered so you can't answer this sort of question, as well as making people spend more without need.

- Per resource cost is turned off by default. Turning it on costs money. How much? ¯\_(ツ)_/¯

- Services like S3 will bill you per API call. How do you know who is calling your APIs? Using Cloud trail, which is turned off by default. Turning it on for this sort of analysis might cost a lot $$$ for a reasonably used bucket.

- CloudWatch will bill you per custom metric and adds up quickly. Why do I have so many custom metrics? Good luck finding out.

- Many costs are tied with stuff that's really hard to control e.g. egress traffic.

- Using many accounts is by far the best way to segregate unrelated projects. But then, you will end up paying for many criminally underutilized NAT GWs, VPN connections, Load Balancers, etc...

- Many services are ha, multi az only. That makes AWS unnecessarily expensive for dev environments / non critical services.

The list goes on. I know a fair number of consulting companies making the money of their lives reducing AWS bills.

> - Using many accounts is by far the best way to segregate unrelated projects. But then, you will end up paying for many criminally underutilized NAT GWs, VPN connections, Load Balancers, etc...

And you can have business support only on one account... obviously you can go Enterprise and have the same support level on everything but pay a lot more $$$

Is this the same for child accounts created via AWS Orgs?

I guess so, unless we were severely misinformed. We had historically one big AWS account, we are now splitting it into more fine-controlled accounts and migrated under a "root" account following AWS best-practices to have unified billing and we can enable the Business support only on the big (legacy) account where we run 100% prod and other things. I don't think we had the option to enable it on the root account and get all the other accounts covered.

This is not the case for us. Support is automatically enabled for all the accounts we create in our Organization. Not sure if this depends on company size though

Each AWS account can have it's own support plan.. for a while we had some accounts with a mix of Enterprise and business support.. in fact, every child we made for business support for a while even though we had enterprise support across our AWS organizations

I think you can do that only if you have Enterprise at some point, but I might be wrong

Offline contracts could require support to be applied across all spend.

Million dollar question - at which point it gets easier to learn how to host things on your own vs keeping up with this cloud mumbo-jumbo bull type of knowledge?

Old School IT guy here (19 years as Director of IT). I Host 80% of our main programs in house, Financial System, data-management, web servers, email servers, security door lock system, camera systems, etc. I have turned down offers of "cloud" for years we have a block of 32 IP addresses and this has been enough to handle our externally available resource needs.

It just always seems a bit salesie to me. Bring your network switching to the "cloud".. yadda yadda. Why? Everything is working great.

I always get the feeling that most cloud ideas come straight out of California, where hiring even an unexperienced administrator could cost you a couple hundred grand a year. Of course cloud services look pretty attractive if you can just instead hire one consultant for a one time fee and pay way less until your business grows.

The problem is that mostly anywhere else in the world, you can get personell for much cheaper.

If you're a startup with VC money, cloud is for you. The game is scaling as fast as possible, costs be damned. But then came the cargo cult, where if you wanted to be hip, smart, and innovative, why you no cloud? This has led to a lot of orgs jumping in and finding out, "holy shit this is expensive AF and we're not getting a lot of value out of this spend." I have seen this a lot in financial services for example, but there are savvy orgs: Bank of America saves $2B/year by running their own internal cloud infra.

For highly volatile workloads, ephemeral workloads, anything where go forward capacity is fuzzy or unknown, cloud is awesome. For anything where your margins aren't robust, or your workload is known and consistent, you're better off going through the schlep of internal IT management (as long as your org isn't dysfunctional and how you operate this is a business enabler, not a hinderance, forcing shadow IT attempts). If you can't generate more revenue or margin, you've gotta look at cost savings, including the cloud premium.

> For highly volatile workloads, ephemeral workloads, anything where go forward capacity is fuzzy or unknown

The thing is, I've yet to see any startup even with VC money that has a product which has this need. Most startups nowadays factor away common crud, like booking a service or providing organisational tools. These high pressure, intense workloads are often incredibly unrealistic.

Since the beginning.

The scalable cloud is useful for scalable workloads (AKA how many times is you peak demand larger than the average?). If you want managed infrastructure, you are looking for renting a managed server or VPS.

>- Using many accounts is by far the best way to segregate unrelated projects. But then, you will end up paying for many criminally underutilized NAT GWs, VPN connections, Load Balancers, etc...

Not if you use shared VPCs.

Reply on news.ycombinator.com