Hi, I work in ESB/message-integration technical support, and I built and maintain a small open-source project called MQ Watcher.
It started after a few broker incidents where I needed to inspect an ActiveMQ Store without starting another broker against it or allowing a recovery path to change the evidence. Raw string searches were useful, but it was easy to read too much into them, and correlating messages, acknowledgements, subscriptions and transactions by hand became cumbersome. MQ Watcher is a local, read-only ActiveMQ Classic Store evidence workbench. It can currently: - inspect selected structured KahaDB journal evidence - correlate messages, ACK/remove records, subscriptions and transactions - compare two Store snapshots - keep local investigation notes and evidence references - browse evidence by journal file and offset - export a redacted offline evidence bundle The boundaries are intentional: - it does not start a broker or perform recovery - it does not write to or upload the selected Store - it does not determine the root cause automatically - journal evidence is not presented as proof of the broker's current state - unsupported values remain Unknown, Partial or Unsupported Broker-generated fixture scenarios currently cover ActiveMQ Classic 5.13.5, 5.15.16 and 5.18.7. This is not intended to claim compatibility with every version, patch level or damaged Store. The repository includes standalone Windows and Linux builds: https://github.com/kutaelee/mq-watcher I am not affiliated with the Apache ActiveMQ project. I am sharing it because the need came directly from support work, and it may save someone else some investigation time. Feedback from ActiveMQ operators would be particularly useful around Store layouts that do not parse, evidence normally needed during incidents, and cases where existing scripts or tools still provide a better workflow. Regards, Kutae Lee
