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

Hetzner launches three new dedicated servers

hetzner.com

Post reply on HN
Print view View original

271–280 of 320 posts

Re: Hetzner launches three new dedicated servers

#271

Post by Strom »

silksowed wrote: post #252
input_sh wrote: post #132

Earlier quoted context omitted.

The biggest catch is that you're getting more desktop-grade hardware than server-grade (notice Intel's i series instead of Xeon, non-ECC RAM). Doesn't make a lot of difference for the vast majority of use cases, but something to keep in mind. You can get server-grade hardware from them, but then the pricing difference isn't so significant when compared to other providers.

i really wonder how they are housing the desktop grade hardware. im so used to rack and stack servers (1U/2U), but do you really need that big of a chassis for a desktop cpu, a couple ram dimm's and some ssd? what are you're thoughts?

They have ATX cases on shelves. You can take a look at this tour of their datacenter [1] for more insight. ATX cases are visible around the 3 minute mark.

--

[1] https://www.youtube.com/watch?v=5eo8nz_niiM

Re: Hetzner launches three new dedicated servers

#272

Post by nailer »

darkwater wrote: post #149
nailer wrote: post #128

Earlier quoted context omitted.

The expectation with AWS and GCP is that the constant intermittent failures are shouldn’t be visible to end users. * So filing tickets is a big difference in user requirements. * Though GCP (back when it was just AppEngine) wasn’t always this way and as GAE users inside Google we had to write our own code for what we expected to fail, retries, backoffs etc.

I still have a few times per month mails from AWS with an "instance retirement", for which then you have to plan accordingly (basically by not having a single instance as a single point of failure). If you do the same with Hetzner, you would not notice the failure either (failed disk on a single machine? just business as usual, it will eventually be replaced). Obviously replacing AWS features like RDS multi-AZ master…

Hah me too! EC2 instance retirement was so long ago I’d forgotten. :D

Re: Hetzner launches three new dedicated servers

#273

Post by e12e »

abricot wrote: post #256
akvadrako wrote: post #114

Earlier quoted context omitted.

Not my experience. Hetzner is really a discount provider, though they may have gotten better. I used to work for a team that rented dozens of servers from them and we had disk failures almost every other week, which required creating a support ticket and asking them to swap out the drive so we could rebuild the RAID array. They used regular SATA consumer drives and they were probably pretty old or refurbished or some…

It's up to you what type of harddrive you order in your servers. So I guess you can blame your own team for ordering consumer sata?

Sounds like op might have been using the cheapest option without ecc ram, server grade cpus etc?

Re: Hetzner launches three new dedicated servers

#274

Post by renaudg »

If you care about your carbon footprint, hosting in Germany (where Hetzner has most of its DCs) is unfortunately not a great option.

France (OVH), Sweden, Norway, and (usually) Switzerland are much better (thanks to nuclear for France, and hydro for the others).

See https://app.electricitymaps.com

Re: Hetzner launches three new dedicated servers

#275

Post by e12e »

OmarAssadi wrote: post #62
speedgoose wrote: post #26

Earlier quoted context omitted.

I rent one AX101 and it has been extremely good value. The thing is so cheap and fast. However my first one did often reboot randomly and the support wasn’t very helpful. They told me to just rent another one, which I did. The second one rebooted randomly once in about a year. I guess the first one went on auction and still happily reboots. Hetzner feels like a hard discount cloud provider. I still prefer them over A…

> Hetzner feels like a hard discount cloud provider. Not to discount your experience, but honestly, my experiences with Hetzner support have generally been unexpectedly good. They've been very quick to respond, tend to immediately start on whatever my issue is if I provide enough info in the initial ticket, etc. And unlike OVH, I haven't felt like I needed to call them on the phone to get decent service. Kind of surp…

> being able to burst a bit higher than gigabit occasionally without paying the €1/TB bandwidth fee would be beautiful.

IIRC you get 30tb/month included - so it's not "pay nothing vs pay from first tb" - but I could be wrong - I've not yet had any projects where 10gbps made sense.

Re: Hetzner launches three new dedicated servers

#276

Post by fidgewidge »

nmjohn wrote: post #236
izacus wrote: post #229

Earlier quoted context omitted.

