Hacker News Reader

> And honestly it’s not hard to run modern Linux machines, that’s an old crones tale.

I totally agree with that. I'm a dev, not a sysadmin. But I run several Linux servers and can set up the network(s), the firewall, the apps and all that. It's not that hard.

You can buy incredibly powerful machines today at a very cheap price. The 20 cores Asahi Linux ARM workstation, 128 GB of RAM recently posted here comes to mind... In the blog's author's word: "The Mac blew the AWS c6g.12xlarge Graviton instance with 48 ARM64 cores instance out of the water."

Or this slightly older blog entry about someone running 25 Gbps fiber to the home. Not to the office. Not to some colo. To the home. Sure, it's not available in many places, but that's quite something. That is someone, at home, having a personal pipe wider than what cloud instances do provide to most.

Heck, my brother, a poweruser (not a dev), just assembled a 7950X / 64 GB of RAM with a 4090 GPU. I think many, used to their laptops, don't realize how powerful such machines are (laptops aren't exactly known for running 4090).

It's honestly a bit insane to me that several devs have, at home, wider pipes and faster machines than what their apps are going to be deployed on.

I'm "only" at 2 Gbps FTTH and don't have a particularly fast machine (just ordered a 7700X / 32 GB of RAM / WD SN850X Black as suggested by someone else here yesterday) but even that is sufficient to make me question a huge lot of cloud instances offerings.

Why not just run IaaS on the cloud? Maintaining a datacenter is expensive. Now I’m not sure if there are companies that provide management of actual physical infrastructure - battery backup, generator, cabling, certifications of premises etc.

> Maintaining a datacenter is expensive. Now I’m not sure if there are companies that provide management of actual physical infrastructure

I knew the cloud marketing machine had done the new generation dirty. I didn't know it was this bad.

So much this! PXE is apparently too hard for engineers to learn to bootstrap a VM, but they'll go to $5k worth of training to learn Terraform to deploy a CRUD app and call that a win :/

I mean, PXE and bootstrapping VMs aren't even close to being what people use terraform for. VM provisioning, in general, is basically just one of the many (and also most trivial) steps in the usual Terraform deployment.

Most terraform flows are just as trivial as any workflow, create/update/delete of infrastructure. Not like we haven't had a hundred frameworks to accomplish that over the last few decades of onprem. Short term enterprise memory just believes there's only one way/place to do that now.

What is ur opinion about hashicorp Tool terraform and vault who can help with multiclod provisioning can bring down the cost of cloud migration.

Didn't see your reply - but terraform can do a bit in the space of multi-cloud provisioning. I say a bit, because most of the work will be on the user, as your terraform plans are gonna be -MESSY- to support multi-cloud. It really isn't seamless or easy in my experience. AWS and Azure operate with enough differences that you'll have more exceptions to your workflow than actual steps. I find multi-cloud to be burdensome and an awful experience in general sadly, even with the "best" tools.

I’m not new gen. I’ve worked in telcos providing 5 9’s infra availability. A datacenter is not just servers. There is a whole host of civil, mechanical and electrical engineering that plays into it.

Edit: also OSHA rules, cable splicing, how to handle ladders, etc, etc.

> A datacenter is not just servers.

If you buy colocation, it is tho.

So basically trading one provider for another

This has to stop at some point.

That's called colocation: you provide the servers, they provide the racks

The new term for this old concept is now being called, "Data Center as a Service".

Do you need people in the colo to wire stuff up or is that handled by the colo folks?

You drop off the box and they wire it up and give you remote console access. You can even schedule an on-site meeting with your equip as needed. Some colo partners will also keep replacement parts for you like if a disk or power supply goes bad.

As others have mentioned, there's usually a remote hands feature that you can contract out. If you wanted to, many colos will rack and wire up all your equipment for you, for a cost. It really depends on how much you're wanting to offload on to the provider. They might do it all for you, they might just give you a cabinet or cage with power and a network drops.

You can usually pay the data center staff for their time to push a button or connect a few wires. Seems expensive at first when they quote you 50€/15 minutes but often way cheaper than somebody driving there and going through security to do the same.

Unless you don't want strangers touching your assigned rack, then you need to bring the key.

In addition to the other comments, after it's up & running some colos will also offer "remote hands" assistance, like replacing a failed drive. In that case ideally the enclosure will or can be made to make that slot's LED red or something so they know exactly which drive to pull.

Came here to suggest this - I'd have first run the numbers if running on a desirable cost IaaS to see if that was a comfortable middleground. It still amounts to self hosting everything yourself, except for compute, networking, storage - the basics (which are deceptively difficult when you account for secure access). And you get nice APIs to automate with still. Don't get me wrong, it might end up being more expensive at a threshold still, but do you really want to do your own... everything? Load balancing, DNS? What a total waste.

That said, some hardware companies have APIs to do most of that stuff - a little human runs out to fulfill your request if need be in some cases I think. Still like... much to worry about.

Uh. That's how essentially how the entire internet is run.

Yes. I know, I work for a telco and have worked with central office switches and telco edge clouds. I know what goes into designing a data center to maintain 5 9’s availability. I don’t think it’s a level of complexity most companies need though.

So to your point it’s how the internet is run. But your midsize enterprise isn’t a telco. They have no expertise with this stuff.

> 9 cents per gigabyte out if you’re lucky - that’s a startling price.

For plain S3. Other methods have lower prices. All methods have automatic volume discounts and almost everyone has a free tier.

> Machines are old and out of date and left behind.

A problem I've seen on every single VM stack, regardless of where it is hosted.

> Vendor lock-in is a real problem .. you probably can’t exit even if you want to.

Build your model first, then map it into services, don't start with the services and map your model based on them.

> And worse, clouds often don’t even have the computing resources you need with “requesting quota” and waiting for your “quota” to be granted some time in the next 24 hours, the opposite of elastic and scalable.

I've only experienced this on services that have a potential for abuse. DNS and SMTP being all that I can remember.

> Or buy a couple of fast fiber connections and buy some 16/32 core machines with 128 gig Ram.

And "requesting quote" isn't a delay item here? I don't know anything about fiber, how to price it, or how to receive it... and the capitalization process on all this is going to eat up a bunch of my departments time.

> And honestly it’s not hard to run modern Linux machines, that’s an old crones tale.

It's much easier to just run a lambda. It's a great compute environment. It does 90% of everything I've ever needed.

Yep, Azure has a 20 Public IP default quota which is blocking progress on a project.

requesting quota: poking the SRE or sysadmin to troll up and down the aisles of the colo datacenter sometime this week to find a QEMU hypervisor thats underprovisioned to the tune of $cores/$ram you want.

On AWS the default quota for GPU instances is ZERO.

Recently I had to do some performance testing. In the past I would have quickly started 5 or 6 instances of different types, run my performance tests, shut them down and be done within 2 hours.

Since AWS introduced zero quota GPU instances, I had to request quota for one specific GPU instance type in one region and then waited 24 hours till it was “approved”.

At that point I completely gave up on AWS as being a viable part of any GPU architecture.

It gets no less scalable or elastic than that.

I literally could have gone to the shop and bought a fleet of machines and built them at home in the time it took AWS to approve a quota of one GPU instance in one region.

That's bitten me too. So annoying.

we have vcenter these days...

But then you'd be back to paying pretty much cloud-like costs for your infra. The vmware suite (and similar tools) pricing is often not taken into account when comparing on-premise to cloud, which I find pretty ironic considering how much we talk about the hidden costs of the cloud. And the (crazy high) costs are usually tied to the numbers of cores/sockets you run, so not necessarily cheaper to scale.