Cockroach DB deleted columns

### Version Information

Server Version:
Docker Container - (hasura/graphql-engine:latest) [ v2.39.2 ]
Docker Container - (cockroachdb/cockroach:latest) [ v23.2.5 ]

### Environment

OSS

### What is the current behaviour?

Tables with deleted columns trigger error on the dashboard when trying to view the table or create a relationship with the table.

### What is the expected behaviour?

Not to show error and allow me to access the table when viewing it or creating foreign keys on another table

### How to reproduce the issue?

1. Create a cockroach db docker container for testing.
2. Create 2 tables that are related by a foreign key.
3. Delete a column from one of the tables.

### Screenshots or Screencast


### Please provide any traces or logs that could help here.

```
{
"arguments": [],
"error": {
"description": null,
"exec_status": "FatalError",
"hint": null,
"message": "has_column_privilege(): column 10 of relation store_integration does not exist",
"status_code": "42703"
},
"prepared": false,
"statement": "SELECT\n COALESCE(Json_agg(Row_to_json(info)), '[]' :: json) AS tables\n FROM (\n WITH partitions AS (\n SELECT array(\n WITH partitioned_tables AS (SELECT array(SELECT oid FROM pg_class WHERE relkind = 'p') AS parent_tables)\n SELECT\n child.relname AS partition\n FROM partitioned_tables, pg_inherits\n JOIN pg_class child ON pg_inherits.inhrelid = child.oid\n JOIN pg_namespace nmsp_child ON nmsp_child.oid = child.relnamespace\n where ((nmsp_child.nspname='private'))\n AND pg_inherits.inhparent = ANY (partitioned_tables.parent_tables)\n ) AS names\n )\n SELECT\n pgn.nspname AS table_schema,\n pgc.relname AS table_name,\n CASE\n WHEN pgc.relkind = 'r' THEN 'TABLE'\n WHEN pgc.relkind = 'f' THEN 'FOREIGN TABLE'\n WHEN pgc.relkind = 'v' THEN 'VIEW'\n WHEN pgc.relkind = 'm' THEN 'MATERIALIZED VIEW'\n WHEN pgc.relkind = 'p' THEN 'PARTITIONED TABLE'\n END AS table_type,\n obj_description(pgc.oid) AS comment,\n COALESCE(json_agg(DISTINCT row_to_json(isc) :: jsonb || jsonb_build_object('comment', col_description(pga.attrelid, pga.attnum))) filter (WHERE isc.column_name IS NOT NULL), '[]' :: json) AS columns,\n COALESCE(json_agg(DISTINCT row_to_json(ist) :: jsonb || jsonb_build_object('comment', obj_description(pgt.oid))) filter (WHERE ist.trigger_name IS NOT NULL), '[]' :: json) AS triggers,\n row_to_json(isv) AS view_info\n FROM partitions, pg_class as pgc\n INNER JOIN pg_namespace as pgn\n ON pgc.relnamespace = pgn.oid\n /* columns */\n /* This is a simplified version of how information_schema.columns was\n ** implemented in postgres 9.5, but modified to support materialized\n ** views.\n */\n LEFT OUTER JOIN pg_attribute AS pga\n ON pga.attrelid = pgc.oid\n LEFT OUTER JOIN (\n SELECT\n nc.nspname AS table_schema,\n c.relname AS table_name,\n a.attname AS column_name,\n a.attnum AS ordinal_position,\n pg_get_expr(ad.adbin, ad.adrelid) AS column_default,\n CASE WHEN a.attnotnull OR (t.typtype = 'd' AND t.typnotnull) THEN 'NO' ELSE 'YES' END AS is_nullable,\n CASE WHEN t.typtype = 'd' THEN\n CASE WHEN bt.typelem <> 0 AND bt.typlen = -1 THEN 'ARRAY'\n WHEN nbt.nspname = 'pg_catalog' THEN format_type(t.typbasetype, null)\n ELSE 'USER-DEFINED' END\n ELSE\n CASE WHEN t.typelem <> 0 AND t.typlen = -1 THEN 'ARRAY'\n WHEN nt.nspname = 'pg_catalog' THEN format_type(a.atttypid, null)\n ELSE 'USER-DEFINED' END\n END AS data_type,\n coalesce(bt.typname, t.typname) AS