Post AkUkNBoTfrekgNIdPs by miah@hachyderm.io
 (DIR) More posts by miah@hachyderm.io
 (DIR) Post #AkUkLqx7kQFth9hrs0 by miah@hachyderm.io
       2024-07-31T15:35:40Z
       
       0 likes, 0 repeats
       
       @arichtman I think VMWare ESX (or whatever they call it nowadays) or Proxmox are probably all 80% of the people using k8s need.Maybe they need a wrapper (like Hashicorp Packer) to generate the containers in a automated fashion too.I think the only "big change" in tech from ~90's-00s to where we are now is software is even more bloated, so maybe it needs "more resources". All the other problems are exactly the same though.
       
 (DIR) Post #AkUkNBoTfrekgNIdPs by miah@hachyderm.io
       2024-07-31T15:58:02Z
       
       0 likes, 0 repeats
       
       @arichtman @pete_wright I think the value is mostly for the business. It lowers the bar for who you can hire as a sysadmin. All you need to know is how to write a helm chart and run a few commands. Understanding how anything actually works is optional.
       
 (DIR) Post #AkUkQL8PTQSwSf7o3c by feld@bikeshed.party
       2024-07-31T16:20:52.272979Z
       
       0 likes, 0 repeats
       
       @miah @pete_wright that's my thought too. It makes your entire IT dept replaceable with a new batch of idiots. Nobody will really understand anything enough to debug it, but they'll be able to modify those helm charts!
       
 (DIR) Post #AkUl6iIVkESmSj3T1M by pete_wright@nlogic.systems
       2024-07-31T16:14:54.113636Z
       
       0 likes, 0 repeats
       
       @miah @arichtman that's a great point, i was looking for an experienced SRE for some contract work on AWS (automating some APIs) and our agency just sent me a bunch of senior devops people who had like three years of running heml charts.  they struggled to understand i needed someone with more than a superficial understanding of how distributed systems operate.