Tracing of Large-scale Actor Systems
summary
Provides a concrete Scala tracing library called Akka Tracing Tool that instruments send calls via an SBT/aspect step and collects traces into local CouchDB instances with eventual CouchDB replication to a central store. Tested on a traffic simulation up to about 50 nodes, it proves scalable by node count but imposes a heavy runtime cost of roughly 39–45% FPS loss and shows a sharp overhead when capturing 100% of messages; visualization is basic and time is not shown.
AI-written summary · about
chapters (6)
- 0:00Start
- 4:37Scala Library
- 8:12Tracing of Large Scale Actor Systems
- 9:26Scalability of the Library
- 14:49The Causes of the Performance Overhead of Your Library
- 19:00Can You Dynamically Switched It On and off or for Example Trace Messages Originating Only from One Specific Actor without Turning Off the System
related talks
31mDistributed Application Tracing with Trace4Cats
57mDistributed Tracing: How the Pros Debug Concurrent and Distributed Systems
15mTRACING AKKA STREAMS
34mDistributed Tracing, Lessons Learned
41mMonitoring Reactive Applications
39mSupervise Your Akka Actors
54mA TRACE TO REMEMBER: COMPOSITIONAL TRACING
49mDistributed Tracing: How the Pros Debug Concurrent and Distributed Systems