Re: Plans with bad estimates/paths

2021-11-18 Thread Joe Wildish
Hi Tomas, On Tue, 16 Nov 2021, at 22:03, Tomas Vondra wrote: > It sure smells like a case of correlated data for some clients but not > others, but who knows ... Hard to say without seeing the nodes below the > join. If the lower nodes are estimated accurately, then it's the join > selectivity th

Re: check scripts after database code change

2021-11-18 Thread Adrian Klaver
On 11/18/21 06:45, Dennis wrote: Hi, Are there any scripts that we can check after the changes we made in the database code? Who made what changes in what code? In summary, are there any automatic post-development scripts before they come to the beta stage? Development of what and whose b

Re: check scripts after database code change

2021-11-18 Thread Karsten Hilbert
Am Thu, Nov 18, 2021 at 05:45:37PM +0300 schrieb Dennis: > Are there any scripts that we can check after the changes we made in the > database code? > In summary, are there any automatic post-development scripts before they > come to the beta stage? What do you want those script to do ? Karsten

Re: General Performance Question

2021-11-18 Thread Tom Lane
DAVID ROTH writes: > The code I am migrating uses the EXECUTE immediate form for almost > everything. The path of least resistance for me is to just copy the code to > Postgres and change "EXECUTE IMMEDIATE" to "EXECUTE". > I am asking about performance differences to see if it is worth convert

restore with pg engine upgrade in a middle

2021-11-18 Thread Marcin Giedz
Hi, the case is: one day we did full backup of database version 12 using standard procedure of continuous backup so: pg_start_backup/rsync/pg_stop_backup snip of WAL files archive directory: ver 12 PG Jul 28 04:37 000103340004 Jul 28 04:41 000103340005 Jul 28 05:16 000

Re: General Performance Question

2021-11-18 Thread DAVID ROTH
Agreed. The code I am migrating uses the EXECUTE immediate form for almost everything. The path of least resistance for me is to just copy the code to Postgres and change "EXECUTE IMMEDIATE" to "EXECUTE". I am asking about performance differences to see if it is worth converting code that does

check scripts after database code change

2021-11-18 Thread Dennis
Hi, Are there any scripts that we can check after the changes we made in the database code? In summary, are there any automatic post-development scripts before they come to the beta stage? Dennis

Re: General Performance Question

2021-11-18 Thread Thomas Kellerer
DAVID ROTH schrieb am 18.11.2021 um 15:15: > I am working on a large Oracle to Postgres migration. > The existing code frequently constructs a string and then uses Oracle's > "EXECUTE IMMEDIATE" to run it. > "EXECUTE" has the same functionality in Postgres. > > For example: > CREATE or REPLACE FUN

General Performance Question

2021-11-18 Thread DAVID ROTH
I am working on a large Oracle to Postgres migration. The existing code frequently constructs a string and then uses Oracle's "EXECUTE IMMEDIATE" to run it. "EXECUTE" has the same functionality in Postgres. For example: CREATE or REPLACE FUNCTION djr_foo_fnc (p_emp_no IN number) RETURN VARCHAR2 A

Re: Execute command in PL/pgSQL function not executing

2021-11-18 Thread Pavel Stehule
čt 18. 11. 2021 v 12:24 odesílatel Abdul Mohammed napsal: > Hello everyone, > Please I am having a problem with a function I am writing. The first part > uses a loop that pulls the values from a column and concatenates them into > a string. This first part works fine. The second part tries to use

Execute command in PL/pgSQL function not executing

2021-11-18 Thread Abdul Mohammed
Hello everyone, Please I am having a problem with a function I am writing. The first part uses a loop that pulls the values from a column and concatenates them into a string. This first part works fine. The second part tries to use the string to build a pivot table using the crosstab function. The