Ensure availability of backend vm during scheduled running of job

[zigouras]

Is your feature request related to a problem? Please describe.
Our spot vm gets preempted by Google and may be offline when it should be running the scheduled job. We have GCP send a Slack message to the backend channel when the preemption happens and someone has to manually restart the vm from the cloud console.

Describe the solution you'd like

1. An automated way of restarting the vm if it is preempted, or...
2. An automation to only run the vm for the hour a week it takes to run the job and then shut it down. This will also save money by not having the vm running when it is not being used.

GCP automation tools may allow us to do one or both of the above. It has to be researched. Team leads will still be able to restart and ssh into the vm if needed to work there.

Describe alternatives you've considered
Have a script running on a team member's personal machine that is always up that can do 1. or 2. above with the Google cloud sdk.

Additional context
Since we have not been loading new vacant properties data from the city since July, 2024 because the data is corrupted, this is less of a priority.

[github-actions]

This issue has been marked as stale because it has been open for 30 days with no activity.