On Tue, Apr 17, 2018 at 10:23 AM, Tomas Vondra <tomas.vondra@xxxxxxxxxxxxxxx> wrote:
On 04/17/2018 04:01 PM, Hackety Man wrote:
On Tue, Apr 17, 2018 at 6:49 AM, Tomas Vondra <tomas.vondra@xxxxxxxxxxxxxxx <mailto:tomas.vondra@2ndquadrant.com >> wrote:
On 04/16/2018 10:42 PM, Hackety Man wrote:
...
The first thing I did was to run some baseline tests using
the basic
queries inside of the IF() checks found in each of the
functions to
see how the query planner handled them. I ran the
following two
queries.
EXPLAIN (ANALYZE, BUFFERS) SELECT COUNT(*) FROM
zz_noidx1 WHERE
LOWER(text_distinct) = LOWER('Test5000001');
EXPLAIN (ANALYZE, BUFFERS) SELECT 1 FROM zz_noidx1 WHERE
LOWER(text_distinct) = LOWER('Test5000001');
Those are not the interesting plans, though. The EXISTS only cares
about the first row, so you should be looking at
EXPLAIN (ANALYZE, BUFFERS) SELECT 1 FROM zz_noidx1 WHERE
LOWER(text_distinct) = LOWER('Test5000001') LIMIT 1;
Okay. I tested this query and it did return an execution time on par with my tests of the "zz_spx_ifexists_noidx" function.
*
***//___^
I moved on to test the other function with the following
query...
EXPLAIN (ANALYZE, BUFFERS) SELECT * FROM
zz_spx_ifcount_noidx('Test5000001');
and I got the following "auto_explain" results...
2018-04-16 14:58:34.134 EDT [12616] LOG: duration:
426.279 ms plan:
Query Text: SELECT EXISTS (SELECT 1 FROM zz_noidx1 WHERE
LOWER(zz_noidx1.text_distinct) = LOWER(p_findme))
Result (cost=4.08..4.09 rows=1 width=1) (actual
time=426.274..426.274 rows=1 loops=1)
Buffers: shared read=5406
InitPlan 1 (returns $0)
-> Seq Scan on zz_noidx1 (cost=0.00..20406.00
rows=5000
width=0) (actual time=426.273..426.273 rows=0 loops=1)
Filter: (lower(text_distinct) = 'test5000001'::text)
Rows Removed by Filter: 1000000
Buffers: shared read=5406
2018-04-16 14:58:34.134 EDT [12616] CONTEXT: SQL statement
"SELECT EXISTS (SELECT 1 FROM zz_noidx1 WHERE
LOWER(zz_noidx1.text_distinct) = LOWER(p_findme))"
PL/pgSQL function zz_spx_ifexists_noidx(text) line 4
at IF
2018-04-16 14:58:34.134 EDT [12616] LOG: duration:
428.077 ms plan:
Query Text: explain (analyze, buffers) select * from
zz_spx_ifexists_noidx('Test5000001')
Function Scan on zz_spx_ifexists_noidx (cost=0.25..0.26
rows=1 width=32) (actual time=428.076..428.076 rows=1
loops=1)
Buffers: shared hit=30 read=5438
Definitely not the execution time, or query plan, results I was
expecting. As we can see, no workers were employed here and my
guess was that this was the reason or the large execution time
difference between these 2 tests? 199 milliseconds versus 428
milliseconds, which is a big difference. Why are workers
not being
employed here like they were when I tested the query found
inside of
the IF() check in a standalone manner? But then I ran
another test
and the results made even less sense to me.
The plan difference is due to not realizing the EXISTS essentially
implies LIMIT 1. Secondly, it expects about 5000 rows matching the
condition, uniformly spread through the table. But it apparently
takes much longer to find the first one, hence the increased duration.
Ah. I did not know that. So EXISTS inherently applies a LIMIT 1, even though it doesn't show in the query plan, correct? Is it not possible for parallel scans to be implemented while applying an implicit, or explicit, LIMIT 1?
It doesn't add a limit node to the plan, but it behaves similarly to that. The database only needs to fetch the first row to answer the EXISTS predicate.
I don't think this is inherently incompatible with parallel plans, but the planner simply thinks it's going to bee very cheap - cheaper than setting up parallel workers etc. So it does not do that.
