Обсуждение: Postgres not using index on views

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely for the addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended recipient is unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender. Please note that any views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts no liability for any damage caused by any malicious code or virus transmitted by this e-mail.

On Mon, Apr 06, 2020 at 02:19:59PM +0000, Rick Vincent wrote:
> I am seeing a performance problem with postgresql v 11.7 on views, and I am wondering if anyone can tell me why or
hasany suggestion.
 
> 
> A table is created as:
> 
> CREATE TABLE "FBNK_CUSTOMER" (RECID VARCHAR(255) NOT NULL PRIMARY KEY, XMLRECORD VARCHAR)
> 
> And contains only 180 rows.
> 
> Doing an explain plan on the view created over this gives:
> 
> EXPLAIN ANALYZE
> select RECID from "V_FBNK_CUSTOMER"
> 
> 
> Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19014.60 rows=180 width=7) (actual time=459.601..78642.189 rows=180
loops=1)
>   ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180 width=14575) (actual time=459.600..78641.950
rows=180loops=1)
 
> 
> Yet an Explain plan on the underlying table( on select RECID from "FBNK_CUSTOMER") gives:
> 
> Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..22.80 rows=180 width=7) (actual time=0.004..0.272 rows=180 loops=1)

It still did a seq scan on the table, so I'm not sure what this has to do with
index scans ?

> The following query takes an extremely long time for only 180 rows, and what this means is that we would have to
indexanything appearing in the where clause for every table in order to use views because the views seem not to
considerthe select clause.  Why is that and does anyone know a way around this?
 

Is there a reason why you don't store the extracted value in its own column ?
And maybe keep it up to date using an insert/update trigger on the xmlrecord
column.

-- 
Justin

Justin Pryzby <pryzby@telsasoft.com> writes:
> On Mon, Apr 06, 2020 at 02:19:59PM +0000, Rick Vincent wrote:
>> The following query takes an extremely long time for only 180 rows, and what this means is that we would have to
indexanything appearing in the where clause for every table in order to use views because the views seem not to
considerthe select clause.  Why is that and does anyone know a way around this? 

> Is there a reason why you don't store the extracted value in its own column ?

The planner seems to be quite well aware that the slower query is going to
be slower, since the estimated costs are much higher.  Since it's not
choosing to optimize into a faster form, I wonder whether it's constrained
by semantic requirements.  In particular, I'm suspicious that some of
those functions you have in the view are marked "volatile", preventing
them from being optimized away.

Beyond that guess, though, there's really not enough info here to say.
The info we usually ask for to debug slow-query problems is explained
at

https://wiki.postgresql.org/wiki/Slow_Query_Questions

            regards, tom lane

On Mon, 2020-04-06 at 14:19 +0000, Rick Vincent wrote:
> I am seeing a performance problem with postgresql v 11.7 on views, and I am wondering if
> anyone can tell me why or has any suggestion.

Your account is somewhat confused - too many questions rolled into one
rant, I would say.

There are two points that may clear up the case:

- If you have no WHERE clause, a sequential scan of the table is usually
  the best way to do it.  The exception is an index only scan if the index
  contains all that is required, but in PostgreSQL you need a recently
  VACUUMed table for that.

- The expensive part in your view is the "extractValueJS" function.
  Try to tune that for better performance.

If any of your problems are not explained by that, please say so.

Yours,
Laurenz Albe
-- 
Cybertec | https://www.cybertec-postgresql.com

Hi Justin,

You said, " Is there a reason why you don't store the extracted value in its own column ?"

RV>> It simply is the way the application stores the data.  For Oracle we are storing in XML and JSON format, for
postgres,due do limitations of XML api, we are storing in VARCHAR.  We can't break it out into columns very easily
becauseof the legacy application. 

You said, "It still did a seq scan on the table, so I'm not sure what this has to do with index scans ?"

RV>> On Oracle it will use the primary key index because it detects that all of the columns in the select clause are
indexable. With Postgres, it might be doing a seq scan but on a 180 rows, a select on the underlying table is many
timesfaster than the same select on the view.  It seems all of the view columns are being triggered which makes it
incrediblyslow. 

