AKKA CLUSTER MANAGEMENT AND SPLIT BRAIN RESOLUTION
summary
Akka cluster uses a gossip CRDT membership, heartbeat failure detector and deterministic leader with seed nodes; unreachable flags stop convergence so joins/removals block. Common split‑brain options include static quorum, majority, keep‑oldest and keep‑referee, and Akka’s weekly-up flag. A production resolver for two datacenters schedules a majority check (every ~7s), compares reachable vs unreachable counts, and issues down commands to the minority side as nodes are migrated between DCs; this approach resolves partitions and ran successfully in production.
AI-written summary · about
chapters (14)
- 0:00Start
- 1:23akka-cluster
- 3:47membership lifecycle
- 5:48seed nodes
- 6:24wait for cluster
- 7:34stopping actor system
- 9:32and then things go south
- 10:00network partitions
- 11:23split brain resolution
- 15:28register for cluster events
- 16:51current cluster state
- 18:00health indicator
- 18:23unreachable members
- 19:56downing
related talks
44mLITHIUM: AN OSS SPLIT-BRAIN RESOLVER FOR AKKA-CLUSTER
29mIS THERE ANYBODY OUT THERE?
41mHands on Introduction to Distributed Systems Concepts with Akka Clustering
45mIs there anybody out there?
25mGo distributed (and scale out) with Actors and Akka Clustering
01h 59mAkkaCluster and AkkaTyped DeepDive Into the ActorModelWith a BitofEyeCandy.
37mAkka Present and Future
48mDistributed Consensus Algorithms