data_type_name\n FROM (pg_attribute a LEFT JOIN pg_attrdef ad ON attrelid = adrelid AND attnum = adnum)\n JOIN (pg_class c JOIN pg_namespace nc ON (c.relnamespace = nc.oid)) ON a.attrelid = c.oid\n JOIN (pg_type t JOIN pg_namespace nt ON (t.typnamespace = nt.oid)) ON a.atttypid = t.oid\n LEFT JOIN (pg_type bt JOIN pg_namespace nbt ON (bt.typnamespace = nbt.oid))\n ON (t.typtype = 'd' AND t.typbasetype = bt.oid)\n LEFT JOIN (pg_collation co JOIN pg_namespace nco ON (co.collnamespace = nco.oid))\n ON a.attcollation = co.oid AND (nco.nspname, co.collname) <> ('pg_catalog', 'default')\n WHERE -- (NOT pg_is_other_temp_schema(nc.oid))\n -- AND\n a.attnum > 0 AND NOT a.attisdropped AND c.relkind in ('r', 'v', 'm', 'f', 'p')\n AND (-- pg_has_role(c.relowner, 'USAGE')\n -- OR\n has_column_privilege(c.oid, a.attnum,\n 'SELECT, INSERT, UPDATE, REFERENCES'))\n ) AS isc\n ON isc.table_schema = pgn.nspname\n AND isc.table_name = pgc.relname\n AND isc.column_name = pga.attname\n /* triggers */\n LEFT OUTER JOIN pg_trigger AS pgt\n ON pgt.tgrelid = pgc.oid\n LEFT OUTER JOIN information_schema.triggers AS ist\n ON ist.event_object_schema = pgn.nspname\n AND ist.event_object_table = pgc.relname\n AND ist.trigger_name = pgt.tgname\n /* This is a simplified version of how information_schema.views was\n ** implemented in postgres 9.5, but modified to support materialized\n ** views.\n */\n LEFT OUTER JOIN (\n SELECT\n nc.nspname AS table_schema,\n c.relname AS table_name,\n pg_get_viewdef(c.oid),\n -- CASE WHEN pg_has_role(c.relowner, 'USAGE') THEN pg_get_viewdef(c.oid) ELSE null END AS view_definition,\n CASE WHEN pg_relation_is_updatable(c.oid, false) & 20 = 20 THEN 'YES' ELSE 'NO' END AS is_updatable,\n CASE WHEN pg_relation_is_updatable(c.oid, false) & 8 = 8 THEN 'YES' ELSE 'NO' END AS is_insertable_into,\n CASE WHEN EXISTS (SELECT 1 FROM pg_trigger WHERE tgrelid = c.oid AND tgtype & 81 = 81) THEN 'YES' ELSE 'NO' END AS is_trigger_updatable,\n CASE WHEN EXISTS (SELECT 1 FROM pg_trigger WHERE tgrelid = c.oid AND tgtype & 73 = 73) THEN 'YES' ELSE 'NO' END AS is_trigger_deletable,\n CASE WHEN EXISTS (SELECT 1 FROM pg_trigger WHERE tgrelid = c.oid AND tgtype & 69 = 69) THEN 'YES' ELSE 'NO' END AS is_trigger_insertable_into\n FROM pg_namespace nc, pg_class c\n WHERE c.relnamespace = nc.oid\n AND c.relkind in ('v', 'm')\n -- AND (NOT pg_is_other_temp_schema(nc.oid))\n AND (-- pg_has_role(c.relowner, 'USAGE')\n -- OR\n has_table_privilege(c.oid, 'SELECT, INSERT, UPDATE, DELETE, TRUNCATE, REFERENCES, TRIGGER')\n OR has_any_column_privilege(c.oid, 'SELECT, INSERT, UPDATE, REFERENCES'))\n ) AS isv\n ON isv.table_schema = pgn.nspname\n AND isv.table_name = pgc.relname\n WHERE\n pgc.relkind IN ('r', 'v', 'f', 'm', 'p')\n and ((pgn.nspname='private'))\n GROUP BY pgc.oid, pgn.nspname, pgc.relname, table_type, isv.*\n ) AS info;"
}
```

### Any possible solutions/workarounds you're aware of?

The only workaround I have been managing is to re-create the table, but that is not a good workaround.

### Keywords

has_column_privilege of relation does not exists

Start with the dashboard table-view and relationship metadata path, using the logged catalog query as the entry point. Reproduce the deleted-column case against CockroachDB, then verify that viewing the table and creating a relationship no longer returns the has_column_privilege error.

Written by the indexing model from the issue text.