Re: How is this possible "publication does not exist" - Mailing list pgsql-bugs

From operations i
Subject Re: How is this possible "publication does not exist"
Date
Msg-id CANWRaJx+SWirb-wL0Zax748hgEZ_x1D0KYbLUzuypD63fvFnPA@mail.gmail.com
Whole thread Raw
In response to Re: How is this possible "publication does not exist"  (operations i <ioperations.c@gmail.com>)
List pgsql-bugs
Ok, I have got the real problem, It's very kind of you for your quick response.

On Thu, Jun 2, 2022 at 6:58 PM operations i <ioperations.c@gmail.com> wrote:
cool , you have send faster than me for about 5 seconds

On Thu, Jun 2, 2022 at 6:57 PM operations i <ioperations.c@gmail.com> wrote:
after several times debugging, I 've realized the key point is the order of publication creation and insert statement, and both two have reached catcache.c:SearchCatCacheMiss() .and below has the difference explained
1. publication create first and then insert data into table
       the expression ```HeapTupleIsValid(ntp = systable_getnext(scandesc))``` return true

2.insert data into table and then create publication
      the expression ```HeapTupleIsValid(ntp = systable_getnext(scandesc))``` return false

so in case 2 , when decoding the first change , system table snapshot do not have Publication 'pub_test' , so throw the error
        " publication 'pub_test'" not exists 


pgsql-bugs by date:

Previous
From: operations i
Date:
Subject: Re: How is this possible "publication does not exist"
Next
From: "Collins, John D."
Date:
Subject: Error installing Stack Builder on Windows 64bit