#	exclusive	inclusive	rows x	rows	loops	node
1.	0.016	7,348.110	↑ 1.0	100	1	→ Limit (cost=273,701.59..273,713.26 rows=100 width=809) (actual time=7,342.684..7,348.110 rows=100 loops=1) Buffers: shared hit=218 read=19,191
2.	60.650	7,348.094	↑ 91,746.4	100	1	→ Gather Merge (cost=273,701.59..1,344,150.99 rows=9,174,640 width=809) (actual time=7,342.682..7,348.094 rows=100 loops=1) Workers Planned: 2 Workers Launched: 2 Buffers: shared hit=218 read=19,191
3.	1,297.238	7,287.444	↑ 45,873.2	100	3 / 3	→ Sort (cost=272,701.57..284,169.87 rows=4,587,320 width=809) (actual time=7,287.438..7,287.444 rows=100 loops=3) Sort Key: pgbench_accounts.abalance Sort Method: top-N heapsort Memory: 51kB Buffers: shared hit=218 read=19,191
4.	963.737	5,990.206	↑ 1.4	3,333,330	3 / 3	→ Hash Left Join (cost=308.07..97,377.50 rows=4,587,320 width=809) (actual time=267.920..5,990.206 rows=3,333,330 loops=3) Hash Cond: (pgbench_accounts.bid = pgbench_tellers.bid) Buffers: shared hit=204 read=19,191
5.	172.197	4,896.152	↑ 1.4	333,333	3 / 3	→ Hash Join (cost=28.07..44,325.43 rows=459,926 width=457) (actual time=137.494..4,896.152 rows=333,333 loops=3) Hash Cond: (pgbench_accounts.bid = pgbench_branches.bid) Buffers: shared hit=94 read=19,136
6.	4,703.754	4,703.754	↑ 1.4	333,333	3 / 3	→ Parallel Index Scan using pgbench_accounts_pkey on public.pgbench_accounts (cost=0.57..43,085.50 rows=459,926 width=97) (actual time=117.167..4,703.754 rows=333,333 loops=3) Index Cond: (pgbench_accounts.aid < 1000000) Buffers: shared read=19,131
7.	0.206	20.201	↑ 1.0	1,000	3 / 3	→ Hash (cost=15.00..15.00 rows=1,000 width=364) (actual time=20.200..20.201 rows=1,000 loops=3) Buckets: 1,024 Batches: 1 Memory Usage: 48kB Buffers: shared hit=10 read=5
8.	19.995	19.995	↑ 1.0	1,000	3 / 3	→ Seq Scan on public.pgbench_branches (cost=0.00..15.00 rows=1,000 width=364) (actual time=13.090..19.995 rows=1,000 loops=3) Buffers: shared hit=10 read=5
9.	1.937	130.317	↑ 1.0	10,000	3 / 3	→ Hash (cost=155.00..155.00 rows=10,000 width=352) (actual time=130.317..130.317 rows=10,000 loops=3) Buckets: 16,384 Batches: 1 Memory Usage: 558kB Buffers: shared hit=110 read=55
10.	128.380	128.380	↑ 1.0	10,000	3 / 3	→ Seq Scan on public.pgbench_tellers (cost=0.00..155.00 rows=10,000 width=352) (actual time=16.691..128.380 rows=10,000 loops=3) Buffers: shared hit=110 read=55

Limit  (cost=273701.59..273713.26 rows=100 width=809) (actual time=7342.684..7348.110 rows=100 loops=1)
   Buffers: shared hit=218 read=19191
   ->  Gather Merge  (cost=273701.59..1344150.99 rows=9174640 width=809) (actual time=7342.682..7348.094 rows=100 loops=1)
         Workers Planned: 2
         Workers Launched: 2
         Buffers: shared hit=218 read=19191
         ->  Sort  (cost=272701.57..284169.87 rows=4587320 width=809) (actual time=7287.438..7287.444 rows=100 loops=3)
               Sort Key: pgbench_accounts.abalance
               Sort Method: top-N heapsort  Memory: 51kB
               Buffers: shared hit=218 read=19191
               ->  Hash Left Join  (cost=308.07..97377.50 rows=4587320 width=809) (actual time=267.920..5990.206 rows=3333330 loops=3)
                     Hash Cond: (pgbench_accounts.bid = pgbench_tellers.bid)
                     Buffers: shared hit=204 read=19191
                     ->  Hash Join  (cost=28.07..44325.43 rows=459926 width=457) (actual time=137.494..4896.152 rows=333333 loops=3)
                           Hash Cond: (pgbench_accounts.bid = pgbench_branches.bid)
                           Buffers: shared hit=94 read=19136
                           ->  Parallel Index Scan using pgbench_accounts_pkey on public.pgbench_accounts  (cost=0.57..43085.50 rows=459926 width=97) (actual time=117.167..4703.754 rows=333333 loops=3)
                                 Index Cond: (pgbench_accounts.aid < 1000000)
                                 Buffers: shared read=19131
                           ->  Hash  (cost=15.00..15.00 rows=1000 width=364) (actual time=20.200..20.201 rows=1000 loops=3)
                                 Buckets: 1024  Batches: 1  Memory Usage: 48kB
                                 Buffers: shared hit=10 read=5
                                 ->  Seq Scan on public.pgbench_branches  (cost=0.00..15.00 rows=1000 width=364) (actual time=13.090..19.995 rows=1000 loops=3)
                                       Buffers: shared hit=10 read=5
                     ->  Hash  (cost=155.00..155.00 rows=10000 width=352) (actual time=130.317..130.317 rows=10000 loops=3)
                           Buckets: 16384  Batches: 1  Memory Usage: 558kB
                           Buffers: shared hit=110 read=55
                           ->  Seq Scan on public.pgbench_tellers  (cost=0.00..155.00 rows=10000 width=352) (actual time=16.691..128.380 rows=10000 loops=3)
                                 Buffers: shared hit=110 read=55

Per node type stats

node type	count	sum of times	% of query
Gather Merge	1	60.650 ms	0.8 %
Hash	2	2.143 ms	0.0 %
Hash Join	1	172.197 ms	2.3 %
Hash Left Join	1	963.737 ms	13.1 %
Limit	1	0.016 ms	0.0 %
Parallel Index Scan	1	4,703.754 ms	64.0 %
Seq Scan	2	148.375 ms	2.0 %
Sort	1	1,297.238 ms	17.7 %

Per table stats

Table name	Scan count	Total time	% of query
scan type	count	sum of times	% of table
public.pgbench_accounts	1	4,703.754 ms	64.0 %
Parallel Index Scan	1	4,703.754 ms	100.0 %
public.pgbench_branches	1	19.995 ms	0.3 %
Seq Scan	1	19.995 ms	100.0 %
public.pgbench_tellers	1	128.380 ms	1.7 %
Seq Scan	1	128.380 ms	100.0 %

explain.depesz.com

PostgreSQL's explain analyze made readable

Result: xFD7

Color mode:

Visible columns:

Per node type stats

Per table stats