Leader
node-1
stable
Term
7
current epoch
Quorum
5/5
Committed Writes
0
Commit Latency
8 ms
p50
BFT Tolerance
1 faulty
f = (5-1)/3
📊 Commit Throughput
commits/sproposals/s
💉 Quorum Votes
Votes received5 / 5
Majority needed3
Election stateSettled
🕒 Term History
🧩 Role Distribution
🔒 Protocol Status
AlgorithmRaft + BFT
Heartbeat150 ms
Election timeout300–600 ms
Log entries0
Snapshot index12480
🌐 Raft Cluster
5 nodes · term 7📤 Log Replication
per followerLeader commit index0
Max follower lag0 entries
⚡ Heartbeat Latency
RTT to followers4 ms
Elections (30d)
11
avg term 0.4/day
Committed (30d)
1.9 M
writes
Split-brain
2
auto-fenced
Avg Commit p99
21 ms
last 7d
📅 Commits per Hour
24h📐 Commit Latency Distribution
🌡 Leadership (node × hour)
5×24🧩 Vote Outcomes
Granted88%
Rejected (stale)9%
Timeout3%
📋 Event Log
All
| Time | Event | Severity | Detail |
|---|
🔔 Live Stream
tail🎛 Raft Tuning
150 ms
Shorter heartbeats detect leader loss faster but raise WiFi chatter.
300 ms
10000
🔒 Safety & BFT
Byzantine fault tolerance
Reject equivocating peers (3f+1)
Split-brain fencing
Demote minority partition leader
Pre-vote phase
Avoid disruptive term inflation
Read from followers
Serve stale reads off leader lease
📋 Cog Details
DescriptionRaft-based leader election with Byzantine fault tolerance. Quorum-based writes ensure consistency across the swarm. Split-brain protection.
Events
HardwareNone (WiFi only)
InputConsensus protocol
Versionv1.0.0 · 16 KB · Hard