Thats kind of what I am doing at the moment and i enjoy writing those scripts but i get so nervous running them. Even though I’m really confident they won’t break something i still feel I’m unlucky enough for the script to accidentally delete our prod dB.
Lol, same. I spend more time writing guard rails, setting up verbose debugging output/logs, figuring out how I can test without blowing up shit, backing up data before, backing up what it’s going to do, and then trying to set up automated confirmation of success or failure than I do coding the purely functional parts.
I’ve run the “person has quit, yeet their access” script I made on people days early more than I should admit. Had to put in a lot of extra checks on that.
If you’re at a place still using VMWare, a tip: if you’re trying to automate shutting down all the VMs for some hardware moves (get list of VMs, send Guest OS shutdown command, wait a certain amount of time, if VM still showing online force shutoff), VSphere will return the VM management devices/servers/whatever they’re called (what vSphere runs on) in the list of VMs, and it will accept Guest OS shutdown commands sent to the thing hosting VM management/vSphere. Halfway through shutting it all down for a move I started getting “cannot reach VSphere” errors. Added an extra hour or two to that weekend project as we had to get into things through some other back end shanigans to finish shutdowns manually.
Thats kind of what I am doing at the moment and i enjoy writing those scripts but i get so nervous running them. Even though I’m really confident they won’t break something i still feel I’m unlucky enough for the script to accidentally delete our prod dB.
Lol, same. I spend more time writing guard rails, setting up verbose debugging output/logs, figuring out how I can test without blowing up shit, backing up data before, backing up what it’s going to do, and then trying to set up automated confirmation of success or failure than I do coding the purely functional parts.
I’ve run the “person has quit, yeet their access” script I made on people days early more than I should admit. Had to put in a lot of extra checks on that.
If you’re at a place still using VMWare, a tip: if you’re trying to automate shutting down all the VMs for some hardware moves (get list of VMs, send Guest OS shutdown command, wait a certain amount of time, if VM still showing online force shutoff), VSphere will return the VM management devices/servers/whatever they’re called (what vSphere runs on) in the list of VMs, and it will accept Guest OS shutdown commands sent to the thing hosting VM management/vSphere. Halfway through shutting it all down for a move I started getting “cannot reach VSphere” errors. Added an extra hour or two to that weekend project as we had to get into things through some other back end shanigans to finish shutdowns manually.