Thread: publication not found, yet it is there.??

publication not found, yet it is there.??

From
Dave Cramer
Date:
How is this possible?

I have checked pg_replication_slots and the slot exists. I have checked pg_publications and the publication exists.



2020-01-07 20:45:44.111 GMT [3768] DEBUG:  received replication command: CREATE_REPLICATION_SLOT debezium  LOGICAL pgoutput
2020-01-07 20:45:44.133 GMT [3768] DEBUG:  searching for logical decoding starting point, starting at 0/20B9640
2020-01-07 20:45:44.133 GMT [3768] LOG:  logical decoding found initial starting point at 0/20B9640
2020-01-07 20:45:44.133 GMT [3768] DETAIL:  Waiting for transactions (approximately 1) older than 629 to end.
2020-01-07 20:45:46.439 GMT [3768] LOG:  logical decoding found consistent point at 0/20BB328
2020-01-07 20:45:46.439 GMT [3768] DETAIL:  There are no running transactions.
2020-01-07 20:45:46.440 GMT [3768] LOG:  exported logical decoding snapshot: "00000005-00000009-1" with 0 transaction IDs
2020-01-07 20:45:46.798 GMT [3768] DEBUG:  received replication command: IDENTIFY_SYSTEM
2020-01-07 20:45:46.800 GMT [3768] LOG:  duration: 0.751 ms  statement: SELECT COUNT(1) FROM pg_publication WHERE pubname = 'dbz_publication'
2020-01-07 20:45:46.802 GMT [3768] LOG:  duration: 0.642 ms  statement: CREATE PUBLICATION dbz_publication FOR ALL TABLES;
2020-01-07 20:45:46.805 GMT [3768] DEBUG:  received replication command: START_REPLICATION SLOT debezium LOGICAL 0/20BB360 ("publication_names" 'dbz_publication', "proto_version" '1')
2020-01-07 20:45:46.805 GMT [3768] LOG:  starting logical decoding for slot "debezium"
2020-01-07 20:45:46.805 GMT [3768] DETAIL:  Streaming transactions committing after 0/20BB360, reading WAL from 0/20B9640.
2020-01-07 20:45:46.806 GMT [3768] LOG:  logical decoding found initial starting point at 0/20B9640
2020-01-07 20:45:46.806 GMT [3768] DETAIL:  Waiting for transactions (approximately 1) older than 629 to end.
2020-01-07 20:45:46.806 GMT [3768] LOG:  logical decoding found consistent point at 0/20BB328
2020-01-07 20:45:46.806 GMT [3768] DETAIL:  There are no running transactions.
2020-01-07 20:45:46.807 GMT [3768] DEBUG:  "PostgreSQL JDBC Driver" has now caught up with upstream server
2020-01-07 20:45:46.988 GMT [3768] DEBUG:  serializing snapshot to pg_logical/snapshots/0-20C2590.snap
2020-01-07 20:45:48.970 GMT [3768] ERROR:  publication "dbz_publication" does not exist
2020-01-07 20:45:48.970 GMT [3768] CONTEXT:  slot "debezium", output plugin "pgoutput", in the change callback, associated LSN 0/20BB360
Dave Cramer

Re: publication not found, yet it is there.??

From
Tomas Vondra
Date:
On Wed, Jan 08, 2020 at 08:35:28AM -0500, Dave Cramer wrote:
>How is this possible?
>
>I have checked pg_replication_slots and the slot exists. I have checked
>pg_publications and the publication exists.
>

Isn't this essentially the same issue as the one you reported about a
month ago on pgsql-hackers:

https://www.postgresql.org/message-id/CADK3HH%2B9uPg%3DanB-thAMojVhsZk3tYWMwux4YmHB2%2BByJXcE3w%40mail.gmail.com

I don't think we really identified what's going wrong there :-(


regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



Re: publication not found, yet it is there.??

From
Dave Cramer
Date:


On Wed, 8 Jan 2020 at 09:02, Tomas Vondra <tomas.vondra@2ndquadrant.com> wrote:
On Wed, Jan 08, 2020 at 08:35:28AM -0500, Dave Cramer wrote:
>How is this possible?
>
>I have checked pg_replication_slots and the slot exists. I have checked
>pg_publications and the publication exists.
>

Isn't this essentially the same issue as the one you reported about a
month ago on pgsql-hackers:

https://www.postgresql.org/message-id/CADK3HH%2B9uPg%3DanB-thAMojVhsZk3tYWMwux4YmHB2%2BByJXcE3w%40mail.gmail.com

I don't think we really identified what's going wrong there :-(

It is and I was hoping to get a few more eyes on this here.

I am going to do some more digging today.

Thanks 

Dave