5 ms·
I work in IT and I happened to be the poor bastard on call when Clown Strike took out the majority of infrastructure. If it wasn't for my own personal refusal t
by Brett_Riverboat 2y ago
I work in IT and I happened to be the poor bastard on call when Clown Strike took out the majority of infrastructure. If it wasn't for my own personal refusal to use cloud based bullshit we would have probably been down for days instead of hours. The fact that people like my IT director saw nothing wrong with this and is taking 0 steps to negate such bullshit makes me quite worried that I will soon have to deal with some other catastrophic cloud based failure in the near future.
I keep repeating ad-nauseam "only idiots rely on other peoples computers" And I stand behind that statement 100%
- xyst 2y agoCan only do so much when idiot CTOs take their advice from CTO summits, consultants with their own perverse incentives, and of course random conferences
- system2 2y agoI 100% agree. Additionally, I am amazed to see that people can pay outrageous fees for cloud services such as Azure VD. For the fraction of the yearly cloud budget, companies can create crazy stable, offline-capable infrastructure themselves.
- cqqxo4zV46cp 2y agoYou sound like a very aggressive person. I don’t think I’d like to work with you regardless of whether or not you were right. Maybe you’d get your point across better if you weren’t so aggressive about it.
- Brett_Riverboat 2y agok
- citrin_ru 2y agoWhat strange that there are mangers/executives who understand why it’s bad to have a SPOF (single point of failure) in the internal infrastructure but who are OK with having a product/service from an external vendor as an SPOF. Like if you have a contract and are paying money for it means that it is created and maintained by infallible superhuman (as opposite to internal engineers they don’t trust). Such misplaced trust puzzles me.
- Brett_Riverboat 2y agoMy director mistakenly thinks the more it costs the better it is, he refuses to even consider anything FOSS. Before he came along we ran everything in house and we still had issues of course but down time was nearly nonexistent because we could take action immediately instead of waiting for whatever cloud service messed up today to feel like getting around to it. And of course we now get to pay monthly for the privilege of being at someone else's mercy as opposed to before when we paid once and went on our merry way.