Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Django - postgres: How to create an index on a JsonB field

I want to allow indexing on JsonB field on an ID which is a few levels deep into the json data in our Django project. Here's what the JSONB data looks like:

  "foreign_data":{
      "some_key": val
      "src_data": {
              "VEHICLE": {
                  "title": "615",
                  "is_working": true,
                  "upc": "85121212121",
                  "dealer_name": "CryptoDealer",
                  "id": 1222551
              }
        }
 }

I want to index on the field id using Django views but not sure how to achieve that. Happy to post my Django ViewSet if it helps.

like image 238
noobcoder Avatar asked Nov 27 '25 18:11

noobcoder


1 Answers

t=# create table d(i bigserial, j jsonb);
CREATE TABLE
t=# insert into d(j) select ('{"foreign_data":{
      "some_key": '||g||',
      "src_data": {
              "VEHICLE": {
                  "title": "615",
                  "is_working": true,
                  "upc": "85121212121",
                  "dealer_name": "CryptoDealer",
                  "id": '||g||'
              }
        }
 }}')::jsonb from generate_series(1,1222600) g;
INSERT 0 1222600
t=# create index ji on d (cast (j->'foreign_data'->'src_data'->'VEHICLE'->>'id' as int));
CREATE INDEX

in order to use such fn() based index youhave to "repeat" function in query:

t=# explain analyze select * from d 
where cast (j->'foreign_data'->'src_data'->'VEHICLE'->>'id' as int) = 1222551;
                                                          QUERY PLAN
------------------------------------------------------------------------------------------------------------------------------
 Index Scan using ji on d  (cost=0.43..8.45 rows=1 width=215) (actual time=0.021..0.021 rows=1 loops=1)
   Index Cond: ((((((j -> 'foreign_data'::text) -> 'src_data'::text) -> 'VEHICLE'::text) ->> 'id'::text))::integer = 1222551)
 Planning time: 1.585 ms
 Execution time: 0.045 ms
(4 rows)

as you see cost is tiny and execution is cheap over index. but if you "skip" formalities and run:

t=# explain analyze select * from d 
where j->'foreign_data'->'src_data'->'VEHICLE'->>'id' = '1222551';
                                                         QUERY PLAN
-----------------------------------------------------------------------------------------------------------------------------
 Gather  (cost=1000.00..50122.31 rows=6113 width=215) (actual time=335.996..336.000 rows=1 loops=1)
   Workers Planned: 2
   Workers Launched: 2
   ->  Parallel Seq Scan on d  (cost=0.00..48511.01 rows=2547 width=215) (actual time=223.548..332.213 rows=0 loops=3)
         Filter: (((((j -> 'foreign_data'::text) -> 'src_data'::text) -> 'VEHICLE'::text) ->> 'id'::text) = '1222551'::text)
         Rows Removed by Filter: 407533
 Planning time: 0.096 ms
 Execution time: 343.090 ms
(8 rows)

index will not be used

like image 108
Vao Tsun Avatar answered Nov 30 '25 09:11

Vao Tsun



Donate For Us

If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!