In order to make your issue reports as actionable as possible, please provide the following information, depending on the issue type.

Bug report

For bug reports, provide this information, please:

What Debezium connector do you use and what version?

io.debezium.connector.postgresql.PostgresConnector

version 1.9.2 Final

What is the connector configuration?

tasks.max: 1
plugin.name: pgoutput
database.sslmode: "require"
database.server.name: publish
database.hostname: "postgresql-ha-pgpool"

database.port: 5432
database.user: "${directory:/opt/kafka/external-configuration/connector-config-postgresql-username:postgresql-username}"
database.password: "${directory:/opt/kafka/external-configuration/connector-config-postgresql-password:postgresql-password}"
database.dbname: mydb
table.include.list: "public.recommendation_p([0-9]+)"
column.include.list: "public.recommendation_p([0-9]).status,public.recommendation_p([0-9]).path,public.mytable_p([0-9]+).site_id"
message.key.columns: "public.recommendation_p([0-9]+):site_id"
transforms: Reroute,filter,unwrap
transforms.Reroute.type: io.debezium.transforms.ByLogicalTableRouter
transforms.Reroute.topic.regex: publish.public.recommendation_p([0-9]+)
transforms.Reroute.topic.replacement: publish.public.recommendation
transforms.Reroute.key.enforce.uniqueness: false
transforms.filter.type: io.debezium.transforms.Filter
transforms.filter.language: jsr223.groovy
transforms.filter.condition: value.op == 'u' && value.after.status == 'PUBLISH_INITIATED'
transforms.filter.topic.regex: publish.public.recommendation
transforms.unwrap.type: io.debezium.transforms.ExtractNewRecordState

What is the captured database version and mode of depoyment?

(E.g. on-premises, with a specific cloud provider, etc.)

Postgres 13.4

It is deployed together with repmgr and pgpool on the Amazon EKS cluster for high availability.

What behaviour do you expect?

It's very low-latency from creating the data change events in the postgres to getting messages from the kafka topic. Also, there shouldn't be any errors in the logs of debezium.

What behaviour do you see?

We see there are some networking issue logs between debezium and postgres. Seems like debezium are reading from/writing to the other end which is already closed. The error looks similar to this issue, but it said it was fixed already in the earlier version. Also, the performance of debezium is not consistent, sometimes it takes few seconds to receive the messages from kafka, sometimes it takes up to several minutes for the consumer to receive. We are wondering whether this behavior is expected or not.

Do you see the same behaviour using the latest relesead Debezium version?

(Ideally, also verify with latest Alpha/Beta/CR version)

We only use 1.9.2 Final version.

Do you have the connector logs, ideally from start till finish?

(You might be asked later to provide DEBUG/TRACE level log)

The connector log is attached below. Basically the log messages in the attachement is repeated as the debezium is running.

How to reproduce the issue using our tutorial deployment?

Our debezium, postgres and kafka cluster are all deployed on the Amazon EKS environment, the tutorial here might not reproduce the issue.

Feature request or enhancement

For feature requests or enhancements, provide this information, please:

Which use case/requirement will be addressed by the proposed feature?

<Your answer>

Implementation ideas (optional)

<Your answer>