Fetch goal state before sending telemetry

[narrieta]

Some of the code that emits telemetry is relying on something else initializing the container ID (which comes from the goal state).

I made an explicit call to fetch the goal state in the extension handler.

For the daemon, provisioning must be completed before being able to fetch the goal state, so I moved some initialization code (whose telemetry we are interested in) after provisioning. As a result of this, resource usage would not be monitored/enforced for provisioning on non-systemd distros, but I think that is OK (in fact, I am considering whether enforcement should include provisioning on systemd distros).

There is still one telemetry event without a container ID. It is emitted at the end of provisioning; I am hesitant to touch that code at this point so I will leave it alone for now.

I verified this change using the telemetry emitted by DCR.

---

This change is 

[codecov]

Codecov Report

❗ No coverage uploaded for pull request base (develop@5b60264). Click here to learn what that means.
The diff coverage is 100%.

@@            Coverage Diff             @@
##             develop    #1751   +/-   ##
==========================================
  Coverage           ?   68.57%           
==========================================
  Files              ?       82           
  Lines              ?    11741           
  Branches           ?     1646           
==========================================
  Hits               ?     8051           
  Misses             ?     3346           
  Partials           ?      344

Impacted Files	Coverage Δ
azurelinuxagent/ga/update.py	88.46% <100%> (ø)
azurelinuxagent/daemon/main.py	73.68% <100%> (ø)

---

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 5b60264...9d6e574. Read the comment docs.

[larohra]

LGTM