Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I returned to AWS and was reminded why I left

fourlightyears.blogspot.com

Post reply on HN
Print view View original

661–670 of 684 posts

Re: I returned to AWS and was reminded why I left

#661

Post by hunterpayne »

bornfreddy wrote: post #528
hunterpayne wrote: post #477

Earlier quoted context omitted.

"They have a problem with their business model, then" Ok, then don't be surprised when the most popular license becomes the FairSource license. Under this license, you have no rights, no ability to fork and no ability to modify, no ability to legally change the software in any way, but hey...you can see the source right. I feel like you don't understand the tragedy of the commons somehow.

That's a huge misrepresentation of fair source licenses. They prevent competing with the original vendor, but still try to retain Right to Repair as much as possible, for example: > The Fair Core License, or FCL, is a mostly-permissive non-compete Fair Source license that eventually transitions to Open Source after 2 years.

The license is what I say it is when I release. You get no rights. You won't be able to fork. You won't be able to legally compile it. You get no rights. Because you have abused the trust of the devs. This is the new normal. Congrats on destroying one of the biggest sources of value in society. And nobody else using it will care because they aren't sociopaths that think they have the right to steal from others.

Re: I returned to AWS and was reminded why I left

#662

Post by firesteelrain »

hunterpayne wrote: post #660
firesteelrain wrote: post #514

Earlier quoted context omitted.

Ok - So you are annoyed I am using something for free and per the license that the authors set themselves and wanted no compensation. Got it.

Nope, I'm not going to use a license that allows that for my next project. And neither will anyone else. Enjoy doing without.

And you are exactly?

Re: I returned to AWS and was reminded why I left

#663

Post by dotwaffle »

cyberax wrote: post #655
dotwaffle wrote: post #650

Earlier quoted context omitted.

GCS / Azure can survive a total AZ failure too, think of 3 x Replicas as RAID-1, whereas Erasure Coding is more like RAID-6. Only it's actually more resilient: Let's say you have 10 data blocks, and you have 4 parity blocks. You can now lose 4 servers containing a block and still be able to repair the data, whereas in 3 x Replica you can only lose 2, and have to store everything 300% of size, instead of only 140%. An…

The problem with erasure coding is that if a disk fails, you suddenly need to read 3-5x more data to reconstruct the missing data from parity blocks. This is especially problematic if your replicas are split across zones. The inter-zonal bandwidth is large, but not infinite. So I'm pretty sure that you need to have at least 2 full copies in different AZs, and then likely at least some additional redundancy within a s…

That's what Locally Recoverable Codes (including Pyramid Codes) are designed to address: you can repair a missing block using only a subset of the blocks, which you can make sure are placed in just one zone, eliminating the cross-az bandwidth requirements. Sure, if you lose multiple blocks, you are going to end up needing that extra bandwidth, but the chances of having two blocks offline at once is very low -- if it's just for a failure rather than extended maintenance, then you have probably already repaired the first block by the time the second fails.

In fact, the RS configuration is often in the >50 data blocks and >10 parity blocks range (albeit with an LRC/nested RS config) for object stores because it's more important to have that recoverability than repair efficiency. While one large provider I worked at did have a system whereby they did effectively have two copies of the RS-encoded data (so that 130% turned into 260%) across two AZs, they were actively in the process of swapping to the blocks being evenly distributed across the AZs, near-halving the total required disk space.

As I said before, most object storage is not on SSD, it's on hard disk: it's 20% of the price per TB, and most objects are read very infrequently. I can promise you that they're not paying $200 for 1TB of SSD either... I realise prices are higher than sensible in the last 6 months, but it was fairly easy to pick up SSDs for under $50/TB at retail pricing (and hard disks for under $10/TB) only a year ago.

Re: I returned to AWS and was reminded why I left

#664

Post by dns_snek »

Orygin wrote: post #594
dns_snek wrote: post #587

Earlier quoted context omitted.

> Why isn't this a problem for other databases then? It is? - MongoDB went from AGPL to SSPL - Redis went from BSD to SSPL - Elasticsearch went from AGPL to SSPL - CockroachDB went from Apache to BSL - TimescaleDB went from Apache to Apache + TLS - Graylog went from GPL to SSPL > It's because the business model for ES is direct competition with AWS and others, and they got out competed. So they had to play licenses g…

Yes and the result is these databases got forked, and the community got rightfully mad. But other databases don't need it, and stayed truly open source, because their business model doesn't rely on being the only hosting provider. > You didn't incur any costs in producing the thing you're selling, duh! Indeed, you gave it away for free, saying I could sell it... It doesn't take a business genius to know AWS can under…

No "buts", this is an extremely common trend not some one-off example like you tried to claim.

Other databases are already funded by cloud companies, predate the cloud, or they're too niche to bother.

Re: I returned to AWS and was reminded why I left

#665

Post by dns_snek »

dmantis wrote: post #614
Orygin wrote: post #594

Earlier quoted context omitted.

