Live data from Hacker News

AWS services explained in one line each

adayinthelifeof.nl

291–300 of 355 posts

Re: AWS services explained in one line each

#291

Sometimes it pays to be boring. One of the hardest parts of using AWS is learning all their silly names for everything. I know it's tempting to be cute when naming things, but everyone else wishes you would just be clear and descriptive. I've seen this play out at startups that love to name servers after galaxies or cartoon characters. It's all fine until your new employee onboarding guide comes with a massive memori…

I worked at a company that had names for all the conference and meeting rooms based on various science fiction names and other "nerdy" stuff instead of some sort of building-floor-room_number scheme. It was neat at first, but quickly became agonizing trying to remember what _building_ a conference room was in much less where in that building. It added an extra few minutes to every meeting for me where I had to either look it up on the internal wiki or go ask someone at the front desk where I was supposed to go or make sure at least one person nearby attending the meeting already knew where we were going. It took months to get the hang of for even the more routine rooms.

Re: AWS services explained in one line each

#292

I made this list for my buddy a while back. It's meant to be more humorous than exactly correct Route 53 - Holy shit! It's NSD WAF - Holy shit! It's modsecurity SES - Holy shit! It's Postfix Inspector - Holy shit! It's OSSEC GuardDuty - Holy shit! It's Snort Data Pipeline - Holy shit! It's Cron and Bash Athena - Holy shit! It's Prestodb Glue - Holy shit! It's Hive Metastore and Spark OpsWorks - Holy shit! It's Chef V…

This is what I thought the article was going to be until I clicked through.

Re: AWS services explained in one line each

#293

I made this list for my buddy a while back. It's meant to be more humorous than exactly correct Route 53 - Holy shit! It's NSD WAF - Holy shit! It's modsecurity SES - Holy shit! It's Postfix Inspector - Holy shit! It's OSSEC GuardDuty - Holy shit! It's Snort Data Pipeline - Holy shit! It's Cron and Bash Athena - Holy shit! It's Prestodb Glue - Holy shit! It's Hive Metastore and Spark OpsWorks - Holy shit! It's Chef V…

This guy does a great job of explaining it architecturally -

https://www.youtube.com/watch?v=Z3SYDTMP3ME

Re: AWS services explained in one line each

#294

Earlier quoted context omitted.

I've been horrified at how bad this is on Azure, I'm sorry to hear that AWS has the same issue. I've signed in some mornings only to discover that I've been charged $20 for a MongoDB instance that's been sitting idle all night. Support can only come back with some hand-waving about RUs, a term that defies any clear definition.

Dealing with Azure causes me physical pain. AWS you can poke around and figure it out, Azure is a Kafkaesque nightmare of infinitely confusing UI/UX, a permission system that is like trying to hit a moving target in a dark room, being moved by someone that hates you, doing anything seems needlessly complex, and trying to replicate/test anything locally is horrible (looking at you Azure functions).

Semi-shameless plug (I work for Oracle) but have you tried OCI? You might find both the console UI/UX and the cloud calculator pretty decent. That's just my personal feel, coming over to work as an Oracle cloud consultant with previous AWS/Azure experience. I'd argue it's normal those two are where they are; AWS is now a spaghetti of products, while Microsoft are anti-KISS by DNA.

Re: AWS services explained in one line each

#295

I made this list for my buddy a while back. It's meant to be more humorous than exactly correct Route 53 - Holy shit! It's NSD WAF - Holy shit! It's modsecurity SES - Holy shit! It's Postfix Inspector - Holy shit! It's OSSEC GuardDuty - Holy shit! It's Snort Data Pipeline - Holy shit! It's Cron and Bash Athena - Holy shit! It's Prestodb Glue - Holy shit! It's Hive Metastore and Spark OpsWorks - Holy shit! It's Chef V…

> Neptune - Holy shit! It's Neo4j Neptune - Holy shit! It's Blazegraph https://blazegraph.com/

It's sadly missing the RDF inferencing of Blazegraph though.

Re: AWS services explained in one line each

#296
post #59

Earlier quoted context omitted.

I think it's because naming things is hard, that link proves it by coming up with worse names for almost everything they tried to rename, and often far, far worse. Imagine the confusion if S3 were called "Amazon Unlimited FTP Server." That gets every word wrong, except that "Amazon" is merely redundant. It's not unlimited (having to pay for a thing is a limit), it's not using FTP, and it's a service, not a server. Or…

Oracle Cloud Infrastructure has been trying really hard to make sure things are named in as straightforward a way as possible. It was a very early decision pre-launch, and unsurprisingly not that hard to stick to. Marketing people didn't argue, either, but maybe that's a difference between the marketing team backgrounds? Enterprise company CIOs etc. don't want to have a translation guide when it comes to making purch…

It does help, there are some early decisions that were made pretty well (the first or second time, at least).

