Ekinops Dark Fiber Learning Path
Stage 5

Ekinops360 — Platform Overview

Now map the physics onto the platform. This page stays at the concept level on purpose — it teaches how Ekinops360 packages transport into shelves and modules, not a parts list. Learn the shape of the platform and you can read any Ekinops design; the exact modules you confirm from current documentation.

Common mistake — read before you spec anything

Do not memorize or quote module/SKU part numbers, port counts, or rate limits from training material. Supported modules, rates, and options change across hardware generations and software releases. Verify exact module/SKU/rate support against current Ekinops documentation or the project BOM every time. This page deliberately names no part numbers and gives no fixed rate table — those belong in the docs, not your memory.

What Ekinops360 is

Ekinops360 is an optical transport platform: a chassis/shelf system that hosts pluggable modules (transponders, muxponders, amplifiers, filters, line-system elements) which together turn dark fiber into managed WDM services. Everything from the previous lessons — client side/line side, transponders, muxponders, mux/demux, amplification, OADM/ROADM, OSC — is implemented here as modules you select and populate into a shelf.

It's managed as a system (see the Celestis NMS lesson), so you get inventory, alarms, performance data, and end-to-end service visibility across every shelf in the network rather than box-by-box.

Mental model

Think chassis + modules. The shelf is a chassis that gives you slots, power, cooling, and a management interface. You populate its slots with modules, and each module plays one of the optical roles — convert, combine/split, amplify, or carry management. Designing an Ekinops link is mostly deciding which roles you need and which modules fill them, then choosing a shelf with enough slots to hold them.

Platform building blocks

EKINOPS360 as a stack of building blocks ┌──────────────────────────────────────────────────────────┐ │ CELESTIS NMS (management on top — inventory/alarms/PM) │ └──────────────────────────────────────────────────────────┘ ▲ managed as one system ┌──────────────────────────────────────────────────────────┐ │ SHELF / CHASSIS │ │ power · cooling · management/control · N slots │ │ ┌───────┐ ┌───────┐ ┌───────┐ ┌───────┐ ┌───────┐ │ │ │ CLIENT│ │ LINE │ │ MUX- │ │ AMP │ │ OADM/ │ ... │ │ │ module│ │ module│ │ ponder│ │ (EDFA)│ │ ROADM │ │ │ └───┬───┘ └───┬───┘ └───────┘ └───────┘ └───────┘ │ └──────┼─────────┼───────────────────────────────────────────┘ grey │ client │ line side (coloured λ) optics▼ side ▼════════════════► onto the DWDM line system routers/switches/SAN (the "road" the wavelengths drive on)

Shelf / chassis

The frame that provides slots, power, cooling, and management. Shelves come in different sizes (from compact edge units to larger aggregation shelves). Slot count and power draw are design constraints — verify capacities against current docs.

Slots

Where modules seat. A design's density and growth headroom come down to slots available vs slots consumed. Reserve slots for growth and for the protect side of protected services.

Line modules

The line-side transport cards — they generate the coloured, tunable DWDM wavelength that travels the fiber. This is where line rate, modulation, and reach live. Often the FlexRate-style cards below.

Client modules

The client-facing cards / ports that accept grey optics from routers, switches, and SAN. The client/line split can be within one transponder card or across separate modules — verify the specific module's architecture.

The client / line module split

Mental model

Keep the two faces from the Optical Transport lesson front of mind: client modules face your world (grey optics, standard rates), line modules face the optical network (coloured λ, ITU channel, tunable). On some Ekinops cards both faces live on one module (a self-contained transponder/muxponder); on others the client and line functions are separate modules that pair up. When you read a BOM, always ask which model you're looking at — it changes slot count and cabling.

FlexRate — the provisionable-rate concept

FlexRate is the idea of a line module whose line rate and format are provisionable rather than fixed in hardware — one card can be set to operate across a range of line rates, so a single card type serves several service speeds and evolves with need instead of being replaced. Conceptually this is like a NIC that negotiates its speed, versus one hard-wired to a single rate.

The payoff is fewer distinct cards to stock, and the ability to raise a link's rate by re-provisioning instead of re-carding. The exact supported rates, modulation formats, and reach for any FlexRate-style module must be verified against current Ekinops documentation or the BOM — do not assume a rate is supported because it exists elsewhere in the family.

Common mistake

Treating "FlexRate" as "any rate, any reach, always." Provisionable ≠ unlimited. A given module supports a specific set of rates/formats, and higher rates typically cost reach and OSNR margin. Always confirm the exact rate and the reach at that rate for the actual module in the design.

The two platform roles: WDM transport + OTN aggregation

WDM transport

The platform's central job: carry many services over one fiber pair using CWDM/DWDM wavelengths, combined and separated by its optical line system. This is the "put colours on the fiber and move them" role.