Yes and the result is these databases got forked, and the community got rightfully mad. But other databases don't need it, and stayed truly open source, because their business model doesn't rely on being the only hosting provider. > You didn't incur any costs in producing the thing you're selling, duh! Indeed, you gave it away for free, saying I could sell it... It doesn't take a business genius to know AWS can under…

> don't really care about open source. Exactly. You can sell the products of your work all the way you want. But pretending to share with the world and then push back when the world actually use it under these same open terms is a hypocrisy.

Who's pretending? If I share something with everyone for any purpose except one specific purpose that's endangering my project's existence that's not "pretending".

And even that's overstating it because there's no prohibition of any kind. Cloud providers are free to use SSPL licensed software as long as they release all associated platform code.

After all we're sharing, right?

Re: I returned to AWS and was reminded why I left

#666

Post by cantalopes »

ovaistariq wrote: post #628
cantalopes wrote: post #560

sorry to hijack this but since we're talking the AWS egress exorbitant prices, could someone recommend a reliable production-tested s3 service that will not ruin me financially? i have been with hetzner and while their vps server are great, they are absolutely terrible at s3. random downtimes, incomplete uploads/downloads, capacity issues, terrible api key management, etc - but the failure rate is really high

Check out Tigris: https://www.tigrisdata.com/

It looks quite nice on the paper, but honestly, i have browsed the website and see fal.ai in the case studies which is a very red flag for me.

We use fal heavily in our manual production pipeline and their cdn/storage speed is really bad. Like, my colleagues that generate videos in the platform sit on the platform sometimes wait for minutes before a generated 5-10 video downloads from the servers, the speeds are very slow. Sometimes it works but half of time i hear them b!tching about fal being slow, when i checked the network requests, the content that's getting fetched from the v3b.fal.media for example goes reeeeally bad

Re: I returned to AWS and was reminded why I left

#667

Post by rpdillon »

hunterpayne wrote: post #661
bornfreddy wrote: post #528

Earlier quoted context omitted.

That's a huge misrepresentation of fair source licenses. They prevent competing with the original vendor, but still try to retain Right to Repair as much as possible, for example: > The Fair Core License, or FCL, is a mostly-permissive non-compete Fair Source license that eventually transitions to Open Source after 2 years.

The license is what I say it is when I release. You get no rights. You won't be able to fork. You won't be able to legally compile it. You get no rights. Because you have abused the trust of the devs. This is the new normal. Congrats on destroying one of the biggest sources of value in society. And nobody else using it will care because they aren't sociopaths that think they have the right to steal from others.

> aren't sociopaths that think they have the right to steal from others.

I cannot possibly fathom how you're getting from "I deployed some GPLv3 software as part of my hosting service" to "sociopath that think(s) they have a right to steal from others."

Copyright, by default, reserves all rights for the author. The authors then chose to license under a license that explicitly gives all those rights back, and when users leverage those rights, it is being cast as theft. That's...not coherent.

These companies want it both ways: open-source to gain traction, closed-source to monetize. In other areas, we call this "enshittification", and that's what it is here, too.

Re: I returned to AWS and was reminded why I left

#668

Post by bcjdjsndon »

Our_Benefactors wrote: post #647
devin-2030 wrote: post #603

Earlier quoted context omitted.

Used to be different when AWS started properly 20 years ago with S3 and EC2. But yes they went from catering to startups to SMBs to enterprise. Unfortunately a lot of developers still believe the AWS / GCP / Azure propaganda (tech evangelism, cloud advocacy, or whatever else you're calling it). Nowadays you shouldn't use them at all when starting up.

What should someone use? Selfhost? Cheap VPS? Genuinely curious to hear your thoughts.

Same thing we did before aws was invented... cloud is just a managed server half the time, not really anything new. And what kind of IT professional is daunted by bread and butter shit like running a server? That's like an accountant that outsources the accounting to someone else cos "it's too complex"

Re: I returned to AWS and was reminded why I left

#669

Post by simonebrunozzi »

andrewstuart wrote: post #642
simonebrunozzi wrote: post #624

Earlier quoted context omitted.

There's many reasons to dislike a company or service. I know. It is not a good reason to look for employment to "get in and out". If I were your hiring manager, I would be upset to know. Mike Culver, great guy and a bit of a mentor to me. Yes, very sadly he passed away several years ago now. RIP.

Hey Simone you misunderstand. You keep mentioning employment… but I was a customer never an employee of AWS. Please reread the post….

Oh, my apologies. You are right. Everything makes sense now.

Re: I returned to AWS and was reminded why I left

#670

Post by htx80nerd »

throwaway2037 wrote: post #635
htx80nerd wrote: post #626

Earlier quoted context omitted.

Hetzner has the most unintuitive UI out of any host I've ever used in 20+ yrs of using different web host. Spending hours trying to do basic things, and even AI couldnt help. Switched to one of their competitors after 2(!) days of trouble shooting and got the same basic task done in ~10 mins.

> Hetzner has the most unintuitive UI out of any host This is the first negative thing about Hetzner I have read on HN. Does anyone else dis/agree? > one of their competitors Can you share which one?

Digital Ocean. I should add going into the whole situation I had a positive feelings about Hetzner and negative feelings about Digital Ocean. It's why I tried Hetzner first. I didnt like D.O. or want to use them. But in the end their interface was way easier to work with.
Post reply on HN