workload/schemachange: screen for errors in operations

[jayshrivastava]

workload/schemachange: screen for errors in operations

These changes add error screening for the following (#56119):

• createEnum op
• createSchema op
• dropColumn op
• dropColumnDefault op
• dropColumnNotNull op
• dropSchema op
• renameColumn op
• renameTable op
• renameView op

Also, errors caused by schema changes after writes in the same txn are handled (#56230)

It also resolves a bug related to ambiguity in column names that was caused by views/tables referencing the same columns (#56235).

[cockroach-teamcity]

This change is 

[ajwerner]

just made it through 2 commits

Reviewed 2 of 2 files at r1, 2 of 2 files at r2.
Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @ajwerner and @jayshrivastava)

---

pkg/workload/schemachange/error_screening.go, line 90 at r1 (raw file):

Quoted 13 lines of code…

	return scanBool(tx, `SELECT EXISTS (
		SELECT fd.descriptor_name
		  FROM crdb_internal.forward_dependencies AS fd
		 WHERE fd.descriptor_id =
				(
				SELECT c.oid
				  FROM pg_catalog.pg_class AS c
					JOIN pg_catalog.pg_namespace AS ns
            ON ns.oid = c.relnamespace
         WHERE c.relname = $1
           AND ns.nspname = $2
				)
	)`, tableName.Object(), tableName.Schema())

nit: sqlfum.pt

SELECT EXISTS(
        SELECT fd.descriptor_name
          FROM crdb_internal.forward_dependencies AS fd
         WHERE fd.descriptor_id
               = (
                    SELECT c.oid
                      FROM pg_catalog.pg_class AS c
                      JOIN pg_catalog.pg_namespace AS ns ON
                            ns.oid = c.relnamespace
                     WHERE c.relname = $1 AND ns.nspname = $2
                )
       );

---

pkg/workload/schemachange/error_screening.go, line 95 at r2 (raw file):

}

func columnIsDependedOn(tx *pgx.Tx, tableName *tree.TableName, columnName string) (bool, error) {

commentary on how this works would be helpful. Namely explain the structure of the data you are picking apart.

Also, it feels like you could do this is sql with only modestly more clunkiness.

Also, this doesn't seem to work on foreign key references :(

---

pkg/workload/schemachange/error_screening.go, line 154 at r2 (raw file):

func colIsPrimaryKey(tx *pgx.Tx, tableName *tree.TableName, columnName string) (bool, error) {
	return scanBool(tx, `SELECT EXISTS(

nit: sqlfumpt

[ajwerner]

Okay, made it through all of it. Mostly just would like to see a bit more commentary where stuff is getting big. Otherwise

Reviewed 1 of 1 files at r11.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @jayshrivastava)

---

pkg/workload/schemachange/error_screening.go, line 95 at r2 (raw file):

Previously, ajwerner wrote…

commentary on how this works would be helpful. Namely explain the structure of the data you are picking apart.

Also, it feels like you could do this is sql with only modestly more clunkiness.

Also, this doesn't seem to work on foreign key references :(

yeah I'm worried you may also need a join on pg_constraint, depending on what you're intending to use this for eventually. Comment that this is about dependencies which prevent drops and not foreign keys.

---

pkg/workload/schemachange/operation_generator.go, line 567 at r11 (raw file):

	}

	uniqueColumnNames := map[string]bool{}

commentary here is starting to be in order.

[ajwerner]

Reviewed 1 of 2 files at r13, 1 of 1 files at r22.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @jayshrivastava)

[jayshrivastava]

bors r=ajwerner

[craig]

Build failed:

• GitHub CI (Cockroach)

[jayshrivastava]

bors retry

[craig]

Build succeeded:

• GitHub CI (Cockroach)