Comment 4 for bug 1927029

I suspect it might be an issue with snapd not being ready to execute snaps, this might help confirm.

Is this a persistent issue or does it eventually resolve with retries?

Did this happen early in the machine uptime, perhaps in cloud-init?