Oracle's EC2 is called "compute instances", S3 "object storage", SES "email delivery", Lambda "functions" etc.

And object storage could have been Casper, load balancer Flamingo, or data transfer Rhino.

Re: AWS services explained in one line each

#297
post #52

Earlier quoted context omitted.

Have you ever used AWS? There's a whole suite of tools they provide around pricing and budgeting. > It's quite a tour de force how Amazon have taken "separation of concerns", applied it to web services and used it to create complex and difficult to understand or predict pricing to print money. Is applying separation of concerns to web services really that bad? Look, if you're a small company with a simple product, yo…

> There's a whole suite of tools they provide around pricing and budgeting. It could be there are such tools. But they are impossible to find, setup or understand for someone who is not already an AWS expert. I recently got a $500 bill from AWS. It was impossible for me to find out which S3 buckets (or even specific files) caused these costs. I looked at this for half an hour and have used AWS before. They just don't…

Well, was most of that $500 for storage costs, or for network? Surely you don't expect the monthly bill to include details about particular files or bucket sizes? Have you enabled CloudTrail logging and/or VPC flow logs?

It's not my intention to victim-blame here, yet do bear in mind that there are many, many ways to use the services and to expose resources to the bottomless appetite of the internet. Maybe Digital Ocean or Linode or the like would be a better option for starting out.

Re: AWS services explained in one line each

#298
post #81

Earlier quoted context omitted.

You're right, it's shitty of them to not provide that out of the box. However, if it's such a big problem for you, as others have pointed out, it's easy to solve. If you are paying $500/month for S3, you must be processing quite a large amount of data - as another poster pointed out, there's an easy to fix it. If you weren't going to use S3 for this, what would you use instead for this use-case?

I look forward to migrating to Linode's S3-compatible object storage. I appreciate that you are agreeing that it's shitty of AWS. But I disagree that it's "easy to solve". Like I said, I already spent quite some time looking into how to solve it. And having to set up logging for a bucket (and learning how to do this in AWS's complex interface), and if I understand correctly creating another bucket to receive those lo…

What drew you to using S3 in the first place while also discounting the risk that it might get a lot of use, and thus a lot of cost? Like, if a 100MB file in a public bucket got a bazillion downloads, did you not want that to happen?

Re: AWS services explained in one line each

#299
post #285

I made this list for my buddy a while back. It's meant to be more humorous than exactly correct Route 53 - Holy shit! It's NSD WAF - Holy shit! It's modsecurity SES - Holy shit! It's Postfix Inspector - Holy shit! It's OSSEC GuardDuty - Holy shit! It's Snort Data Pipeline - Holy shit! It's Cron and Bash Athena - Holy shit! It's Prestodb Glue - Holy shit! It's Hive Metastore and Spark OpsWorks - Holy shit! It's Chef V…

I'm not sure if the sarcastic tone is meant to imply AWS is just a rehash of existing tools, but that's obviously not the case. A list of actual competing software/products/services would be pretty useful.

Especially since “EC2” is not just a VM. There are dozen services at least that come under EC2.

Also if you start trying to use DynamoDB like you would Mongo, you’re going to be really upset. The use cases, design considerations and performance considerations of DynamoDB compared to Mongo is night and day.

Re: AWS services explained in one line each

#300
post #286

Earlier quoted context omitted.

>I mean, they're operating this stuff at Internet scale and have never (AFAIK) raised prices Prices should have gone down because they've continued to scale up and go into custom cooling/hardware/backbones/etc. You're basically saying that you're cutting them slack because their margins have gotten much larger and they haven't decided to raise prices to make them even bigger.

> Prices should have gone down because they've continued to scale up and go into custom cooling/hardware/backbones/etc. Why should that make prices go down? Prices aren't based on cost. They're based on what the market will bear, i.e., what customers will pay. If you think AWS is too costly, you can go find some other solution. If enough customers start doing that that AWS's financial state suffers, AWS will have to…

In a highly competitive market prices tend toward marginal costs plus some small profit margin. Think agriculture or plastic trinkets. The fact their prices and profits have remained high for so long indicates lower competition, likely due to high barriers to entry from competitors and high switching costs for consumers.
Post reply on HN