Thanks,
Rick



-----Original Message-----
From: Justin Pryzby <pryzby@telsasoft.com>
Sent: Tuesday, April 7, 2020 6:59 AM
To: Rick Vincent <rvincent@temenos.com>
Cc: pgsql-performance@postgresql.org; Manoj Kumar <manojkumar@temenos.com>; Herve Aubert <haubert@temenos.com>
Subject: Re: Postgres not using index on views

On Mon, Apr 06, 2020 at 02:19:59PM +0000, Rick Vincent wrote:
> I am seeing a performance problem with postgresql v 11.7 on views, and I am wondering if anyone can tell me why or
hasany suggestion. 
>
> A table is created as:
>
> CREATE TABLE "FBNK_CUSTOMER" (RECID VARCHAR(255) NOT NULL PRIMARY KEY,
> XMLRECORD VARCHAR)
>
> And contains only 180 rows.
>
> Doing an explain plan on the view created over this gives:
>
> EXPLAIN ANALYZE
> select RECID from "V_FBNK_CUSTOMER"
>
>
> Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19014.60 rows=180 width=7) (actual time=459.601..78642.189 rows=180
loops=1)
>   ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180
> width=14575) (actual time=459.600..78641.950 rows=180 loops=1)
>
> Yet an Explain plan on the underlying table( on select RECID from "FBNK_CUSTOMER") gives:
>
> Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..22.80 rows=180 width=7)
> (actual time=0.004..0.272 rows=180 loops=1)

It still did a seq scan on the table, so I'm not sure what this has to do with index scans ?

> The following query takes an extremely long time for only 180 rows, and what this means is that we would have to
indexanything appearing in the where clause for every table in order to use views because the views seem not to
considerthe select clause.  Why is that and does anyone know a way around this? 

Is there a reason why you don't store the extracted value in its own column ?
And maybe keep it up to date using an insert/update trigger on the xmlrecord column.

--
Justin

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely
forthe addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended
recipientis unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender.
Pleasenote that any views or opinions presented in this e-mail are solely those of the author and do not necessarily
representthose of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts
noliability for any damage caused by any malicious code or virus transmitted by this e-mail. 

Hi Tom,

 

The function is defined as below, so no use of VOLATILE.  Let me know if you need any other information.  I am hoping the below will further clarify the issue.

 

CREATE OR REPLACE FUNCTION extractValueJS (sVar text, nfm INTEGER, nvm INTEGER)

RETURNS VARCHAR as $$

declare

sRet text := '';

nSize int := 0;

retVal int := 0;

cVar text[] := regexp_split_to_array(sVar,'');

idx int := 1;

nStart int := 0;

nEnd int := 0;

begin

etc...

        return sRet;

end;

$$ LANGUAGE plpgsql;

 

After reading you link…..

 

Here is a better explain plan:

 

Explain on the table:

 

EXPLAIN (analyze,BUFFERS)

select RECID from "FBNK_CUSTOMER"

Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..22.80 rows=180 width=7) (actual time=0.011..0.073 rows=180 loops=1)

  Buffers: shared hit=21

Planning Time: 0.056 ms

Execution Time: 0.091 ms

 

Explain on the view:

 

EXPLAIN (analyze,BUFFERS)

select RECID from "V_FBNK_CUSTOMER"

 

Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19014.60 rows=180 width=7) (actual time=455.727..76837.097 rows=180 loops=1)

  Buffers: shared hit=204

  ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180 width=14575) (actual time=455.726..76836.791 rows=180 loops=1)

        Buffers: shared hit=204

Planning Time: 1.109 ms

Execution Time: 76838.505 ms

 

Explain on view with a column:

 

EXPLAIN (analyze,BUFFERS)

SELECT RECID FROM "V_FBNK_CUSTOMER" WHERE "TESTER" = '5.00' ORDER BY RECID

Sort  (cost=19015.06..19015.06 rows=1 width=7) (actual time=76033.475..76033.475 rows=1 loops=1)

  Sort Key: "V_FBNK_CUSTOMER".recid

  Sort Method: quicksort  Memory: 25kB

  Buffers: shared hit=21

  ->  Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19015.05 rows=1 width=7) (actual time=66521.952..76033.434 rows=1 loops=1)

        Filter: (("V_FBNK_CUSTOMER"."TESTER")::text = '5.00'::text)

