Effects of Clock Drift

I am in the IT department at our organization. According to our engineers and developers, they are attributing some issues in the plant to clock drift that happens each night when the VM running ignition backups. We are using Veeam. I believe the 3 second pause or so is caused by the snapshot being taken and/or merged back. I asked them if they were saying we couldn’t tolerate a 3 second pause once a night and this is their response.

“It doesn’t happen every time, but this clock drift can put tags into some kind of erred state that they won’t update again until manually toggled or updated. It affects every app on the gateway and could potentially shut the plant down if it effected enough tags and someone isn’t available that is aware of what is happening and how to fix it.”

My question should this be the case? If it is, then how do we backup the VM? We already use the scheduled backups of the gateway, but I would like to maintain the VM as well. The longest drift has been 4425ms, but generally its less than that. often 1.5-3 seconds. I dont believe a couple of seconds once or twice a night should be this problematic. Is this normal? Or is it bad application design or architecture?

Appreciate any feedback.

I think they’ll just need to work (keep working?) with support to figure out what’s happening. This isn’t normal.

A 3-5 second pause could conceivably cause some red overlays if that was long enough for the tags to be considered “stale” or for a connection to time out, but they should come back to good quality afterwards.