Reduce the log level of wickedd-dhcp6

[MiLk]

All our bottlerocket nodes seem to be generating 34k~56k log entries per day for wickedd-dhcp6.
This is quite a sizable proportion of our total log volume.
I can't see what is wrong with our setup, so I was wondering if it would be possible to reduce the log level or at least make it configurable.
The current log level is configured with --log-level debug --debug most.

What I'd like:
Be able to reduce the log level of wickedd-dhcp6.

Any alternatives you've considered:

• Using a bootstrap container, update the systemd service file, and restart the service.
  But I couldn't figure out how to interact with systemctl from the bootstrap container.

• Fix whatever is wrong with our setup, but I don't have any idea of how to proceed.
  One thing I can think of is that our VPC has IPv6 enabled and our nodes are starting in a subnet with assignIpv6AddressOnCreation enabled.
  I believe that even if the nodes have an IPv6, as the cluster uses IPv4, there is no point assigning IPv6 addresses, and it could be disabled, but I don't know if it would help.
  I'm attaching log files of 2 nodes, let me know if you have other ideas.

extract-2022-06-28T23_55_01.965Z.csv

extract-2022-06-28T23_47_48.393Z.csv

[bcressey]

Agreed - the IPv6 logs in particular are very chatty because the lease is renewed so often.

The intent was to provide at least enough of a marker in the host logs to correlate a DHCP event with a hypothetical application level issue happening at the same time, but it's definitely excessive right now.