        Rows Removed by Filter: 179

        Buffers: shared hit=21

        ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180 width=14575) (actual time=462.949..76033.096 rows=180 loops=1)

              Buffers: shared hit=21

Planning Time: 0.819 ms

Execution Time: 76033.731 ms

 

But on the underlying table and not the view but just using the one view column called TESTER:

 

EXPLAIN (analyze,BUFFERS)

SELECT RECID FROM "FBNK_CUSTOMER" WHERE extractValueJS(XMLRECORD, 179, 9) = '5.00' ORDER BY RECID

Sort  (cost=68.26..68.27 rows=1 width=7) (actual time=220.403..220.404 rows=1 loops=1)

  Sort Key: recid

  Sort Method: quicksort  Memory: 25kB

  Buffers: shared hit=21

  ->  Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..68.25 rows=1 width=7) (actual time=193.000..220.397 rows=1 loops=1)

        Filter: ((extractvaluejs((xmlrecord)::text, 179, 9))::text = '5.00'::text)

        Rows Removed by Filter: 179

        Buffers: shared hit=21

Planning Time: 0.045 ms

Execution Time: 220.418 ms

 

Other info:

 

SELECT relname, relpages, reltuples, relallvisible, relkind, relnatts, relhassubclass, reloptions, pg_table_size(oid) FROM pg_class WHERE relname='FBNK_CURRENCY';

 

relname relpages        reltuples       relallvisible   relkind relnatts        relhassubclass  reloptions      pg_table_size

FBNK_CURRENCY   6       93      0       r       2       false   NULL    81920

 

Version is:

PostgreSQL 11.7 (Debian 11.7-2.pgdg90+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 6.3.0-18+deb9u1) 6.3.0 20170516, 64-bit

 

It is a postgres docker image.

 

Thanks,

Rick

 

-----Original Message-----
From: Tom Lane <tgl@sss.pgh.pa.us>
Sent: Tuesday, April 7, 2020 7:09 AM
To: Justin Pryzby <pryzby@telsasoft.com>
Cc: Rick Vincent <rvincent@temenos.com>; pgsql-performance@postgresql.org; Manoj Kumar <manojkumar@temenos.com>; Herve Aubert <haubert@temenos.com>
Subject: Re: Postgres not using index on views

 

Justin Pryzby <pryzby@telsasoft.com> writes:

> On Mon, Apr 06, 2020 at 02:19:59PM +0000, Rick Vincent wrote:

>> The following query takes an extremely long time for only 180 rows, and what this means is that we would have to index anything appearing in the where clause for every table in order to use views because the views seem not to consider the select clause.  Why is that and does anyone know a way around this?

 

> Is there a reason why you don't store the extracted value in its own column ?

 

The planner seems to be quite well aware that the slower query is going to be slower, since the estimated costs are much higher.  Since it's not choosing to optimize into a faster form, I wonder whether it's constrained by semantic requirements.  In particular, I'm suspicious that some of those functions you have in the view are marked "volatile", preventing them from being optimized away.

 

Beyond that guess, though, there's really not enough info here to say.

The info we usually ask for to debug slow-query problems is explained at

 

https://wiki.postgresql.org/wiki/Slow_Query_Questions

 

                        regards, tom lane

 

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely for the addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended recipient is unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender. Please note that any views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts no liability for any damage caused by any malicious code or virus transmitted by this e-mail.

Rick Vincent schrieb am 07.04.2020 um 11:08:
> The function is defined as below, so no use of VOLATILE.

If you don't specify anything, the default is VOLATILE.

So your function *is* volatile.
 
> CREATE OR REPLACE FUNCTION extractValueJS (sVar text, nfm INTEGER, nvm INTEGER)
> RETURNS VARCHAR as $$
> declare
> sRet text := '';
> nSize int := 0;
> retVal int := 0;
> cVar text[] := regexp_split_to_array(sVar,'');
> idx int := 1;
> nStart int := 0;
> nEnd int := 0;
> begin
> etc...
>         return sRet;
> end;
> $$ LANGUAGE plpgsql;

You haven't shown us your actual code, but if you can turn that into a "language sql" function (defined as immutable,
orat least stable), I would expect it to be way more efficient. 

