fix(sql/testing): use wait.ForSQL when readying cockroach

[GeorgeMac]

I've noticed connections on cockroachdb attempting to apply migrations are getting stuck a lot during testing.

I'm unsure if this will help. However, I decided to attempt and migrate to the wait.ForSQL pattern for cockroach.
Just in case there is some kind of race when cockroach HTTP server is ready but SQL is in a bad state.

Example: https://github.com/flipt-io/flipt/actions/runs/3578054959/jobs/6017766294

[codecov-commenter]

Codecov Report

Merging #1183 (6237254) into main (518ec32) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1183   +/-   ##
=======================================
  Coverage   78.83%   78.83%           
=======================================
  Files          35       35           
  Lines        2528     2528           
=======================================
  Hits         1993     1993           
  Misses        435      435           
  Partials      100      100           

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[GeorgeMac]

Hmm didn't help.

Going to push up a change here to see the log output from cockroachdb.

[markphelps]

I was wondering if the cockroach image changed, like if we were using latest but it seems like we haven't updated the image being used in awhile cockroachdb/cockroach:latest-v21.2

[GeorgeMac]

Yeah, its weird. I've literally run this about 10 times now in CI. and run it locally with -count 5 numerous times.
Can't get it to fail 😂 Tempted to merge this change as-is. Then, when it next occurs we will have context, what do you think @markphelps ?

[markphelps]

Do it to it