Maintenance Window Deep Dive
Understanding the 480-block confirmation cycle, timing strategies, and monitoring your window status.
Maintenance Window β Deep Dive
Every Flux node has a maintenance window β a periodic 480-block window (approximately 16 hours at 2 minutes per block) during which the node must be re-confirmed on-chain. Understanding this mechanism is critical for avoiding unexpected downtime and planning upgrades.
How It Works
The Flux blockchain tracks each node's last confirmed height β the block at which the node was last verified as active and healthy. Approximately every 480 blocks, the node must pass a health check and submit a new confirmation transaction. If the node fails to confirm within this window, it risks being placed on the DOS (Denial of Service) list and eventually removed from the active node list.
The maintenance window is NOT a scheduled downtime window. It is the grace period between confirmations. Your node should be online throughout β the "window" refers to the re-confirmation cycle.
Calculating Your Window
To determine your maintenance window status:
Maintenance Window Calculation
Blocks since confirmation = Current Block Height - Last Confirmed Height
Blocks remaining = 480 - Blocks since confirmation
Time remaining β Blocks remaining Γ 2 minutes| Blocks Remaining | Time Remaining | Status |
|---|---|---|
| > 240 | > 8 hours | Safe β plenty of time |
| 120β240 | 4β8 hours | Normal β monitor as usual |
| 60β120 | 2β4 hours | Attention β ensure node is healthy |
| < 60 | < 2 hours | Critical β investigate immediately |
| 0 | Expired | Closed β node needs re-confirmation |
What Happens When the Window Closes?
- 1
Warning state
As blocks remaining approach zero, the node enters a warning state. Watchdog (or ArcaneOS) may attempt to restart services and force a re-confirmation.
- 2
DOS list
If the node fails to confirm, it is placed on the Denial of Service list. This is a temporary state β the node can still recover.
- 3
Start list removal
Prolonged failure to confirm results in removal from the active node list. The collateral remains locked, but the node stops earning rewards.
- 4
Recovery
Restarting the node and ensuring it passes benchmarks will trigger a new confirmation cycle. There is no permanent penalty.
Best Practices
- β’Never reboot during low blocks remaining: Plan restarts and updates when you have 300+ blocks remaining in your window
- β’Use the Fluxme.io monitor: The Node Monitor page shows real-time maintenance window status with color-coded indicators
- β’Set up notifications: Configure Discord/Telegram alerts to warn you when blocks remaining drops below a threshold
- β’ArcaneOS advantage: ArcaneOS automatically manages confirmations and self-heals before the window closes
- β’Coordinate fleet updates: If managing multiple nodes, stagger updates so not all nodes are restarting simultaneously
Monitoring via API
You can check your node's confirmation status directly via the FluxOS API:
Check Maintenance Window via API
# Get current block height
curl -s https://api.runonflux.io/daemon/getblockcount | jq .data
# Get your node's status (includes last confirmed height)
curl -s http://YOUR_NODE_IP:16127/daemon/getfluxnodestatus | jq .dataOther articles in Advanced Operations
ArcaneOS Installation Guide
Step-by-step guide to installing ArcaneOS β the hardened OS for PNR-eligible Flux nodes.
Legacy vs ArcaneOS
Feature comparison, migration path, and decision guide for choosing between Legacy and ArcaneOS.
Multitoolbox Reference
Complete reference for all 14 Multitoolbox options β installation, diagnostics, repair, and configuration.
Troubleshooting & Diagnostics
Common node issues, log analysis, benchmark failures, DOS list recovery, and diagnostic techniques.
FluxOS API Reference
Essential API endpoints for monitoring, automation, and building tools on top of the Flux network.
Backup, Restore & Migration
Back up your node identity, restore from scratch, and migrate between hosts with zero collateral risk.
Multi-Node & Delegate Setup
Run multiple nodes with UPnP, Proxmox Fractus setups, and the P2SH delegate system.