Live data from Hacker News

Reclaiming the lost art of Linux server administration

pietrorea.com

281–290 of 485 posts

Re: Reclaiming the lost art of Linux server administration

#281
post #147

Earlier quoted context omitted.

Of course there are a lot of benefits of using hosted databases. I like hosted databases and use them for both work and personal projects. What I have a problem with is: - the premium over bare metal is just silly - maximum vertical scaling being a rather small fraction of what you could get with bare metal - when you pay for a hot standby you can't use it as a read only replica (true for AWS and GCP, idk about Azure…

Readable standby instances is something AWS just recently added actually: https://aws.amazon.com/blogs/database/readable-standby-insta... Though it seems to require you to have 3 instances rather than just letting you read from the standby... I don't quite get the rationale for that.

Even though a hot standby is technically the exact same thing as a read only replica, it solves 2 very different problems (uptime vs scaling). Personally, I never want read only queries going to my hot standby, otherwise I risk the secondary can’t handle the load in case of failover. Probably a reason like that with AWS as well, as they are selling a “hot standby” product to ensure the uptime for your RDS.

Re: Reclaiming the lost art of Linux server administration

#282
post #61

I have over 20 years of Linux/FreeBSD sysadmin experience ranging from universities to major silicon valley companies in both cloud and on-prem. When it comes to companies I mostly support cloud these days but when it comes to me and my family I accept every downside and host as almost all of our digital lives in a 42u rack in a gutted closet in our house with static IPs and business fiber. I know where our data live…

I have done pretty much the same as you, except that instead of an on-premises 42U rack of stuff, a lot of my personal and family infrastructure lives in a single beefy 1U server (32 cores, 512GB of RAM, set up as a hypervisor) in colocation at a major IX point.

Then there's a set of about six or seven $4 per month KVM VM at geographically distributed offsite 3rd party hosting companies for things like secondary authortative nameservers and such.

And an offsite backup system that mirrors everything on the system in colocation to a medium sized disk array that lives in the corner of a closet in a family member's basement.

Re: Reclaiming the lost art of Linux server administration

#283

A simple LAMP stack is a good place to start, and a single droplet from DO or similar is plenty to play around with and standup a real internet-facing system. Bonus points for starting vanilla and installing & configuring the AMP portion from scratch, which is an excellent primer on the basics of system config & admin. Get your hands a little dirtier installing a lightweight desktop environment like LXDE, programming…

oracle cloud's free tier is equivalent to DO's $5 droplets (they also have a second free tier offering with ARM cpu's, with up to 4 vcpu/24gb memory for free(!), but it's high-demand and difficult to get a slot)

Re: Reclaiming the lost art of Linux server administration

#284

Earlier quoted context omitted.

Here's what the bare metal server didn't come with: API access for managing configuration, version updates/rollbacks, and ACL. A solution for unlimited scheduled snapshots without affecting performance. Close to immediate replacement of identical setup within seconds of failure. API-managed VPC/VPN built in. No underlying OS management. (Probably forgot a few...) I get that going bare metal is a good solution for som…

Everything has a cost. The cost of the cloud is that you’re out of control, the cloud provider owns your stuff, and a single vulnerability can make half the Earth get pwned. Whoops. EDIT: I do use cloud services for some stuff. My point isn’t being anti-cloud, just that nothing is perfect.

True, and also it costs more money.

Re: Reclaiming the lost art of Linux server administration

#285

A simple LAMP stack is a good place to start, and a single droplet from DO or similar is plenty to play around with and standup a real internet-facing system. Bonus points for starting vanilla and installing & configuring the AMP portion from scratch, which is an excellent primer on the basics of system config & admin. Get your hands a little dirtier installing a lightweight desktop environment like LXDE, programming…

oracle cloud's free tier is equivalent to DO's $5 droplets (they also have a second free tier offering with ARM cpu's, with up to 4 vcpu/24gb memory for free(!), but it's high-demand and difficult to get a slot)

I'm afraid to touch anything Oracle for fear that a fleet of SUVs filled with teams of software auditors & a mobile law firm will show up on my doorstep with a $5M invoice and a lein on my house.

Re: Reclaiming the lost art of Linux server administration

#286
post #94

Blame the folks demonizing/shaming having "pet" servers and pushing immutable infrastructure. Linux server administration is quite enjoyable, and with how well apps these days can scale vertically, it really takes a special kind of workload to need (and actually saturate) fleets of servers.

In my experience Pet servers are a good starting point (you really should _graduate_ from Pet servers into all the various immutable/cattle stuff), but it can quickly require discipline from the Admins. They can't be doing one-off undocumented config, package, and network/firewall changes which make it impossible to setup another server reliably. At $company I moved us to Terraform+Packer (to get them used to immutab…

Do you have any tips in auto installing an OS on a server/desktop?

I'm completely missing. I have searched arpund and I have some solutions, but back in my head, some people have something else.

The part before Ansible or puppet to kick in.

Re: Reclaiming the lost art of Linux server administration

#287

Earlier quoted context omitted.

I think most folks would argue "cattle" means using imaging to manage/replace your fleet. Using something like puppet or ansible against a fresh install implies a level of individualism towards each system as they "may" have minute details based on when puppet/ansible ran, even if they're part of a dynamic inventory of some sort.

I'm not following, sorry. This is cattle: * PXE boot server(s) * Image contains masterless puppet bootstrap. * Server(s) asks git - "give me the bootstrap for my mac address" * Server(s) gets a list of classes to apply. * Server(s) applies classes. Done.

What kind of software do you use to configure PXE servers? Or is there anything else that I'm missing?

I'm looking for a solution to make unattend installation possible.

Edit: Can use Ansible to make a PXE server. (Egg or chicken thing)

Re: Reclaiming the lost art of Linux server administration

#288
post #167

Earlier quoted context omitted.

People who have been woodworking for decades can own very expensive tools so that they can create very complicated things. People who are experts in cars can own very expensive cars and tools to tune them. People who have been working in music can have very expensive instruments and expensive headphones, microphones, sequencers, etc. We seem to be looking down on experienced "computer experts" and wanting to take the…

bruh nobody's looking down on people running they're own metal Server hardware is fun but it's not trivial to manage, buy or run. So when someone talks about how they've managed servers for 2 decades, own a house where they can install a 42 rack and how much better it is than a hosted solution. A lot of people rightly point out that this is hardly feasible for most people

I started self hosting when I was 12 with whatever last gen hardware at goodwill no one wanted.

Re: Reclaiming the lost art of Linux server administration

#289
post #86

Earlier quoted context omitted.

Yes but you have talent and a lifetime of experience, plus space for a noisy 42u rack full of servers, but not everybody does...

And a tolerant wife/spouse. My closet of RPi's are quiet.

yes

Re: Reclaiming the lost art of Linux server administration

#290
post #87

Earlier quoted context omitted.

I remember how surprised people were when I demoed a $200/month bare metal server outperforming by a huge margin RDS MySQL instance that they were paying something upwards of 16k/month. IIRC we ended up using it as a disposable replica for some non-real time but heavy operations.

Here's what the bare metal server didn't come with: API access for managing configuration, version updates/rollbacks, and ACL. A solution for unlimited scheduled snapshots without affecting performance. Close to immediate replacement of identical setup within seconds of failure. API-managed VPC/VPN built in. No underlying OS management. (Probably forgot a few...) I get that going bare metal is a good solution for som…

Also physical security, power, physical rack cost, networking, salaries to manage everything, and all those other hidden costs
Post reply on HN