I'm running Mint 18.2 (Mate) in a Lenovo 100-15IBD, and before upgrading to 18.3 I installed Timeshift to create a system snapshot.
It started working in the background and some hours later a pop-up window said that my disk had only 500 mb left. (I have only 1 HDD, with 500 Gb, with about 207 gb free).
I tried to execute Timeshift to see what was happening, but I received a message that the program was already running. I tried to kill the application (command 'killall), without success. Tried to erase the snapshot directory (/timeshift/snapshot/2017-12-12-etc), but it said the operation wasn't allowed. Then I made logoff and shut down the system, to try to stop the process.
The problem was that, as the disk was full, I was unable to login. Fortunately, in the forum ( https://forums.linuxmint.com/viewtopic.php?f=90&t=253502&p=1364448&hilit=disk+full#p1364448 ), Rene said to someone with a similar problem to set the root-reserved percentage on the root filesystem from 5% to 0%. And that worked! I was able to login and then manually delete the directory.
I'm not an expert in Linux, but it seems that Timeshift has no instruction to stop before completely filling a disk, so it goes and goes until everything screws up, and you can't login anymore. I've noticed some more 1-2 similar issues in the forum.
Sorry if this seems a little stupid for more experienced users, but for me, as a newbie, was quite unexpected.
I'm running Mint 19 in VirtualBox, and I have a 20 GB virtual disk. After several weeks, it stopped working and would only boot to tty1. It turns out that /dev/sda1 was 100% full, and /timeshift was using about 9.8 GB (half the disk). As soon as I deleted the snapshots, everything started working again (in fact, the login screen immediately appeared, even before I logged out tty1).
So I agree with the original bug description: Timeshift should be aware of how much disk space it's using and avoid filling up the disk.
See also:
https:/
https:/