[24.0] Fix very slow user data table query

[mvdbeek]

by adding an index on extension.
Fixes extremely slow tool form building.

Before:

['Sort  (cost=11152.93..11152.94 rows=1 width=575) (actual time=151.608..152.636 rows=0 loops=1)',
 '  Sort Key: history_dataset_association.id',
 '  Sort Method: quicksort  Memory: 25kB',
 '  ->  Nested Loop  (cost=1001.26..11152.92 rows=1 width=575) (actual time=151.564..152.592 rows=0 loops=1)',
 '        ->  Nested Loop  (cost=1000.84..11144.46 rows=1 width=575) (actual time=151.564..152.592 rows=0 loops=1)',
 '              Join Filter: (history_dataset_association.dataset_id = dataset.id)',
 '              ->  Nested Loop  (cost=1000.42..11143.94 rows=1 width=575) (actual time=151.563..152.591 rows=0 loops=1)',
 '                    ->  Gather  (cost=1000.00..11135.50 rows=1 width=363) (actual time=151.563..152.590 rows=0 loops=1)',
 '                          Workers Planned: 2',
 '                          Workers Launched: 2',
 '                          ->  Parallel Seq Scan on history_dataset_association  (cost=0.00..10135.40 rows=1 width=363) (actual time=139.910..139.911 rows=0 loops=3)',
 "                                Filter: ((NOT deleted) AND (metadata ~~ '\\x2522616c6c5f66617374612225'::bytea) AND ((extension)::text = 'data_manager_json'::text))",
 '                                Rows Removed by Filter: 80928',
 '                    ->  Index Scan using dataset_pkey on dataset dataset_1  (cost=0.42..8.44 rows=1 width=212) (never executed)',
 '                          Index Cond: (id = history_dataset_association.dataset_id)',
 '              ->  Index Scan using dataset_pkey on dataset  (cost=0.42..0.51 rows=1 width=4) (never executed)',
 '                    Index Cond: (id = dataset_1.id)',
 "                    Filter: ((total_size <> file_size) AND ((state)::text = 'ok'::text))",
 '        ->  Index Scan using history_pkey on history  (cost=0.42..8.44 rows=1 width=4) (never executed)',
 '              Index Cond: (id = history_dataset_association.history_id)',
 '              Filter: (user_id = 1)',
 'Planning Time: 9.134 ms',
 'Execution Time: 152.984 ms']

After:

['Sort  (cost=352.72..352.73 rows=1 width=575) (actual time=3.429..3.432 rows=0 loops=1)',
 '  Sort Key: history_dataset_association.id',
 '  Sort Method: quicksort  Memory: 25kB',
 '  ->  Nested Loop  (cost=6.35..352.71 rows=1 width=575) (actual time=3.377..3.379 rows=0 loops=1)',
 '        ->  Nested Loop  (cost=5.93..344.25 rows=1 width=575) (actual time=3.377..3.378 rows=0 loops=1)',
 '              Join Filter: (history_dataset_association.dataset_id = dataset.id)',
 '              ->  Nested Loop  (cost=5.51..343.73 rows=1 width=575) (actual time=3.376..3.377 rows=0 loops=1)',
 '                    ->  Bitmap Heap Scan on history_dataset_association  (cost=5.09..335.29 rows=1 width=363) (actual time=3.376..3.376 rows=0 loops=1)',
 "                          Recheck Cond: ((extension)::text = 'data_manager_json'::text)",
 "                          Filter: ((NOT deleted) AND (metadata ~~ '\\x2522616c6c5f66617374612225'::bytea))",
 '                          Rows Removed by Filter: 126',
 '                          Heap Blocks: exact=55',
 '                          ->  Bitmap Index Scan on hda_ext  (cost=0.00..5.09 rows=89 width=0) (actual time=0.616..0.617 rows=126 loops=1)',
 "                                Index Cond: ((extension)::text = 'data_manager_json'::text)",
 '                    ->  Index Scan using dataset_pkey on dataset dataset_1  (cost=0.42..8.44 rows=1 width=212) (never executed)',
 '                          Index Cond: (id = history_dataset_association.dataset_id)',
 '              ->  Index Scan using dataset_pkey on dataset  (cost=0.42..0.51 rows=1 width=4) (never executed)',
 '                    Index Cond: (id = dataset_1.id)',
 "                    Filter: ((total_size <> file_size) AND ((state)::text = 'ok'::text))",
 '        ->  Index Scan using history_pkey on history  (cost=0.42..8.44 rows=1 width=4) (never executed)',
 '              Index Cond: (id = history_dataset_association.history_id)',
 '              Filter: (user_id = 1)',
 'Planning Time: 23.736 ms',
 'Execution Time: 3.594 ms']

This became important with #17435

How to test the changes?

(Select all options that apply)

• I've included appropriate automated tests.
• This is a refactoring of components with existing test coverage.
• Instructions for manual testing are as follows:
  1. [add testing steps and prerequisites here if you didn't write automated tests covering all your changes]

License

• I agree to license these and all my past contributions to the core galaxy codebase under the MIT license.

[wm75]

not directly of concern here, but:
this tool https://usegalaxy.eu/root?tool_id=toolshed.g2.bx.psu.edu/repos/wolma/mimodd_main/mimodd_map/0.1.9 was particularly strongly affected by the slow loading and what's special about it is that it has a datatable select in multiple branches of a conditional. From this it seems the query is executed multiple times for the same data table in a wrapper?
Avoiding this might be another possible optimization.