Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

TCX1000 Management Architecture

 

This topic describes the TCX1000 management architecture including the hardware operation and management communication ports and port rules.

Understanding the TCX1000 Management Architecture

Figure 1 shows the management network architecture for the TCX1000 Series, which is based on high availability (HA) communication or redundant paths between the proNX Optical Director and TCX1000 devices in the optical network. The management network that connects the proNX Optical Director to the devices that it manages is called the Data Communication Network (DCN).

Figure 1: TCX1000 Management Networking Architecture
TCX1000 Management Networking
Architecture

Along with the management traffic, the DCN carries the optical control traffic that allows the proNX Optical Director to control the optical links on the devices that it manages. In this role, the proNX Optical Director processes streams of optical link metrics from the TCX1000 devices and constantly adjusts the attenuators and amplifiers within each device to ensure optimal link performance at all times even as optical wavelength services are added and deleted from the network. This real-time control loop requires a continuous and reliable communication path between the proNX Optical Director and each TCX1000 device under control and management, which places strict requirements on the DCN.

High Availability

In order to provide the reliability required for the proNX Optical Director to act as an optical link controller, the DCN must be designed for high availability (HA). An HA network provides 99.999% availability, supports diverse links, and has no single point of failure. An HA system is intended for continuous operation and has redundant components and adequate backup and failover strategies. The DCN must be highly fault tolerant where a single failure does not isolate part of the network. The data center where the proNX Optical Director servers are situated must be hardened and be able to survive power outages and disastrous events.

Redundant Management Communications to Each TCX1000 Device

Similarly, every managed TCX1000 device (TCX1000-RDM20 and TCX1000-ILA must have redundant links to your Ethernet DCN HA for management communications. In Figure 1, we can see that all TCX1000 nodes in the optical network are connected to the DCN HA through an external Layer 2 switch and the Ethernet management communication ports on the TCX1000 devices.

Best Practice

The fastest and best method to use for management communications to TCX1000 devices is to connect the device’s Ethernet management communication ports to your Ethernet DCN HA using an external Layer 2 RSTP-enabled switch.

Figure 2: TCX1000-RDM20 and TCX1000-ILA Hardware Management Communication Architecture
TCX1000-RDM20 and TCX1000-ILA
Hardware Management Communication Architecture

TCX1000 IP Management Communications

TCX1000 devices uses two IP addresses for management communications:

  • Local IP address — used only to communicate locally with the TCX1000 device.

  • Remote IP address — used for all communications with the proNX Optical Director.

Note

The proNX Optical Director resides in your HA network and as such, the TCX1000 devices on your DCN HA are considered to be remotely managed.

For remotely managed TCX1000 devices, the proNX Optical Director always addresses management messages to the remote IP address of the TCX1000 device.

The protocol used to communicate with the proNX Optical Director is IP.

How the Communication Ports on TCX1000 Devices Work

Figure 2 illustrates the internal hardware management communications architecture used by the TCX1000-RDM20 (on the left) and TCX1000-ILA (on the right).

Both the TCX1000-RDM20 (left) and TCX1000-ILA (left) have an internal CPU to which all management communications occur. The external ports on each device communicate with the CPU through an internal RSTP-enabled Layer 2 switch. The proNX Optical Director management messages are sent to and from the CPU over external ports to the internal Layer 2 switch.

If a management message is addressed to the local TCX1000 device, the CPU responds to the message out the same port the message was received on. If the message is not addressed to the local TCX1000 device, the CPU consults its address table to identify which port it has seen the address on previously. If it has received messages from the address previously, it will have a record of which port it received the message on, and it will forward the message out that port. If the CPU consults its address table and finds no record of the address, it will broadcast the message out all ports.

TCX1000-RDM20 Management Communication Port Rules

Rules — For redundant management communication paths, you must connect the TCX1000-RDM20 device to your DCN HA using at least two of three ports described in Table 1.

Table 1: TCX1000-RDM20 Management Communications Port Rules

TCX1000-RDM20 Port

How Communications Work

Notes

DCN0

DCN1

Management messages received on either of the DCN ports are sent to the internal CPU. If the message is addressed to the local device, the CPU will act upon the message and respond over its DCN port.

If the message is addressed to a remote TCX1000 device connected to the TCX1000-RDM20 Line port the TCX1000-RDM20 forwards the message out its Line port if OSC forwarding is enabled.

