Today, during deployment on server without remote access:

Me (on the phone calling our data centre Admin): "There's a permissions mismatch. The following paths need write access from the following users..."

Admin: "Okay, okay, slow down... I'm still in the elevator." - 10 minutes later - "Okay, ready."

And I gave him the paths and he said: "Try now."

And I tried and it still didn't work. And then we tried all that again. And again. And finally he said:

Admin: "Okay, I give up, I'm going back down to get the screen."

If you didn't facepalm at this point - he was working blind with just a keyboard. Didn't expect anything more than a reboot would be needed and screens are heavy.

:| .... If it’s a permissions mismatch and he’s blindly using chmod while he HOPES he’s in the right directory you have a serious problem and it isn’t the server issue

Needs wk174 because it is unprofessional

Good one 😁

LMAO!!!! Who needs screens anyway 😹

He needs this: https://github.com/Amt921/...
(It's a cool & "cheap" remote kvm thing mainly using a raspberry pi)

This has some serious Bitwise vibes... If it's not 100% mandatory for me to be validating the system is collecting my inputs properly (mstsc tends to drop or add keyboard presses) I just don't look at the screen... Laze back into my chair, stair at the ceiling and glance down for outputs

I'm totally guilty of this too haha

Screens really are heavy, and sometimes it's dark and bright lights hurt.

Wow that's really weird. Usually when I deploy on a server I built it I am the admin I am in total control. How it it possible to do your job if you have to deal with that?

On a production system this makes my spider senses tingle. If you’re trying to resolve an outage on a production system - use a screen and check your work. I don’t care how good you are at typing, people make mistakes. The difference between chmod -R 777 ./* and chmod -R 777 /* is one keystroke, and the consequences are fairly heavy if you fuck that up. Get a sysadmin who at least recognizes the seriousness of what they’re doing.

To @Diactoros @kescherRant @MagicSowap @S-falken @Diactoros and all other worried professionals, it all depends on circumstances. :D

1) ...it's a "semi-production" environment, meaning it's a brand new bare production server that hasn't been hooked up to any of the network KVM switches yet but it's only hosting an internal service that isn't vital, no other systems on that machine for now,

2) ...it auto-deploys with hooks after merge to `live` branch,

3) ...oh, he can `chmod` with his eyes closed and guess the output by the noise of RAID writes, only he had to change the ACLs,

4) ...worst case scenario he'd restore the image and run the commit-hooks again.

To everyone else:

Yes, it was ridiculously unexpected and made my day as well! XD

@cprn *removes stick from ass*