The last NixOS upgrade stalled on every borgbackup restart

During the current NixOS upgrade, the borgbackup daily job failed to restart.

For now just click a couple of times more. From the error logs, it looks like it could just be the internet weather. Investigate & fix if this happens again.

For example this deployment:

warning: the following units failed: borgbackup-job-daily.service
× borgbackup-job-daily.service - BorgBackup job daily
     Loaded: loaded (/etc/systemd/system/borgbackup-job-daily.service; linked; preset: ignored)
     Active: failed (Result: exit-code) since Mon 2025-06-16 12:31:10 UTC; 10s ago
   Duration: 1.013s
 Invocation: 13b76ce1a7664f53b678943b863237ee
TriggeredBy: ● borgbackup-job-daily.timer
    Process: 1667742 ExecStart=/nix/store/kssda9ymvmhzc6gblyjjq2fd35m41pnk-unit-script-borgbackup-job-daily-start/bin/borgbackup-job-daily-start (code=exited, status=2)
   Main PID: 1667742 (code=exited, status=2)
         IP: 0B in, 0B out
         IO: 0B read, 300K written
   Mem peak: 66.7M
        CPU: 980ms
Jun 16 12:31:09 storage002 systemd[1]: Started BorgBackup job daily.
Jun 16 12:31:10 storage002 borgbackup-job-daily-start[1667753]: {"type": "log_message", "time": 1750077070.7625299, "message": "Connection closed by remote host. Is borg working on the server?", "levelname": "ERROR", "name": "borg.archiver", "msgid": "ConnectionClosedWithHint"}
Jun 16 12:31:10 storage002 systemd[1]: borgbackup-job-daily.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
Jun 16 12:31:10 storage002 systemd[1]: borgbackup-job-daily.service: Failed with result 'exit-code'.
Jun 16 12:31:10 storage002 systemd[1]: borgbackup-job-daily.service: Consumed 980ms CPU time, 66.7M memory peak, 300K written to disk.
Error while activating new configuration.
+ echo '2025-06-16T12:23:19+00:00 7b91af851b58ea1cbdac517b13add48af54fff7e FAIL'
+ exit 1
+ ```