fix(server): set port before instantiating server

[joeldenning]

• This is a bugfix
• This is a feature
• This is a code refactor
• This is a test update
• This is a docs update
• This is a metadata update

For Bugs and Features; did you add new tests?

Yes

Motivation / Use-Case

This resolves #2142.

Breaking Changes

No breaking changes.

Additional Info

The reason this fixes the bug is that we are setting options.port before calling new Server() instead of afterwards. The constructor in Server calls function that need the port to already be present.

[codecov]

Codecov Report

Merging #2143 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #2143   +/-   ##
=======================================
  Coverage   93.91%   93.91%           
=======================================
  Files          33       33           
  Lines        1265     1265           
  Branches      361      361           
=======================================
  Hits         1188     1188           
  Misses         71       71           
  Partials        6        6

---

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 f80e2ae...f962dc2. Read the comment docs.

[alexander-akait]

/cc @hiroppy @Loonride what do you think guys?

[knagaitsev]

This solution looks good to me, just one note.

@evilebottnawi Unfortunately, this problem will make it difficult, or maybe impossible, to move findPort into the API for the CLI refactor.

[joeldenning]

The failed tests are related to a "page crashed" error that seems to be unrelated to my code changes.

[joeldenning]

@Loonride @hiroppy anything else before this is good to merge?

[knagaitsev]

Looks good to me.

[joeldenning]

Anybody know why one of the azure checks is still pending? When I go to Azure it says that everything passed. Is it because a commit on master isn't merged into this branch?

I'm trying out merging master into this branch again.

[knagaitsev]

@joeldenning It looks like a Page crashed issue. Not your fault, there is still some instability with the e2e tests here. Your PR can still be merged if a couple CI builds are failing/timing out for this particular reason since your changes are not causing it.

[joeldenning]

All checks are now passing, after merging master in to retrigger the build.

Time to merge?

[mistic]

@evilebottnawi could you merge this?