Thomas

> RV>> It simply is the way the application stores the data. For Oracle
> we are storing in XML and JSON format, for postgres, due do
> limitations of XML api, we are storing in VARCHAR.

Why not use JSON in Postgres then?
Postgres' JSON functions are at least as powerful as Oracle's  (if not better in a lot of areas).

Would be interesting to see what XML function/feature from Oracle you can't replicate/migrate to Postgres.

Another option might be to upgrade to Postgres 12 and define those columns as generated columns as part of the table,
ratherthan a view. 
Then you only pay the performance penalty of the extracValueJS() function when you update the table, not for every
select.

Thomas

Hi,

 

I was wondering if anyone can explain the below problem.  Should a bug be logged for this?

 

Kind regards,

Rick

 

_____________________________________________
From: Rick Vincent
Sent: Tuesday, April 7, 2020 11:08 AM
To: 'Tom Lane' <tgl@sss.pgh.pa.us>; Justin Pryzby <pryzby@telsasoft.com>
Cc: pgsql-performance@postgresql.org; Manoj Kumar <manojkumar@temenos.com>; Herve Aubert <haubert@temenos.com>
Subject: RE: Postgres not using index on views

 

 

Hi Tom,

 

The function is defined as below, so no use of VOLATILE.  Let me know if you need any other information.  I am hoping the below will further clarify the issue.

 

CREATE OR REPLACE FUNCTION extractValueJS (sVar text, nfm INTEGER, nvm INTEGER)

RETURNS VARCHAR as $$

declare

sRet text := '';

nSize int := 0;

retVal int := 0;

cVar text[] := regexp_split_to_array(sVar,'');

idx int := 1;

nStart int := 0;

nEnd int := 0;

begin

etc...

        return sRet;

end;

$$ LANGUAGE plpgsql;

 

After reading you link…..

 

Here is a better explain plan:

 

Explain on the table:

 

EXPLAIN (analyze,BUFFERS)

select RECID from "FBNK_CUSTOMER"

Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..22.80 rows=180 width=7) (actual time=0.011..0.073 rows=180 loops=1)

  Buffers: shared hit=21

Planning Time: 0.056 ms

Execution Time: 0.091 ms

 

Explain on the view:

 

EXPLAIN (analyze,BUFFERS)

select RECID from "V_FBNK_CUSTOMER"

 

Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19014.60 rows=180 width=7) (actual time=455.727..76837.097 rows=180 loops=1)

  Buffers: shared hit=204

  ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180 width=14575) (actual time=455.726..76836.791 rows=180 loops=1)

        Buffers: shared hit=204

Planning Time: 1.109 ms

Execution Time: 76838.505 ms

 

Explain on view with a column:

 

EXPLAIN (analyze,BUFFERS)

SELECT RECID FROM "V_FBNK_CUSTOMER" WHERE "TESTER" = '5.00' ORDER BY RECID

Sort  (cost=19015.06..19015.06 rows=1 width=7) (actual time=76033.475..76033.475 rows=1 loops=1)

  Sort Key: "V_FBNK_CUSTOMER".recid

  Sort Method: quicksort  Memory: 25kB

  Buffers: shared hit=21

  ->  Subquery Scan on "V_FBNK_CUSTOMER"  (cost=0.00..19015.05 rows=1 width=7) (actual time=66521.952..76033.434 rows=1 loops=1)

        Filter: (("V_FBNK_CUSTOMER"."TESTER")::text = '5.00'::text)

        Rows Removed by Filter: 179

        Buffers: shared hit=21

        ->  Seq Scan on "FBNK_CUSTOMER" a  (cost=0.00..19012.80 rows=180 width=14575) (actual time=462.949..76033.096 rows=180 loops=1)

              Buffers: shared hit=21

Planning Time: 0.819 ms

Execution Time: 76033.731 ms

 

But on the underlying table and not the view but just using the one view column called TESTER:

 

EXPLAIN (analyze,BUFFERS)

SELECT RECID FROM "FBNK_CUSTOMER" WHERE extractValueJS(XMLRECORD, 179, 9) = '5.00' ORDER BY RECID

Sort  (cost=68.26..68.27 rows=1 width=7) (actual time=220.403..220.404 rows=1 loops=1)

