Collapse DB migrations that have already been released (pre-ATCv6)

[shamrickus]

What does this PR (Pull Request) do?

• This PR fixes Database migrations should be collapsed for ATC 6 #6027

Which Traffic Control components are affected by this PR?

• Traffic Ops

What is the best way to verify this PR?

Run the db/admin tool on master and on this fork and compare the DB schema and data to ensure they are the same.
I used pg_dump to fetch the data from the DBs and diff to compare the output.

The following criteria are ALL met by this PR

• This PR includes tests OR I have explained why tests are unnecessary
• This PR includes documentation OR I have explained why documentation is unnecessary
• This PR includes an update to CHANGELOG.md OR such an update is not necessary
• This PR includes any and all required license headers
• This PR DOES NOT FIX A SERIOUS SECURITY VULNERABILITY (see the Apache Software Foundation's security guidelines for details)

Additional Information

My diff:

> diff tops tops2
2685,2726c2685,2698
< 1	0	t	2021-07-28 08:22:43.182007
< 2	20181206000000	t	2021-07-28 08:22:43.185251
< 3	20190219000000	t	2021-07-28 08:22:43.186341
< 4	20190319000000	t	2021-07-28 08:22:43.187132
< 5	20190513000000	t	2021-07-28 08:22:43.188108
< 6	20191004000000	t	2021-07-28 08:22:43.191913
< 7	20191005000000	t	2021-07-28 08:22:43.195014
< 8	20191024000000	t	2021-07-28 08:22:43.202873
< 9	20191215000000	t	2021-07-28 08:22:43.206729
< 10	20200218000000	t	2021-07-28 08:22:43.207674
< 11	20200227000000	t	2021-07-28 08:22:43.209716
< 12	20200313000000	t	2021-07-28 08:22:43.210601
< 13	20200408000000	t	2021-07-28 08:22:43.212077
< 14	20200422101648	t	2021-07-28 08:22:43.215578
< 15	20200507000000	t	2021-07-28 08:22:43.231749
< 16	2020061622101648	t	2021-07-28 08:22:43.241121
< 17	2020062923101648	t	2021-07-28 08:22:43.24213
< 18	2020072700000000	t	2021-07-28 08:22:43.308803
< 19	2020081000000000	t	2021-07-28 08:22:43.311069
< 20	2020081108261100	t	2021-07-28 08:22:43.3147
< 21	2020082000000000	t	2021-07-28 08:22:43.315934
< 22	2020082700000000	t	2021-07-28 08:22:43.316662
< 23	2020103000000000	t	2021-07-28 08:22:43.319846
< 24	2020110200000000	t	2021-07-28 08:22:43.320716
< 25	2021010900000000	t	2021-07-28 08:22:43.322472
< 26	2021010900000001	t	2021-07-28 08:22:43.323647
< 27	2021010900000002	t	2021-07-28 08:22:43.327534
< 28	2021010900000003	t	2021-07-28 08:22:43.328402
< 29	2021012200000000	t	2021-07-28 08:22:43.331042
< 30	2021012700000000	t	2021-07-28 08:22:43.332001
< 31	2021012800000000	t	2021-07-28 08:22:43.333272
< 32	2021020500000000	t	2021-07-28 08:22:43.33405
< 33	2021021900000000	t	2021-07-28 08:22:43.334882
< 34	2021022300000000	t	2021-07-28 08:22:43.338818
< 35	2021030300000000	t	2021-07-28 08:22:43.342562
< 36	2021031400000000	t	2021-07-28 08:22:43.344308
< 37	2021040819481469	t	2021-07-28 08:22:43.349096
< 38	2021041200000000	t	2021-07-28 08:22:43.349876
< 39	2021052400000000	t	2021-07-28 08:22:43.351517
< 40	2021061100000000	t	2021-07-28 08:22:43.355434
< 41	2021070600000000	t	2021-07-28 08:22:43.35952
< 42	2021070800000000	t	2021-07-28 08:22:43.360362
---
> 1	0	t	2021-07-28 11:46:25.799758
> 2	2021012700000000	t	2021-07-28 11:46:25.802868
> 3	2021012800000000	t	2021-07-28 11:46:25.804771
> 4	2021020500000000	t	2021-07-28 11:46:25.805634
> 5	2021021900000000	t	2021-07-28 11:46:25.806486
> 6	2021022300000000	t	2021-07-28 11:46:25.810273
> 7	2021030300000000	t	2021-07-28 11:46:25.813885
> 8	2021031400000000	t	2021-07-28 11:46:25.815709
> 9	2021040819481469	t	2021-07-28 11:46:25.820572
> 10	2021041200000000	t	2021-07-28 11:46:25.821395
> 11	2021052400000000	t	2021-07-28 11:46:25.823174
> 12	2021061100000000	t	2021-07-28 11:46:25.826771
> 13	2021070600000000	t	2021-07-28 11:46:25.831382
> 14	2021070800000000	t	2021-07-28 11:46:25.832155
3105c3077
< SELECT pg_catalog.setval('public.goose_db_version_id_seq', 42, true);
---
> SELECT pg_catalog.setval('public.goose_db_version_id_seq', 14, true);

[shamrickus]

Topology index is incorrect, see #6066

[shamrickus]

phys_location index has a typo, see #6066

[zrhoffman]

Verified that

• create_tables.sql can still be run twice, and
• A DB dump with collapsed migrations has the same sha256 sum as a DB dump from apache/master without collapsed migrations:

  (set -o errexit -o nounset;
  create_migrate_dump() {(
      create_migrate_dump() {
        trap 'echo "Error on line ${LINENO} of ${0}" >/dev/stderr; exit 1' ERR;
        set -o errexit -o nounset -o pipefail;
        {
          docker-entrypoint.sh postgres >/dev/stderr &
          sleep 10;
          <create_tables.sql psql;
          <create_tables.sql psql;
          for migration in migrations/*; do
            <"$migration" sed '/+goose Down/,$d' | psql;
          done;
        } >/dev/stderr;
        pg_dump | sed '/^-- Dumped/d';
      };
      docker run --rm -iw/db \
        -v "$(pwd)/traffic_ops/app/db:/db" \
        -e PGUSER=traffic_ops \
        -e PGPASSWORD=twelve \
        -e POSTGRES_USER=traffic_ops \
        -e POSTGRES_PASSWORD=twelve \
        postgres:13-alpine bash -c "$(type create_migrate_dump | tail -n+2); create_migrate_dump";
  )};
  for remote_branch in apache/master shamrickus/to/6.0-migration; do
    git fetch "${remote_branch/\/*}";
    git checkout "${remote_branch}";
    create_migrate_dump >"${remote_branch/\/*}.sql";
  done;
  diff -u master.sql shamrickus.sql;
  openssl sha256 master.sql shamrickus.sql)