Fix etcd startup

[Confushion]

After applying the fix for issue #1206, timing/race issues were still showing when creating etcd clusters with more than 1 instance.

This patch fixes the race condition by having etcdadm-reconfigure.service depend on service var-lib-etcd2.mount instead of format-etcd2-volume.service (var-lib-etcd2.mount already depends on that service).

Also a potential circulair dependency on cfn-etcd-environment.service has been eliminated.

[codecov-io]

Codecov Report

Merging #1270 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1270   +/-   ##
=======================================
  Coverage   36.63%   36.63%           
=======================================
  Files          63       63           
  Lines        3882     3882           
=======================================
  Hits         1422     1422           
  Misses       2242     2242           
  Partials      218      218

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 185ccc5...889e48b. Read the comment docs.

[mumoshu]

@Confushion Good catch! This change does make sense to me.
Just wanted make sure, but did you actually get to bring new clusters with this change?
Anything you want me to test by hand?
Thanks anyway for your contribution 👍

[Confushion]

Yes, I managed to bring up 2 new clusters with 3 etcd instances each.

Although still with setting ‘waitSignal.enabled: false’, but that seems to be a separate issue...

[mumoshu]

@Confushion Thanks for the confirmation.

So, are you still seeing issues when the wait signal is enabled?
Would you mind sharing logs(even a full output from journalctl can be helpful), statuses or anything you think important?
I'd like to sort these out, too!