How to find a missing value from another table - sql

Question

Imagine if I had two tables like this:
Table Name: user
| id | user_id | password |
Table Name: permissions
| id | admin | write | delete | transfer |
And populated the tables with this:
Inserting into user table:
- 0, joshsh, asdf01
- 1, jakesh, asdf02
- 2, annsh, asdf03
- 3, lamsh, asdf04
Inserting into permissions table:
- 0, yes, yes, yes, yes
- 1, yes, yes, yes, yes
- 2, no, yes, yes, yes
And I didn't add in the 4th value into the permissions table. How would I query a statement as to find which numbers I forgot (in case it was a big database)?

Answer 1

There are many ways to get the same result, these are two possible ways:
With a NOT EXISTS:
SQL> select u.*
2 from user_ u
3 where not exists (
4 select 1
5 from permissions p
6 where u.id = p.id
7 );
ID USER_ID PASSWORD
---------- ---------- ----------
3 lamsh asdf04
Execution Plan
----------------------------------------------------------
Plan hash value: 3342498783
----------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 4 | 160 | 7 (15)| 00:00:01 |
|* 1 | HASH JOIN ANTI | | 4 | 160 | 7 (15)| 00:00:01 |
| 2 | TABLE ACCESS FULL| USER_ | 4 | 108 | 3 (0)| 00:00:01 |
| 3 | TABLE ACCESS FULL| PERMISSIONS | 3 | 39 | 3 (0)| 00:00:01 |
----------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("U"."ID"="P"."ID")
Note
-----
- dynamic sampling used for this statement (level=2)
And with an outer join:
SQL> select u.*
2 from user_ u
3 left outer join permissions p
4 on (u.id = p.id)
5 where p.id is null;
ID USER_ID PASSWORD
---------- ---------- ----------
3 lamsh asdf04
Execution Plan
----------------------------------------------------------
Plan hash value: 3342498783
----------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
----------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 4 | 160 | 7 (15)| 00:00:01 |
|* 1 | HASH JOIN ANTI | | 4 | 160 | 7 (15)| 00:00:01 |
| 2 | TABLE ACCESS FULL| USER_ | 4 | 108 | 3 (0)| 00:00:01 |
| 3 | TABLE ACCESS FULL| PERMISSIONS | 3 | 39 | 3 (0)| 00:00:01 |
----------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
1 - access("U"."ID"="P"."ID")
Note
-----
- dynamic sampling used for this statement (level=2)
Notice that Oracle did the same explain plan for both statements, considering that the test has been made with two very small tables with no indexes and no statistics.
Also, consider that there are many ways to get the same result; performance strongly depends on you data, stats, indexes, ...
ps I used USER_ instead of USER, to avoid problems.

Answer 2

Assuming that the id column is the join between the permissions and user tables based on the id column (and that id is the primary/unique key in both tables), here are a couple of solutions:
select id from permissions
minus
select id from user;
or
select * from user
where id not in (select id from permissions);

Answer 3

By applying count(*) on both the tables if they are not equal then alert will be generated and that altert can be generated using sql server agent job.
or you can do this
Select * from user as u
full outer join permissions as p
on a.id in(select id from user)

Answer 4

The best way is to use LEFT JOIN with and IS NULL in where clause, it's elegant and efficient in every aspects:
SELECT u.*
FROM user u
LEFT JOIN permissions p ON u.id = p.id
WHERE p.id IS NULL