Ciena NCS (Navigator) — network-level control
12 min
Site Manager handles one shelf. NCS (Navigator Network Control Suite, sometimes branded MCP — Manage, Control, Plan — depending on era) handles the whole network. It's the operator-facing software-defined transport (SDN) controller for Ciena gear: a server-side application (web UI + REST/NETCONF APIs) that knows every NE, every degree, every wavelength, and every service spanning them.
If Site Manager is the chassis's front panel, NCS is the network's NOC view.
What NCS does that Site Manager can't
| Capability | Why it matters | |---|---| | Multi-NE topology | One view of every shelf in your network with their photonic and OTN trails drawn between them. Click a fiber and see every wavelength it carries. | | Multi-layer correlation | IP / MPLS layer overlaid on optical — so when a lambda goes down, you immediately see which IP links and customer services it impacts. | | End-to-end service provisioning | Specify "100 Gbps Ethernet from Node A client to Node Z client." NCS computes the path, picks a wavelength (or wavelengths), pushes config to every NE along the way. | | Path Computation Element (PCE) | Constraint-based path selection: shortest, lowest-OSNR-impact, diverse from another service, must avoid a maintenance window, etc. | | Service templates | Reusable patterns: "L2 private line 10G," "100G transport with regen at hop 3," etc. | | Restoration / GMPLS | Configure pre-planned or dynamic photonic restoration paths for failure recovery. | | APIs and northbound integration | REST and NETCONF for OSS/BSS integration, automation, and customer self-service portals. |
A typical NCS workflow — provisioning end-to-end
The same wavelength you provisioned manually in the last lesson (NodeA → NodeB), done with NCS:
- Services → Create → Photonic Service
- Source: Node A, port 11-CLIENT-1
- Destination: Node B, port 8-CLIENT-1
- Service type: 100G transparent (OTU4)
- Constraints: avoid the 6500 at Node C (planned maintenance); diverse from existing service "Cust42-East"; lowest-OSNR-impact wavelength
- Click "Compute paths." NCS shows you 2–3 options with computed OSNR, latency, and shared-risk groups.
- Pick the preferred path. Click Provision.
- NCS pushes the per-NE config to every shelf on the path (Site Manager-equivalent steps, all at once), commissions the wavelength, validates that PMs are clean, and reports success.
What took 10 minutes of careful Site Manager work per NE × 2 NEs × hope-you-got-the-frequency-right is now one form and one button — and it's auditable, repeatable, and the path data is stored in NCS's inventory for the future.
Site Manager vs NCS — when to use which
| Use Site Manager when … | Use NCS when … | |---|---| | Investigating an alarm on one shelf | Investigating a customer service impact spanning multiple shelves | | Replacing a card or upgrading software on one shelf | Coordinating a software upgrade across a region | | First-time commissioning a brand-new NE | Provisioning any service that crosses more than one NE | | Detailed PM drilldown for an individual port | Service-level SLA reporting and trend analysis | | You don't have NCS deployed (smaller operators) | You do — and your processes are mature enough to discourage per-NE pokes |
In a healthy mature operator, day-to-day provisioning goes through NCS; Site Manager is reserved for one-off investigation and break-fix.
The northbound — automation and integration
NCS exposes REST and NETCONF APIs. Real benefits start when you wire it into:
- OSS/BSS for billing and service catalog
- CI/CD-style provisioning pipelines for repeatable service onboarding
- Network-wide assurance tools that compare planned vs actual state
- Customer portals for self-service (within limits)
This is what people mean when they say "the transport network has become SDN-capable" — the per-NE TL1 brittleness is hidden behind a model-driven API that treats the whole network as one programmable object.
What to remember
- NCS = network-level controller; Site Manager = per-NE.
- NCS provides multi-layer topology, PCE-based path computation, and end-to-end service workflows.
- The biggest day-to-day win is one click → consistent config on every NE on the path.
- APIs (REST/NETCONF) make NCS the right hook point for automation and OSS/BSS integration.