[ https://issues.apache.org/jira/browse/CASSANDRA-15704?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
David Capwell updated CASSANDRA-15704: -------------------------------------- Status: Changes Suggested (was: Review In Progress) Feedback in GitHub. The main comment is that Netty keeps appending to the same buffer which leads to double counting, and we output frame as two buffers which leads to skewed histogram on the encoding side. It is a small change to move this logic to the two call sites which know the correct sizes. Other than that, LGTM. > add client request size metrics to netty pipeline > ------------------------------------------------- > > Key: CASSANDRA-15704 > URL: https://issues.apache.org/jira/browse/CASSANDRA-15704 > Project: Cassandra > Issue Type: Improvement > Components: Observability/Metrics > Reporter: Jon Haddad > Assignee: Jon Haddad > Priority: Normal > Labels: pull-request-available > Time Spent: 20m > Remaining Estimate: 0h > > We currently lack metrics around client connection incoming / outgoing bytes. > It’s fairly standard to know how many bytes are read and written to the > network, but that aggregates client facing and internal cluster traffic to a > single number. This patch will help us understand client overhead more > granularly. -- This message was sent by Atlassian Jira (v8.3.4#803005) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org