Vanlightly commented on a change in pull request #2835: URL: https://github.com/apache/bookkeeper/pull/2835#discussion_r730906765
########## File path: site/bps/BP-44-use-metrics.md ########## @@ -0,0 +1,164 @@ +--- +title: "BP-44: USE metrics" +issue: https://github.com/apache/bookkeeper/issues/2834 +state: "Under Discussion" +release: "N/A" +--- + +### Motivation +Based on our experience (at Splunk) running many BookKeeper clusters in production, from very small to very large deployments (in terms of number of bookies, size of VMs and load) we have identified a number of short-comings with the current BookKeeper metrics that make it harder than it should be to identify bottlenecks in performance. The USE method is an effective strategy for diagnosing where bottlenecks in a system lie but the current metrics do not always expose metrics related to utilization and saturation. Also, even if you have a good mental model for how BookKeeper works internally, there are blindspots in the metrics that make it difficult to know what is happening at times. Review comment: Certainly. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
