Diffrence between gateway backups made manually and automatically?

I have a 8.1.17 gateway that takes automatic backups. We have a backup VM (in case production goes down) that we routinely restore these auto backups to that sometimes give us an error and then we manually back up production and restore to the gateway on the backup VM.

Our IT person extracted both an automatic backup and a manual backup to try to figure out why the automatics were failing and found the following difference - automatic on the left, manual on the right -

Are these two files important? Would the be related to this error we often get when trying to restore an automatic backup?

They are absolutely important. user-lib will hold your Python standard library and all your JDBC drivers. backupinfo is a sentinel file for the rest of the system with some metadata.

Is anything being logged by your gateway when the scheduled backups happen? They go through exactly the same code path (there’s only one mechanism in the code to generate a .gwbk, precisely to avoid mixed results).

Any chance virus software could cause this?

I can turn on the automatic backups to be more frequent and get a log of the event. Is there a particular logger I should be setting to trace or similar?

There’s not really any trace or fine grained logging. If there’s anything, it’ll be errors from com.inductiveautomation.ignition.gateway.servlets.BackupServlet.

Interestingly it seems to work now with the user-lib and backupinfo.xml coming through.

Our IT guy rolled back a recent windows update and he thinks it’s working now because of the rollback. Here is the specific update