Comment 3 for bug 1891618

We got a few reports about this problem over last couple of months but so far haven't been able to reproduce it (I tried several times in a VM with network traffic shaping, simulating slow network with packet losses) nor pinpoint a problem. It is under investigation but very elusive, any extra info may help.

If anyone hits this issue, then the following information may help:
- `snap changes` and `snap change <X>` output
- journalctl -u snapd output (ideally, running snapd with SNAPD_DEBUG=1 environment variable set when the problem occured).
- `systemctl status snapd` output
- information on the quality of network in use (stability, whether it drops frequently etc.)