The Fundamental Difference
Simple Queue and Queue Tree are both bandwidth control mechanisms in MikroTik RouterOS, but they work very differently under the hood.
Simple Queue is a flat, per-target rate limiter. You specify a target IP or subnet and a max-limit — done. It is processed in order from top to bottom, and every packet is matched against every queue. This sequential processing is what kills performance at scale.
Queue Tree is a hierarchical system based on Hierarchical Token Bucket (HTB). Traffic is classified first using mangle marks, then the queue tree applies bandwidth policies based on those marks. This parallel classification is why Queue Tree scales to 10,000+ rules.
Feature Comparison
| Feature | Simple Queue | Queue Tree |
|---|---|---|
| Configuration complexity | Easy — target IP + rate | Moderate — requires mangle rules |
| Hierarchy support | Parent-child (limited) | Full HTB tree (unlimited depth) |
| Traffic classification | By IP/subnet only | By mangle marks (any criteria) |
| Processing | Sequential (top to bottom) | Parallel (mark-based lookup) |
| Max practical rules | ~200-500 | 10,000+ |
| PCQ support | Yes (limited) | Yes (full) |
| Burst support | Yes | Yes |
| Priority levels | 1-8 | 1-8 |
| CPU impact at scale | High | Low to moderate |
Simple Queue: Quick Setup
Simple Queue is great for small ISPs (under 200 subscribers) or for testing:
/queue simple add name=subscriber-001 target=10.10.1.2/32 max-limit=10M/10M burst-limit=15M/15M burst-threshold=8M/8M burst-time=10s/10s comment="John Doe - 10Mbps Plan" add name=subscriber-002 target=10.10.1.3/32 max-limit=25M/25M comment="Jane Smith - 25Mbps Plan"
Queue Tree: Scalable Setup
Queue Tree requires two steps: mark traffic with mangle, then apply queue policies:
# Step 1: Mark connections and packets in mangle /ip firewall mangle add chain=prerouting action=mark-connection new-connection-mark=conn-10mbps src-address=10.10.1.0/24 passthrough=yes add chain=prerouting action=mark-packet connection-mark=conn-10mbps new-packet-mark=pkt-10mbps passthrough=no # Step 2: Create queue tree hierarchy /queue tree add name=download parent=global max-limit=1G add name=plan-10mbps parent=download packet-mark=pkt-10mbps max-limit=10M queue=pcq-download-default
Performance Benchmarks
Real-world performance comparison on a MikroTik CCR1036 with 1,000 subscribers:
At 1,000 rules, Queue Tree uses roughly half the CPU. The gap widens as you add more rules.
| Metric | Simple Queue (1000 rules) | Queue Tree (1000 rules) |
|---|---|---|
| CPU usage (1 Gbps traffic) | 78-85% | 35-45% |
| Packet processing latency | 2.1ms average | 0.8ms average |
| Rule addition time | ~50ms per rule | ~5ms per rule |
| Max throughput before drops | ~1.8 Gbps | ~4.5 Gbps |
When to Use Each
- Simple Queue — fewer than 200 subscribers, single router, need quick setup without mangle complexity
- Queue Tree — 200+ subscribers, multiple service tiers, need traffic prioritisation (VoIP, gaming, bulk), or running PCQ for fair queuing
- Hybrid — use Simple Queue for per-subscriber limits and Queue Tree for global traffic prioritisation
How ISPbills Handles Both
ISPbills supports both bandwidth management approaches:
- RADIUS rate limits — ISPbills sets Mikrotik-Rate-Limit attributes that apply via the PPP profile, working with both queue types
- Package-based management — define bandwidth tiers in ISPbills and they are enforced automatically regardless of queue implementation
- Automatic provisioning — ISPbills creates the correct queue entries via MikroTik API when a subscriber is assigned a package
- Migration support — switch from Simple Queue to Queue Tree without changing your ISPbills configuration
Ready to Automate Your ISP?
ISPbills handles billing, network provisioning, and customer management. Use Free Demo to evaluate with up to 100 active subscribers.