Yeah, turns out that if you write code for constant intermittent failures, it'll run well for rare intermittent failures of Hetzner machines too.

With AWS - failure means the instance is automatically retired, and your ASG causes a new instance to automatically be created and put in service without you having to do anything. With hetzner - failure means your monitoring detected disk failure, sent you a pagerduty alert, which you then have to check the alert, figure out what has failed, and send in a support ticket to get the disk replaced. This will take a cou…

That's apples and oranges. With RAID the server is never retired and you don't need to set up auto-scaling and all the scale-out complexity that comes with it. It just keeps running. The replace/resilver cycle may degrade performance whilst the data is re-replicated, but bringing up a new VM will also degrade performance for a while whilst it replicates data from some other node onto itself.

Re: Hetzner launches three new dedicated servers

#277

Post by e12e »

renaudg wrote: post #274

If you care about your carbon footprint, hosting in Germany (where Hetzner has most of its DCs) is unfortunately not a great option. France (OVH), Sweden, Norway, and (usually) Switzerland are much better (thanks to nuclear for France, and hydro for the others). See https://app.electricitymaps.com

Hetzner also has a pop in Finland.

Re: Hetzner launches three new dedicated servers

#278

Post by mdasen »

input_sh wrote: post #132
esperent wrote: post #17

Earlier quoted context omitted.

I really wish I could get dedicated hardware for somewhere close to that price here in Asia. It seems like Hetzner is the only company in the world offering these kind of prices, right? What's the catch?

The biggest catch is that you're getting more desktop-grade hardware than server-grade (notice Intel's i series instead of Xeon, non-ECC RAM). Doesn't make a lot of difference for the vast majority of use cases, but something to keep in mind. You can get server-grade hardware from them, but then the pricing difference isn't so significant when compared to other providers.

If you go with their AMD servers, you'll get ECC RAM which would be the main sticking point for me. ECC memory will be €5.50 extra on their €37 Ryzen 5 3600 (6 core, 12 thread) 64GB RAM server. Even at €43 with EEC RAM, it's still a great deal. It even comes with 2x512 GB NVMe storage.

Yes, non-ECC RAM is an issue, but that's easily upgraded on their AMD servers.

For €63 you'll get a Ryzen 7 7700 (Zen 4, 8 core, 16 thread) box with 64GB ECC RAM and 2x1TB NVMe SSDs. Google Cloud's N2D-Standard-16 with 8 cores (16 vCPU threads, Zen 2 or Zen 3), 64GB ECC RAM, and no storage costs $550/mo. Yes, it's may not be a perfect comparison, but it's also 8x the price - oh, and Google will charge you $0.085/GB for bandwidth that Hetzner throws in for free. Even Google's Spot Pricing is more than double the cost.

I do agree that non-ECC RAM is an issue, but if you're willing to go with AMD servers, it becomes a very cheap issue to fix.

Re: Hetzner launches three new dedicated servers

#279

Post by tosihakkeri »

renaudg wrote: post #274

If you care about your carbon footprint, hosting in Germany (where Hetzner has most of its DCs) is unfortunately not a great option. France (OVH), Sweden, Norway, and (usually) Switzerland are much better (thanks to nuclear for France, and hydro for the others). See https://app.electricitymaps.com

Hetzner has a data center in Finland too

Re: Hetzner launches three new dedicated servers

#280

Post by izacus »

nmjohn wrote: post #236
izacus wrote: post #229

Earlier quoted context omitted.

Yeah, turns out that if you write code for constant intermittent failures, it'll run well for rare intermittent failures of Hetzner machines too.

With AWS - failure means the instance is automatically retired, and your ASG causes a new instance to automatically be created and put in service without you having to do anything. With hetzner - failure means your monitoring detected disk failure, sent you a pagerduty alert, which you then have to check the alert, figure out what has failed, and send in a support ticket to get the disk replaced. This will take a cou…

You're, In my real world experience their reliability beats out AWS reliability by a massive margin.

On AWS, something is constantly breaking. One of the 100s of services will always have performance issues, degraded availability or some other crap going on.

On Hetzner, the hard drive, CPU or RAM on one of the machines will die once every few years. Maybe.

(This changes as your service grows and scales out, but there's a stupid high amount of traffic a few machines can take.)

Post reply on HN