Clean up run results

[kwigley]

resolves #2493

Description

This branches attempts to rationalize run result status handling.

• The resulting status of a run will be stored in a status field and related info in the message field. this is a breaking change
  Example:
  • success status -> status returned from database adapter stored in message
  • error status -> error message stored in the message field
• Instead of including the entire node obj in run results, we just write the unique_id. this is a breaking change
• Renames source freshness results state field to status. this is a breaking change

New schema can be viewed here: https://schemas.getdbt.com/dbt/run-results/v1.json this is definitely slimmer than the previous run results schema.

Checklist

• I have signed the CLA
• I have run this code in development and it appears to resolve the stated issue
• This PR includes tests, or tests are not required/relevant for this PR (I updated existing tests)
• I have updated the CHANGELOG.md and added information about my change to the "dbt next" section.

[kwigley]

This will affect model run event tracking data, I'm not sure if there is anything else to consider here.

[jtcohen6]

thanks for calling it out! we'll just need to account for this on our end

[gshank]

Lots of changes! Looks good. Nice and clean.

[jtcohen6]

This is looking really solid! The new run_results.json feels both more sensible and more concise. There are some small things I want to improve in the future (not this PR), mostly around a more structured version of the message (a la #2580 or #2747). Let's discuss #2747 this afternoon.

Three small things I've noticed in the CLI output:

1. When running a model that fails, the skip looks great, but the errant model itself reports the message instead of the status:

13:36:52 | 1 of 2 START view model dbt_jcohen.model_a........................... [RUN]
13:36:52 | 1 of 2 ERROR creating view model dbt_jcohen.model_a.................. [Database Error in model model_a (models/model_a.sql)
  syntax error at or near ";"
  LINE 7: ;aoushfs;dulfhsdlf
          ^
  compiled SQL at target/run/testy/models/model_a.sql in 0.12s]

I believe the same occurs with seeds as well.

2. When a test has a fail or warn status, the stdout "summary" now reports the statuses as their message:

Failure in test unique_model_a_col3 (models/whatever.yml)
  Status: fail

  compiled SQL at target/compiled/testy/models/whatever.yml/schema_test/unique_model_a_col3.sql

Warning in test unique_model_b_col3 (models/whatever.yml)
  Status: warn

  compiled SQL at target/compiled/testy/models/whatever.yml/schema_test/unique_model_b_col3.sql

Whereas previously it wrapped the message like so:

Failure in test unique_model_a_col3 (models/whatever.yml)
  Got 1 result, expected 0.

  compiled SQL at target/compiled/testy/models/whatever.yml/schema_test/unique_model_a_col3.sql

Warning in test unique_model_b_col3 (models/whatever.yml)
  Got 1 result, expected 0.

  compiled SQL at target/compiled/testy/models/whatever.yml/schema_test/unique_model_b_col3.sql

I'm already hoping to change that message output to be a bit more descriptive (and customizable!) per our proposed test changes in v0.20.0, so I'd prefer to keep it consistent for now.

3. When snapshotting source freshness, if there is a database error (i.e. malformed SQL syntax), we should surface that error in the stdout "summary". (This is distinct from the error produced when a query completes successfully and finds stale sources.)

This PR:

14:52:45 | 1 of 1 ERROR freshness of dbt_jcohen.my_seed......................... [ERROR in 0.07s]
14:52:46 | Done.

Current behavior (v0.18.1):

14:54:27 | 1 of 1 START freshness of dbt_jcohen.my_seed......................... [RUN]
14:54:27 | 1 of 1 ERROR freshness of dbt_jcohen.my_seed......................... [ERROR in 0.04s]

Database Error in source my_seed (models/whatever.yml)
  Expected a timestamp value when querying field 'col3' of table "jerco"."dbt_jcohen"."my_seed" but received value of type 'str' instead
14:54:27 | Done.

[jtcohen6]

I did a quick tour of seed, run, test, snapshot, source snapshot-freshness, and docs generate. All CLI output + run results are looking good. Nice work!