Service providers have traditionally relied on overcapacity in the network to ensure adequate service delivery for their clients. With service provider revenues under pressure from increased competition, justifying the costs associated with maintaining that overcapacity is proving to be more difficult. Numerous service providers have moved to address this issue by deploying large network inventory management systems (NIMS) to better understand actual capacity and, therefore, better manage the requirement for future capacity.
Many of these projects have struggled. Why?
There are two views to the network - the actual network and the perceived network. A NIMS has a perceived view to the network; rarely does it accurately reflect the actual network. If the discrepancy between perceived and actual is too great, the value of the data contained in the NIMS is questionable. Finding a way to reconcile the actual network with the perceived network is critical to success.
Traditionally reconciliation has been a manual process; use a team of people to collect current data from the actual network and then manually compare it to and update the data contained in the NIMS. This is a costly proposition. Not only must it be done during the deployment of the NIMS, it must be performed on a regular basis in order to ensure on going data accuracy.
There are reconciliation solutions in the market that help to automate the collection of data from the actual network. What these solutions fail to address is the on going reconciliation of that data with the NIMS; this remains a costly manual process. This is where NIMS projects struggle. Never-ending resource costs and manual data entry errors attributed to the reconciliation process combine to negatively impact the benefits.
NIMS projects can be successful. What is required is a complete automated solution: synchronization. Discovery of the actual network is automated as well as the reconciliation process with the NIMS. Manual effort is reduced, data accuracy is increased and never ending costs are better managed.
Network Inventory Implementation
Network inventory is an enterprise-wide resource that must be available to multiple teams within the organization such as network planning, design and assign, provisioning, etc. Since multiple teams are involved, either independently or interdependently, the ability to access correct and up-to-date network inventory is essential.
Knowing the actual network configuration in today’s networks is challenging. Additions of new services, dynamic nature of network devices, and unrecorded network maintenance and repair changes, result in the inventory database and the network growing increasingly out of sync.
The cost of not having an accurate network inventory is large as it significantly affects the operations of many teams and workgroups within the service provider.
Such costs are:
• Planning produces incorrect designs
• Provisioning fall-out is increased
• Problem resolutions takes longer
All of these costs continue to have a direct dramatic impact on customer experience.
Reconciliation versus Synchronization
Reconciliation and Synchronization are both first based on discovery. Discovery in both cases is retrieving information from the network elements and using it to populate physical and logical configurations in the network inventory system. Physical information may include rack, shelf, card and port information depending on the equipment involved but at least card and port information is readily available. Logical information may include logical ports, channels and internal cross connections of the device.
Reconciliation is the manual process by which a team of people use the report output from a discovery tool to identify and address discrepancies between the actual network and the inventory management system
FIGURE 1 – Typical Reconciliation Model
As shown above, typical reconciliation scenarios imply only “confirmation” of what has been discovered against the network database, which is the simplest of tasks, and leaving everything else to be worked outside of the inventory application using compiled reports to manually understand and fix discrepancies.
Synchronization is automated reconciliation. TierOne has accomplished this by developing a solution that automatically addresses discrepancies, leaving only those few discrepancies that it can't resolve automatically for manual intervention.
It encompasses the innovative idea of creating logical connectivity oriented services automatically in a hierarchical fashion by “stitching” pieces of information as they are being discovered and synchronizing them against the database through a set of well established predefined rules.
Essentially, this is creating a new database and keeping it synchronized embodied within one solution.
FIGURE 2 – TierOne™ Synchronization Model
The additional advantage of this solution is that yet another application (see previous view) is not being introduced to the end users who typically already have a plethora of systems to manage. The reconciliation is built into the inventory management system allowing the end user to view inventory data that will be using and having the discrepancies available within the same environment.
Key Success Criteria
The key success criteria was to create a solution where the inventory information is periodically synchronized with the network and only where user intervention is required is the end user tasked with resolution problems.
In essence, a feedback loop was created to keep the discrepancy in check and automatically create or modify records in the database to accurately reflect the network.
The focus was also to use TierOne’s extensive understanding of circuits and facilities to automatically “stitch” or create “logical” circuits based on information provided from the network and an understanding of the customers naming convention in a repetitive fashion that will allow the network database to continually stay in sync with the network.
This process of creating services would also allow data migration of network data with minimal user input when deploying a new inventory management system and significantly increase the accuracy of the resulting initial data load.
