justindev

6d

Friday afternoon.
Boss: “Can we push this small fix before the weekend? It’s just a button color.”
Me: “Sure, what could possibly go wrong?”

Fast forward 20 minutes:

Whole CSS is missing
Login page is blank
Server panicked so hard it restarted itself
I’m now "that guy" who deploys on Friday

Moral of the story:
No fix is truly “small” on a Friday. :(

random

Ranter

Comments

8

tosensei

9059

6d

please - describe in detail _how_ you managed this screw up
5

Hazarth

9408

6d

@tosensei my bet is their CI/CD is fucked, or someone fucked it without deploying and OP is just the first one to clean up the mess.

That's my guess at least
2

Waytopranav

122

6d

Backup first, thats the first rule man what you are doing?
3

retoor

9188

6d

@Waytopranav just run the repo on production and let the CI update that. So comfy.
0

justindev

7

2d

@tosensei I deployed a small CSS change without realizing it was tied to a larger stylesheet that hadn't been updated in staging. I skipped the final QA step because I assumed the change was isolated — turns out it wasn’t. Lesson learned: always test even the "harmless" stuff.
0

tosensei

9059

2d

@justindev but how did a CSS change result in "server panicking"? and why are there even external files that aren't part of the deploy package?

devRant © 2021 Hexical Labs LLC
Privacy Policy | Terms of Service