The most common purpose of an interface to a network is Network Reconciliation and Auto Discovery. Auto Discovery in this case refers to potential differences discovered between the IMS database and what's in the live network. IMS will take a snapshot of the network data and compare this data with that held in the IMS database. This snapshot is referred to here as the Reconcile Table.
Reconcile Process
The following steps describe the reconciliation process:
1.Each night IMS collects data from the network.
2.The network data will be stored in the IMS Reconcile table in the database - referred often in this document as Reconcile Data Set
3.Each inventory item (e.g. equipment, connection) has a unique link ID associated to its related network element, as the IMS and network name could differ.
4.Connections running over multiple network platforms might have different network ID’s according to their platform, this enables IMS to accurately synchronize each IMS connection to its counterpart in the various network platforms.
5.IMS will query and compare the data held in the reconcile tables with the equivalent production IMS database tables.
There can be two options to take when differences arise between the NMS and IMS data:
1)Auto-discovery: IMS automatically modifies the data in the IMS database
2)IMS simply reveals the differences in the reconcile form
The frequency of the reconciliation process can be set by the IMS Admin user i.e., daily at a specific time or according to individual use cases.
Auto-Discovery
Here are some examples of discovered data that IMS can auto correct:
-New equipment in the actual network but not yet in IMS
-New connections in the actual network but not yet in IMS
-New inserted cards in the actual network but not yet in IMS
-Serial numbers found in the network that are different in IMS
-Software and Hardware versions
Customers decide which objects are to be automatically updated or simply reported as a mismatch.
Equipment Discovery
The reconcile and auto discovery process will typically collect equipment information such as:
•Equipment name
•Equipment type
•Equipment parameters
•Equipment shelf name
•Equipment shelf type
•Equipment shelf parameters
•Card name
•Card type, codes, and part numbers
•Card parameters
•Card slot positions
Limitation of network equipment data
A network may not always be able to provide sufficient information required by IMS, this is a limitation on the network side, yet VC4-IMS provides features to populate this missing information.
Some Equipment North Bound Interfaces (NBI) are unable to provide SFP/XFP pluggable card information. These cards often contain ports and VC4-IMS will need these ports in order to establish physical connections to it.
To overcome this issue VC4 has implemented the following methodology:
If IMS sees a connection on a port but this port is not present due to missing SFP/XFP information, IMS will create a dummy SFP/XFP with a port. Once IMS creates the port, then it can continue loading more connections into the database.
Should the vendor resolve the issue with the missing SFP/XFP data, IMS will subsequently replace the dummy card with the real one – this is non service affecting process.
Port discovery
Regarding the equipment auto discovery process, ports are not directly populated from the network data. IMS will populate its inventory according to each vendor’s card and SFP/XFP specifications.
So for example, if a card has 2 x 1GBE ports, IMS will automatically create those two ports when inserting this card in a slot. This will ensure business practice conformity.
Physical connection discovery
The successful discovery and loading of physical connections is dependent of the available network data.
Most NMS’ shold provide full information, as they often hold the full network topology. North Bound Interfaces like TMF CORBA and MTOSI provide the A-end and Z-end details of each physical connection.
Element Management Systems (EMS) provide less information as they only hold the equipment and not the connection data. To find the connectivity VC4-IMS must connect directly to the network elements (NE).
There are several ways to connect to NEs with protocols like SNMP, TL1, Telnet, or SSH. Different types of equipment may involve any of these protocols and when chosen, should provide for the best retrieval of information.
For physical connections, it is important to have the following basic information:
Node A, Shelf A, Slot/Card A, Port A and to which Node B, Shelf B, Slot/Card B, Port B.
Connection names are less important, because VC4-IMS can create those names automatically if needed.
The following are typical Port A and Port B discovery commands:
Neighbor commands:
▪Cisco router/switch: show cdp neighbors
▪Cisco router/switch: show lldp neighbors
▪Juniper router: show lldp neighbors
▪Cisco ONS: show topology neighbor detail
Then there are other vendor proprietary commands involving protocols such as:
▪CLI commands (command line interface)
▪TL1 commands
▪SNMP commands – often provide limited information
Due to network security reasons, these neighbor commands may not be enabled and a one time access will be required.
Workaround for when no port A and B commands are available
Often each port will have a comment or description field in the configuration of the NE. This port comment can be used to agree a standard naming convention to register the connected neighbor port, for example: <nodename> <shelfname> <port>.
Logical connection and service discovery
The discovery and loading of logical connections and services is similar to the physical connectivity and success will depend on the available network data.
Most NMS’ should provide full information, as they often hold the full network topology. North Bound Interfaces like TMF CORBA and MTOSI provide the logical connections and services as well as their routing through the network - their carriers i.e., :
-How a L2VPN or L3VPN is routed over a V-LAN.
-How a ODU2 (DWDM) connection is routed over a ODU3 or ODU4.
Element Management Systems (EMS) provide less information as they only hold the equipment and not the connection data. To find the connectivity where no NMS is available, VC4-IMS must connect directly to the network elements (NE).
Thre are several ways to connect to NEs with protocols like SNMP, TL1, Telnet, or SSH. Different types of equipment may involve any of these protocols and when chosen, should provide for the best retrieval of information.
Each vendor equipment type will have its own command set to retrieve the logical connections and service data.
Getting the end-to-end picture
VC4-IMS network discovery will try to construct an end-to-end view that is useful for network engineers. VC4-IMS topology map (seen above) will provide users with the ability to quickly correlate faults with network elements and circuits, thereby providing faster and better quality of service.