Benoit Tellier created JAMES-3817:
-------------------------------------
Summary: Health Check for rabbitMQ dead letter queues
Key: JAMES-3817
URL: https://issues.apache.org/jira/browse/JAMES-3817
Project: James Server
Issue Type: Improvement
Components: rabbitmq
Reporter: Benoit Tellier
h3. The philosophy
If my data is stuck somewhere, I need alerting in place to "unstuck" it.
Health checks is a good way to do this:
- logs as WARN/ERROR
- HTTP calls from eg zabbix
- And hopefully that will be the first page of a webadmin UI!
h3. RabbitMQ deadletter
Data can end up in rabbitMQ errors if:
- Mail processing situation is desperate, cassandra mail repository keeps
failing, we requeued too many time / can't requeue
- Idem for event processing if saving to dead letter fails
- deserialisation errors
h3. Definition of Done
Write a health check that fails as degraded if james dead letter queues are not
empty.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscr...@james.apache.org
For additional commands, e-mail: server-dev-h...@james.apache.org
