jbertram commented on PR #4844: URL: https://github.com/apache/activemq-artemis/pull/4844#issuecomment-1998119785
Generally speaking, exported metrics are not meant for debugging, per se. Exported metrics are really designed for runtime monitoring which is why we don't export everything but only the stuff that really matters for understanding how the broker is functioning in real time. This keeps the metrics' footprint relatively low and performance relatively high. Metrics are often consumed by external systems that are not only used to visualize the data but are also used to alert on specific conditions. With that in mind, metrics should be something that can either be meaningfully visualized over time (i.e. the size of a queue) or whose state might indicate a problem (e.g. a passive backup broker that becomes active might trigger an alert). It's not clear to me where the activation sequence fits in to this framework. It seems like it would be better to log activation sequence changes and if you need to debug something then you can take a look at the log. At this point, I don't see how it would be specifically useful to expose this as a metric. Do you have a particular use-case in mind? Regarding the "synchronizing" metric you added, please elaborate on how it's functioning now and how you expect it to function. Also, it's worth noting that the "synchronizing" and "replica.sync" metrics are looking at the _same_ underlying data so one is redundant and should be removed. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: gitbox-unsubscr...@activemq.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
