could not refresh group config [SQL0060]
Edit this pageAs it arrives in your log — example values stand where yours will:
level=WARN msg="could not refresh group config" code=SQL0060 schema=sqlstreams group=charge-cards
A running instance polls its message_consumer or exception_consumer
row on ConsumeOptions.ConfigRefreshInterval to pick up a redeclared
group config. This read did not land, so the instance keeps consuming
under the copy it already resolved. Claiming never touches the row, so
nothing stalls - the instance is simply running an older config than
whatever is stored.
One line is a database blip and the next poll picks the change up. The
same line repeating is the case that matters: the stored document
cannot be parsed or does not validate, and this instance will never
converge on it. A redeclaration that landed in worker_config then
reaches some replicas and not others, which is the one way replicas in
a group drift apart on config.
The worker attribute names which row failed - a group has two, and
they carry separate documents. Run the diagnose query to see what is
stored:
sqlstreams explain SQL0060Repeats of this line inside a one-minute window collapse to the first,
and the next one carries the dropped total as suppressed_count.
Helpful info.
the config document stored on this group's worker rows
SELECT worker_config.name, worker_config.metadata FROM {schema}.worker_config JOIN {schema}.consumer_group_config ON consumer_group_config.id = worker_config.consumer_group_id WHERE consumer_group_config.name = '{group}' ORDER BY worker_config.name;
Looking for a different code? Search every thread.