I’ve moved these scripts to github. More precisely, I’ve done away with these scripts which I haven’t been using for Waterpoint for quite a while and instead put the scripts I currently use there.
The new scheme involves setting up the checkpoint to happen at predictable times and then having cron kick off a backup script shortly after the checkpoint happens.
This is easier to get right, easier to test, and has fewer moving parts than the old scheme.
There’s a README there.
The old scheme involved a script to tail the MOO log file and kick off the backup script whenever it saw a line indicating that a checkpoint had just completed.
In truth, you may want something tailing the MOO log, but it should be alerting when checkpoints FAIL rather than depending on that to trigger a backup.