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

Migrating from AWS to Hetzner

digitalsociety.coop

Post reply on HN
Print view View original

201–210 of 643 posts

Re: Migrating from AWS to Hetzner

#201

Post by 1a527dd5 »

Love it!

We are unfortunately moving away from self-hosted bare metal. I disagree with the transition to AWS. But it's been made several global layers above me.

It's funny our previous AWS spend was $800 per month and has been for almost 6 years.

We've just migrated some of our things to AWS and the spend is around $4,500 per month.

I've been told the company doesn't care until our monthly is in excessive of five figures.

None of this makes sense to me.

The only thing that makes sense is our parent company is _huge_ and we have some really awesome TAMs and our entire AWS spend is probably in the region of a few million a month, so it really is pennies behind the sofa when global org is concerned.

Re: Migrating from AWS to Hetzner

#202

Post by abujazar »

Aissen wrote: post #180
rightbyte wrote: post #159

Earlier quoted context omitted.

What is old is new again. My employer is so conservative and slow that they are forerunning this Local Cloud Edge Our Basement thing by just not doing anything.

As an infrastructure engineer (amongst other things), hard disagree here. I realize you might be joking, but a bit of context here: a big chunk of the success of Cloud in more traditional organizations is the agility that comes with it: (almost) no need to ask permission to anyone, ownership of your resources, etc. There is no reason that baremetal shouldn't provide the same customer-oriented service, at least for th…

The permissions and ownership part has little to do with the infrastructure – in fact I've often found it more difficult to get permissions and access to resources in cloud-heavy orgs.

Re: Migrating from AWS to Hetzner

#203

Post by swiftcoder »

matt-p wrote: post #178
swiftcoder wrote: post #166

Earlier quoted context omitted.

to be fair, AWS' database restore support is generally only a small part of the picture - the only option available is to spin an entirely new DB cluster up from the backup, so if your data recovery strategy isn't "roll back all data to before the incident", you have to build out all your own functionality for merging the backup and live data...

I think the "strategy" for most people is to do it manually, or make the decision to just revert wholesale to a particular time.

Yeah, and that default strategy tends to become very, very painful the first time you encounter non-trivial database corruption.

For example, one of my employers routinely tested DB restore by wiping an entire table in stage, and then having the on call restore from backup. This is trivial because you know it happened recently, you have low traffic in this instance, and you can cleanly copy over the missing table.

But the last actual production DB incident they had was a subtle data corruption bug that went unnoticed for several weeks - at which point restoring meant a painful merge of 10s of thousands of records, involving several related tables.

Re: Migrating from AWS to Hetzner

#204

Post by belter »

adamcharnock wrote: post #150

I cannot overstate the performance improvement of deploying onto bare metal. We typically see a doubling of performance, as well as extremely predictable baseline performance. This is down to several things: - Latency - having your own local network, rather than sharing some larger datacenter network fabric, gives around of order of magnitude reduced latency - Caches – right-sizing a deployment for the underlying har…

> If you don't want to do this yourself, then we'll do it for you for half the price of AWS (and we'll be your DevOps team too

You might not realize but you are actually increasing the business case for AWS :-) Also those hardware savings will be eaten away by two days of your hourly bill. I like to look at my project costs across all verticals...

Re: Migrating from AWS to Hetzner

#205

Post by mythz »

We've moved most of our Apps off AWS to Hetzner years ago by switching to SQLite/Litestream > Cloudflare R2 replication to avoid needing to using a managed RDBMS and saved a bunch of $$$ [1].

Although for our latest App we've switched to using local PostgreSQL (i.e. app/RDBMS on same server) with R2 backups for its better featureset, same cost as we only pay for the 1x Hetzner VM and Cloudflare R2 storage is pretty cheap.

[1] https://docs.servicestack.net/ormlite/litestream

Re: Migrating from AWS to Hetzner

#206

Post by matt-p »

swiftcoder wrote: post #203
matt-p wrote: post #178

Earlier quoted context omitted.

I think the "strategy" for most people is to do it manually, or make the decision to just revert wholesale to a particular time.

Yeah, and that default strategy tends to become very, very painful the first time you encounter non-trivial database corruption. For example, one of my employers routinely tested DB restore by wiping an entire table in stage, and then having the on call restore from backup. This is trivial because you know it happened recently, you have low traffic in this instance, and you can cleanly copy over the missing table. Bu…

Yeah, but automating a solution for all possible "one off subtle data corruption bugs" is a lot of energy and effort to be honest.

Re: Migrating from AWS to Hetzner

#207

Post by endymion-light »

rs_rs_rs_rs_rs wrote: post #149
endymion-light wrote: post #36

Earlier quoted context omitted.

It's really dismissive and frankly quite ignorant to have an attitude that just because a product doesn't have a massive AWS bill it's a toy project. It's a rotten attitude, and judging a projects worth by an AWS bill is a very poor comparator. I could spin up a massive aws bill doing some pointless machine learning workloads, is that suddenly a valid project in your eyes?

>I could spin up a massive aws bill doing some pointless machine learning workloads, is that suddenly a valid project in your eyes? Can you spin it on a AWS competitor for a fraction of a cost? Absolutely yes I would be interested in reading about it!

I will do - but my latest ML model is attempting to create leylines of different mcdonalds across the country, i don't think it's worthy of being considered product

Re: Migrating from AWS to Hetzner

#208

Post by fakwandi_priv »

kees99 wrote: post #133
jwr wrote: post #76

Earlier quoted context omitted.

I actually benchmarked this and wrote an article several years back, still very much applicable: https://jan.rychter.com/enblog/cloud-server-cpu-performance-...

Did you "preheat" during those tests? It is very common for cloud instances to have "burstable" vCPUs. That is - after boot (or long idle), you get decent performance for first few minutes, then performance gradually tanks to a mere fraction of the initial burst.

> The total wall clock time for the build was measured. The smaller the better. I always did one build to prime the caches and discarded the first result.

The article is worth the read.

Re: Migrating from AWS to Hetzner

#209

Post by CuriouslyC »

likium wrote: post #129
CuriouslyC wrote: post #6

I use Hetzner for this reason, but there are caveats. They're great but their uptime isn't as good as AWS and they don't have great region coverage. I strongly advise people to pair them with Cloudflare. Use Hetzner for your core with K8s, and use R2/D1/KV with Container Durable Objects to add edge coprocessing. I also like to shard customer data to individual DOs, this takes a ton of scaling pressure off your data l…

If customer data is considered edge, then what’s core?

Everything that's shared between customers, internal system state and customer metadata. I use Postgres with FDWs + Steampipe + Debezium to integrate everything, it's more like a control plane than a database. This model lets you go web scale with one decently sized database and a read replica, since you're only hitting PG for fairly static shared data, Cloudflare Hyperdrive gives insane performance.

Re: Migrating from AWS to Hetzner

#210

Post by withinboredom »

jwr wrote: post #66

I've been running my SaaS on Hetzner servers for over 10 years now. Dedicated hardware, clusters in DE and FI, managed through ansible. I use vpncloud to set up a private VPN between the servers (excellent software, btw). My hosting bill is a fraction of what people pay at AWS or other similar providers, and my servers are much faster. This lets me use a simpler architecture and fewer servers. When I need to scale, I…

Similar setup to me (including rethinkdb). Why choose FoundationDB? RethinkDb is still maintained and features added occasionally (I'm on the rethinkdb slack and maintain an async php driver). It just is one guy though, working on it part time.
Post reply on HN