Clock drifts constantly on macOS because fedora's pool is unavailable

[nicdoye]

Describe the bug The clock in the underlying VM on macOS drifts. Even over this weekend, it drifted nearly 2 mins.

Steps to reproduce Start finch on macOS, run a long running container. Wait a couple of days. Run date and see it is no longer accurate

Expected behavior I expect sub-second accuracy of the VM's clock.

Screenshots or logs Sample logs on the VM:

Oct 21 10:07:18 lima-finch chronyd[665]: Source 176.58.127.165 replaced with 178.62.18.76 (2.fedora.pool.ntp.org)
Oct 21 10:08:50 lima-finch chronyd[665]: Detected falseticker 129.250.35.250 (2.fedora.pool.ntp.org)
Oct 21 10:08:50 lima-finch chronyd[665]: Selected source 213.5.132.231 (2.fedora.pool.ntp.org)
Oct 21 10:08:54 lima-finch chronyd[665]: Can't synchronise: no majority
Oct 21 10:09:54 lima-finch chronyd[665]: Detected falseticker 129.250.35.250 (2.fedora.pool.ntp.org)
Oct 21 10:09:54 lima-finch chronyd[665]: Selected source 213.5.132.231 (2.fedora.pool.ntp.org)
Oct 21 10:10:00 lima-finch chronyd[665]: Can't synchronise: no majority
Oct 21 10:10:59 lima-finch chronyd[665]: Detected falseticker 129.250.35.250 (2.fedora.pool.ntp.org)
Oct 21 10:10:59 lima-finch chronyd[665]: Selected source 213.5.132.231 (2.fedora.pool.ntp.org)
Oct 21 10:11:04 lima-finch chronyd[665]: Can't synchronise: no majority

Additional context Replacing the fedora pool in /etc/chrony.conf with: pool time.aws.com iburst and running sudo service chronyd force-reload might fix the problem (it certainly synced the clock this morning, but only time will tell).