We had quite a big issue with our azure web app service. All of a sudden we started getting 'disk out of space' errors. Sure enough, the LogFiles directory was huge. We weren't sure how to clear this directory. After a while and without finding any recommended way of doing it, we figured that it was just log files and it was probably safe to just delete a few files in there.
Via the kudu powershell console, we deleted a dump file LogFile/dumps and we also deleted a bunch a stdout files via a rm stdout_*.log
command.
Shockingly, this completely screwed our slot!
We tried to redeploy but we get getting HTTP 503 errors and there was nothing obvious to us on how to fix this.
Luckily, we deleting these files on the staging slot and not production so there was no downtime. We ended spinning a brand new slot and deploying there, and then deleting the previous slot.
Not a great experience for sure. Can anyone enlighten me about what might have happened? We have a pretty simple asp.net core 2.1 app.
Can deleting log files really mess up a slot???!