Hacker News Reader

Comment by rglover | Hacker News Reader

You can accomplish a lot by just having a basic knowledge of Linux sysadmin. I was clueless and then learned some systemd-and-curl-fu. Will never forget the "holy sh*t, this is deceptively simple" moment. A bit more research and I found that beyond convenience and specialty APIs, you really just don't need a lot of this stuff to run a healthy system (since reducing absolute cloud dependence, my reliability has gone through the roof).

View on HN

100%. I'm not really sure why we all agreed that deployment is somehow the hardest thing that you need to outsource when setting the linux server is one the richest experience you can get and it will pay dividents forever.

Stop with this "we" cliche already, please. I never agreed to it and I'm in the profession for 24 years. You don't speak for me.

Executives will always prefer to transfer liability and responsibility to someplace else.

Who's calling the shots in an organization? Engineers or executives?

There's a multibillion dollar industry that lives only because they managed to successfully convince an entire generation of "engineers" to become helpless and not be able to serve an HTTP response using their own hardware even if their life depended on it.

It’s easy to convince developers of a thing if it starts with: “you don’t need to learn X anymore”

Just wait until you learn about system tools like perf, gdb, bpf -- the amount of low-level detailed information you can get about running processes means you'll reduce the amount of guesswork involved with troubleshooting or performance optimization to a minimum.

How do you deal with the few minutes of downtime when you do kernel/OS/software upgrades?

I’m pretty sure for most systems that does not matter in the slightest.

How is this an issue in a world where load balancers exist? I was part of a Unicorn that ran prod on 8 boxes and literally never had customer facing outages due to infrastructure updates.

What was the most stressful part of that job?

Depending on the deployment and any SLAs, I either don't worry about it (just do a late night rollout when nobody is on the system) or rely on my deployment architecture's sibling checks (I can see when a given machine is still versioning and requeue subsequent rollouts to other machines).

You spin up a second host and load balance

You put nginx or Haproxy in front of the hosts, drop the one that needs maint from the pool, and re-add once it’s ready.

Reply on news.ycombinator.com