EXPLAIN Request Modifier and Recursion | Interpreting EXPLAIN Output | Vantage - 17.10 - EXPLAIN Request Modifier and Recursion - Advanced SQL Engine - Teradata Database

Teradata Vantageā„¢ - SQL Request and Transaction Processing

Product
Advanced SQL Engine
Teradata Database
Release Number
17.10
Release Date
July 2021
Content Type
Programming Reference
User Guide
Publication ID
B035-1142-171K
Language
English (United States)

Recursive SELECT Example

The following example demonstrates a recursive query:

EXPLAIN 
WITH RECURSIVE temp_table (employee_number, depth) AS
(SELECT root.employee_number, 0 as depth
 FROM Employee root
 WHERE root.manager_employee_number = 801
 UNION ALL
  SELECT indirect.employee_number, seed.depth+1 as depth
  FROM temp_table seed, Employee indirect
  WHERE seed.employee_number = indirect.manager_employee_number
  AND depth <= 20)
SELECT employee_number, depth FROM temp_table;

EXPLAIN generates the following report for this request:

...
2) Next, we lock PERSONNEL.root for read.
3) We do an all-AMPs RETRIEVE step from PERSONNEL.root by
   way of an all-rows scan with a condition of (
   "PERSONNEL.root.manager_employee_number = 801") into
   Spool 3 (all_amps), which is built locally on the AMPs. The size
   of Spool 3 is estimated with no confidence to be 1 row. The
   estimated time for this step is 0.06 seconds.
4) We do an all-AMPs RETRIEVE step from Spool 3 by way of an all-rows
   scan into Spool 2 (all_amps), which is built locally on the AMPs.
   The size of Spool 2 is estimated with no confidence to be 1 row.
   The estimated time for this step is 0.07 seconds.
5) We execute the following steps in parallel.
     1) We do an all-AMPs RETRIEVE step from Spool 3 (Last Use) by
        way of an all-rows scan with a condition of ("(DEPTH <= 20)
        AND (NOT (EMPLOYEE_NUMBER IS NULL ))") into Spool 4
        (all_amps), which is duplicated on all AMPs. The size of
        Spool 4 is estimated with no confidence to be 8 rows. The
        estimated time for this step is 0.03 seconds.
     2) We do an all-AMPs RETRIEVE step from
        PERSONNEL.indirect by way of an all-rows scan with a
        condition of
        ("NOT (PERSONNEL.indirect.manager_employee_number IS NULL)")
        into Spool 5 (all_amps), which is built locally on the AMPs.
        The size of Spool 5 is estimated with no confidence to be 8
        rows. The estimated time for this step is 0.01 seconds.
6) We do an all-AMPs JOIN step from Spool 4 (Last Use) by way of an
   all-rows scan, which is joined to Spool 5 (Last Use) by way of an
   all-rows scan. Spool 4 and Spool 5 are joined using a single
   partition hash join, with a join condition of ("EMPLOYEE_NUMBER =
   manager_employee_number"). The result goes into Spool 6
   (all_amps), which is built locally on the AMPs. The size of Spool
   6 is estimated with no confidence to be 3 rows. The estimated
   time for this step is 0.08 seconds.
7) We do an all-AMPs RETRIEVE step from Spool 6 (Last Use) by way of
   an all-rows scan into Spool 3 (all_amps), which is built locally
   on the AMPs. The size of Spool 3 is estimated with no confidence
   to be 4 rows. The estimated time for this step is 0.07 seconds.
   If one or more rows are inserted into spool 3, then go to step 4.
8) We do an all-AMPs RETRIEVE step from Spool 2 (Last Use) by way of
   an all-rows scan into Spool 7 (all_amps), which is built locally
   on the AMPs. The size of Spool 7 is estimated with no confidence
   to be 142 rows. The estimated time for this step is 0.07 seconds.
...

Step 3 indicates the processing of the seed statement inside the recursive query and produces the initial temporary result set.

Steps 4 through 7 correspond to the processing of the recursive statement inside the recursive query and repeat until no new rows are inserted into the temporary result set. Although steps 4 through 7 indicate a static plan, each iteration can produce spools with varying cardinalities; thus, the level of confidence for the spool size in these steps is set to no confidence.

Step 8 indicates the processing of the final result that is sent back to the user.