OTN aggregation

Layered in via muxponders/OTN mapping: wrap and groom several client services into a common digital container so they share one wavelength while staying independently monitored and switchable (see the Protection & OTN page).

The optical line system — the "road" the wavelengths drive on

Transponders and muxponders make wavelengths; the optical line system is the infrastructure those wavelengths travel over between sites. It's the amplifiers (EDFA/Raman), mux/demux, OADM/ROADM, and OSC working together as one managed road. A common way to reason about an Ekinops design is to separate the services (the transponder/muxponder cards that add/drop traffic) from the road (the line system that carries and manages the light).

SERVICES ride on the ROAD [transponder]──┐ ┌──[transponder] [muxponder ]──►│ add/drop │ ═══ line system ═══ │ add/drop │◄──[muxponder] └──────────┘ └──────────┘ mux/demux · EDFA/Raman · OADM/ROADM · OSC <----------------- the ROAD -----------------> Design tip: size the road (reach, amps, add/drop) and the services (rates, wavelengths) as two related but separate problems.

Passive vs active building blocks

TypeExamplesPowered?What to check
PassiveFilters, mux/demux, fixed OADM, patch/splitter modulesNo power, no alarms of their ownInsertion loss, correct port/channel mapping, clean connectors — a passive fault shows up as loss/absence downstream
ActiveTransponders, muxponders, amplifiers, ROADM/WSS, OSCPowered, managed, alarmedPower levels, config, alarms, software/firmware, redundancy
Key idea

Passive elements are silent: they never raise an alarm, so their faults appear as loss or missing light at the next active element. When you troubleshoot, the passive gear is often the culprit the alarms point through, not at. Account for every passive insertion loss in the loss budget.

Open / disaggregated & white-box line systems

Modern optical architectures increasingly disaggregate the line system from the transponders — mixing an open/white-box optical line system with transponders (or router pluggables) from another vendor. Ekinops participates in this open model, so you may see designs where Ekinops provides the line system, or the transponders, or both.

This is where alien-wavelength support matters: the line system carries a third-party wavelength transparently. It buys flexibility and vendor choice, but it moves OSNR/reach responsibility for that alien λ onto the design team rather than a single-vendor turnkey tool. Confirm which parts are Ekinops and which are alien, and who owns the optical engineering for each.

Where Ekinops360 fits (conceptual)

Metro / regional

Rings and linear add/drop across a city or region — many add/drop sites, moderate reach. The bread-and-butter of dark-fiber transport.

DCI

High-capacity, low-latency point-to-point between data centers — often coherent, sometimes router pluggables as alien wavelengths.

Enterprise / service provider

From a single enterprise leasing dark fiber between two campuses, to a carrier selling wavelength services and grooming many customers with OTN.

These are conceptual placements to build intuition, not a positioning statement. Which shelf/module family fits a given deployment must be confirmed against current Ekinops documentation.

Key idea

You don't need to memorize the Ekinops catalog to be effective. If you can state the role each part of a link must play — client conversion, aggregation, wavelength combine/split, amplification, add/drop/routing, management — you can read any Ekinops design and ask precise questions about the modules chosen to fill those roles.

What to ask when someone hands you an Ekinops design

Treat this as your intake checklist. Each item lists why it matters and what the answer changes — so a checked box means you understood the consequence, not just read the field.

Field checklist — Ekinops design intake (with the "so what")
Common mistake

Accepting a design without a loss budget and a management plan. A shelf full of correct modules still fails if the span exceeds budget or you can't reach the remote shelf. Tie this checklist back to the loss-budget lesson and the OSC/management path every time — and remember every SKU on the BOM is a claim to verify against current Ekinops documentation, not a fact to trust.

1. Why does this page refuse to list module SKUs and fixed rate tables?

Because supported modules/SKUs, rates, and reach change across hardware generations and software releases; quoting them from training risks a wrong BOM. Always verify against current Ekinops documentation or the project BOM.

2. What does "FlexRate" describe conceptually, and what's the catch?

A line module whose line rate/format is provisionable across a range, so one card serves several service speeds. The catch: provisionable ≠ unlimited — each module supports a specific set of rates, and higher rates typically cost reach/OSNR margin. Verify the exact rate and its reach.

3. In an open/disaggregated design, an Ekinops line system carries a third-party 400G wavelength. Who owns the OSNR/reach engineering for it?

The design team — it's an alien wavelength, so the line system carries it transparently and can't see inside it. That flexibility moves optical-engineering responsibility for that λ off a single-vendor turnkey tool and onto you.

4. Someone hands you an Ekinops BOM with no wavelength plan and no loss budget. What do you send back?

Which ITU channel each service uses, whether it's native or alien, and how that maps to mux/demux ports — plus a per-span loss budget and the management/OSC reachability plan. Without those, the module list can't be validated.