If the message is addressed to a remote TCX1000 device connected through the TCX1000-RDM20’s other DCN port such, as in a multi-degree ROADM node configuration or when the DCN ports are daisy-chained together, the message is forwarded out the other DCN port

Best Practice: The DCN ports shown in Figure 2) are directly connected through the internal RSTP-enabled Layer 2 switch to the CPU and the DCN ports are connected to the DCN HA through an external RSTP-enabled Layer 2 switch. This is the recommended method of deployment.

 

Line

Management messages received on the Line port are sent to the internal CPU.

If the message is addressed to the local TCX1000-RDM20, the CPU will act upon the message and respond over its Line port.

If the message is addressed to a remote TCX1000 device connected through the TCX1000-RDM20’s DCN ports such, as in a multi-degree ROADM node configuration or when the DCN ports are daisy-chained together, the message is forwarded out the appropriate DCN port.

Note: In order for the TCX1000-RDM20 to exchange management messages over its Line port, the OSC forwarding feature must be enabled on the device. When this feature is enabled, the OSC signal is combined with the line WDM signal and sent over the fiber span connected to the line port. The OSC is a point-to-point channel that allows optical devices to communicate with each other for management and control purposes.

Note: TCX1000-RDM20 OSC forwarding is disabled by default.

Note: You cannot use OSC forwarding if the TCX1000-RDM20 line port is connected to a TCX1000-RDM20 or TCX1000-ILA that is on a different IP subnet.

ETH CRAFT

This port is used for local management communications to the internal CPU of the TCX1000-RDM20.

Note: The ETH CRAFT port (not shown in Figure 2) is directly connected to the internal CPU and the CPU uses a dedicated IP address to communicate with the ETH CRAFT port.

Remote Management Communications in Multi-Degree TCX1000-RDM20 Nodes

Multi-degree ROADM nodes have multiple TCX1000-RDM20 elements in a single node for multi-directional switching. Each TCX1000-RDM20 in the node can receive management messages over its line port, assuming you have enabled OSC forwarding on each device.

Daisy Chain the DCN Ports in Multi-Degree TCX1000-RDM20 Nodes for Redundant Management Communications

As a second path for management messages, in remotely connected multi-directional TCX1000-RDM20 nodes, you can daisy chain the DCN ports on the co-located TCX1000-RDM20s within the node in order to pass along management messages addressed to the co-located TCX1000-RDM20. If the management message is addressed to a TCX1000-RDM20 that is further down in the optical network, the co-located TCX1000-RDM20 forwards the message over its Line port as long as OSC forwarding is enabled.

TCX1000-ILA Management Communication Port Rules

The right-hand side of Figure 2 shows the internal hardware architecture of the TCX1000-ILA. You can communicate with the TCX1000-ILA in several ways as described in Table 2.

Note

The proNX Optical Director always addresses management messages to the IP address of the TCX1000-ILA internal CPU, regardless of whether the message arrives on one of the Line ports or the MGMT port.

The TCX1000-ILA has separate OSC signals for Line A and Line B that are used to exchange management messages addressed to remotely connected TCX1000-RDM20 and TCX1000-ILA devices. If the TCX1000-ILA receives a management message for a remotely connected TCX1000-RDM20 or TCX1000-ILA, it forwards the message out the appropriate Line port. This forwarding of management messages over its line ports is not configurable like it is on the TCX1000-RDM20; you cannot disable management message forwarding on the TCX1000-ILA.

For redundant management communication paths, you must connect the TCX1000-ILA to your DCN HA using at least two of three ports described in Table 2.

Table 2: TCX1000-ILA Management Communication Port Rules

TCX1000-ILA Port

How Communications Work

Notes

MGMT (Ethernet) port

Management messages received on the MGMT port are sent to the internal CPU. If the message is addressed to the local device, the CPU will act upon the message and respond over its MGMT port.

If the message is addressed to a remote TCX1000 device, the CPU consults its address table and forwards the message out either Line A Out or Line B Out port if the address is known, or broadcasts the message out both line ports.

Best Practice: Connecting the TCX1000-ILA MGMT port to an external Layer 2 switch that is connected to the same DCN HA as the proNX Optical Director is the best method for communicating with the TCX1000-ILA.

Line A

Management messages received on the Line A In port are sent to the internal CPU. If the message is addressed to the device, the CPU will act upon the message and respond over its Line A Out port.

If the message is addressed to a remote TCX1000 device, the CPU consults its address table and forwards the message out either Line B Out port or the MGMT port.

 

Line B

Management messages received on the Line B In port are sent to the internal CPU. If the message is addressed to the device, the CPU will act upon the message and respond over its Line B Out port.

If the message is addressed to a remote TCX1000 device, the CPU consults its address table and forwards the message out either Line A Out port or the MGMT port.