Understood. Any chance of the planner possibly being enhanced in the future to come to a better conclusion as to whether, or not, a parallel scan implementation would be a better choice during EXISTS condition checks? :-)
How did you generate the data?
I used generate_series() to create 1 million records in sequence at the same time that I created the table using the following script...
CREATE TABLE zz_noidx1 AS SELECT generate_series(0, 999999) AS
int_distinct, 'Test'::text || generate_series(0, 999999)::text AS
text_distinct;
Which means that there are actually no matching rows for 'Test5000001'. So the database will scan the whole table anyway, in order to answer the EXISTS condition. The estimate of 5000 matching rows is a default value (0.5% out of 1M rows), because the value is entirely out of the data range covered by the histogram.
The easiest solution probably is adding an index on that column, which will make answering the EXISTS much faster (at least in this case).
Yes. I did test that scenario, as well. Adding an index does put the EXISTS condition check on par with the IF(SELECT COUNT(*) FROM...) condition check. The one scenario where the EXISTS condition check dominated over the IF(SELECT COUNT(*) FROM...) condition check was when no index was used and a matching string *was* found, as opposed to this test particular test where we're looking for a string that will *not* be found. I just wanted to test all possible scenarios.
That's a hard-coded value. The first 5 executions are re-planned using the actual parameter values, and then we try generating a generic plan and see if it's cheaper than the non-generic one. You can disable that, though.
When I ran the above query the first 5 times after starting my
Postgres service, I got the same results each time (around 428
milliseconds), but when running the query 6 or more times, the
execution time jumps up to almost double that. Here are the
"auto_explain" results running this query a 6th time...
This is likely due to generating a generic plan after the fifth
execution. There seems to be only small difference in costs, though.
--"auto_explain" results after running the same query 6
or more
times.
2018-04-16 15:01:51.635 EDT [12616] LOG: duration:
761.847 ms plan:
Query Text: SELECT EXISTS (SELECT 1 FROM zz_noidx1 WHERE
LOWER(zz_noidx1.text_distinct) = LOWER(p_findme))
Result (cost=4.58..4.59 rows=1 width=1) (actual
time=761.843..761.843 rows=1 loops=1)
Buffers: shared hit=160 read=5246
InitPlan 1 (returns $0)
-> Seq Scan on zz_noidx1 (cost=0.00..22906.00
rows=5000
width=0) (actual time=761.841..761.841 rows=0 loops=1)
Filter: (lower(text_distinct) = lower($1))
Rows Removed by Filter: 1000000
Buffers: shared hit=160 read=5246
2018-04-16 15:01:51.635 EDT [12616] CONTEXT: SQL statement
"SELECT EXISTS (SELECT 1 FROM zz_noidx1 WHERE
LOWER(zz_noidx1.text_distinct) = LOWER(p_findme))"
PL/pgSQL function zz_spx_ifexists_noidx(text) line 4
at IF
2018-04-16 15:01:51.635 EDT [12616] LOG: duration:
762.156 ms plan:
Query Text: explain (analyze, buffers) select * from
zz_spx_ifexists_noidx('Test5000001')
Function Scan on zz_spx_ifexists_noidx (cost=0.25..0.26
rows=1 width=32) (actual time=762.154..762.155 rows=1
loops=1)
Buffers: shared hit=160 read=5246
As you can see, the execution time jumps up to about 762
milliseonds. I can see in the sequence scan node that the
LOWER()
function shows up on the right side of the equal operator,
whereas
in the first 5 runs of this test query the plan did not
show this. Why is this?
It doesn't really matter on which side it shows, it's more about a
generic plan built without knowledge of the parameter value.
Right. I was more wondering why it switched over to a generic plan, as you've stated, like clockwork starting with the 6th execution run.
So on that note, in the planner's eyes, starting with the 6th execution, it looks like the planner still thinks that the generic plan will perform better than the non-generic one, which is why it keeps using the generic plan from that point forward?
Similar to the parallel scans, any chance of the planner possibly being enhanced in the future to come to a better conclusion as to whether, or not, the generic plan will perform better than the non-generic plan? :-)
regards
--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
Thanks for all the help! I really appreciate it!
Ryan
