I just have encountered a strange use case with my Postgres. I run the following commands:

CREATE INDEX idx_index_value_and_text ON index_test ( value, text );

explain (analyze true, verbose true, buffers true) select * from index_test r where r.text='some value';

...where 'value' and 'text' are int and varchar columns from the previous setup. And I have the following reply from Postgres:

'Index Scan using idx_index_value_and_text on index_test r (cost=0.00..262.76 rows=1 width=610) (actual time=0.543..0.543 rows=0 loops=1)'

' Output: id, text, last_modified, value, n_text, n_last_modified, n_value, item_type'

' Index Cond: ((r.text)::text = 'some value'::text)'

' Buffers: shared hit=73'

'Total runtime: 0.568 ms'

What happens here? It looks like Postgres is doing Index Scan, though it cannot do it because in WHERE clause I use the SECOND column of the index. Could you explain this, please?