  Sort Key: recid

  Sort Method: quicksort  Memory: 25kB

  Buffers: shared hit=21

  ->  Seq Scan on "FBNK_CUSTOMER"  (cost=0.00..68.25 rows=1 width=7) (actual time=193.000..220.397 rows=1 loops=1)

        Filter: ((extractvaluejs((xmlrecord)::text, 179, 9))::text = '5.00'::text)

        Rows Removed by Filter: 179

        Buffers: shared hit=21

Planning Time: 0.045 ms

Execution Time: 220.418 ms

 

Other info:

 

SELECT relname, relpages, reltuples, relallvisible, relkind, relnatts, relhassubclass, reloptions, pg_table_size(oid) FROM pg_class WHERE relname='FBNK_CURRENCY';

 

relname relpages        reltuples       relallvisible   relkind relnatts        relhassubclass  reloptions      pg_table_size

FBNK_CURRENCY   6       93      0       r       2       false   NULL    81920

 

Version is:

PostgreSQL 11.7 (Debian 11.7-2.pgdg90+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 6.3.0-18+deb9u1) 6.3.0 20170516, 64-bit

 

It is a postgres docker image.

 

Thanks,

Rick

 

-----Original Message-----
From: Tom Lane <tgl@sss.pgh.pa.us>
Sent: Tuesday, April 7, 2020 7:09 AM
To: Justin Pryzby <pryzby@telsasoft.com>
Cc: Rick Vincent <rvincent@temenos.com>; pgsql-performance@postgresql.org; Manoj Kumar <manojkumar@temenos.com>; Herve Aubert <haubert@temenos.com>
Subject: Re: Postgres not using index on views

 

Justin Pryzby <pryzby@telsasoft.com> writes:

> On Mon, Apr 06, 2020 at 02:19:59PM +0000, Rick Vincent wrote:

>> The following query takes an extremely long time for only 180 rows, and what this means is that we would have to index anything appearing in the where clause for every table in order to use views because the views seem not to consider the select clause.  Why is that and does anyone know a way around this?

 

> Is there a reason why you don't store the extracted value in its own column ?

 

The planner seems to be quite well aware that the slower query is going to be slower, since the estimated costs are much higher.  Since it's not choosing to optimize into a faster form, I wonder whether it's constrained by semantic requirements.  In particular, I'm suspicious that some of those functions you have in the view are marked "volatile", preventing them from being optimized away.

 

Beyond that guess, though, there's really not enough info here to say.

The info we usually ask for to debug slow-query problems is explained at

 

https://wiki.postgresql.org/wiki/Slow_Query_Questions

 

                        regards, tom lane

 

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely for the addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended recipient is unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender. Please note that any views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts no liability for any damage caused by any malicious code or virus transmitted by this e-mail.

On Friday, April 17, 2020, Rick Vincent <rvincent@temenos.com> wrote:

Hi,

 

I was wondering if anyone can explain the below problem.  Should a bug be logged for this?

 

Kind regards,

Rick

 

_____________________________________________
From: Rick Vincent
Sent: Tuesday, April 7, 2020 11:08 AM
To: 'Tom Lane' <tgl@sss.pgh.pa.us>; Justin Pryzby <pryzby@telsasoft.com>
Cc: pgsql-performance@postgresql.org; Manoj Kumar <manojkumar@temenos.com>; Herve Aubert <haubert@temenos.com>
Subject: RE: Postgres not using index on views

 

 

Hi Tom,

 

The function is defined as below, so no use of VOLATILE.  Let me know if you need any other information.  I am hoping the below will further clarify the issue.

 

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely for the addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended recipient is unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender. Please note that any views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts no liability for any damage caused by any malicious code or virus transmitted by this e-mail.

The information in this e-mail and any attachments is confidential and may be legally privileged. It is intended solely for the addressee or addressees. Any use or disclosure of the contents of this e-mail/attachments by a not intended recipient is unauthorized and may be unlawful. If you have received this e-mail in error please notify the sender. Please note that any views or opinions presented in this e-mail are solely those of the author and do not necessarily represent those of TEMENOS. We recommend that you check this e-mail and any attachments against viruses. TEMENOS accepts no liability for any damage caused by any malicious code or virus transmitted by this e-mail.