Case Study : Network Data Reconciliation

Client: Top US MNO
Network Data Reconciliation Sincera
Business Problem:
Numerous instances of incorrect data in the inventory database vs. actual network deployment causing significant automation fallouts and manual processes.
  • Engineering and planning processes requiring manual swivel chair resulting in delays and inefficiencies
  • Service Assurance functions such as troubleshooting, planned maintenance require manual correlation between different network data sources and inventory system
  • Producing reports of existing network assets (e.g. active, spare, pending, etc.) are inaccurate resulting in CAPEX inefficiencies
  • Various automated provisioning and activation processes (e.g. Network Elements IP address assignment/release) are hampered with fallouts resulting in inefficient manual intervention

Using 1Data, network data from various sources (EMSs) and are ingested and
normalized for comparison to Inventory System data.

Data Sources:

  • Inventory System RDBMS containing 800K+ network devices and 3M+ circuits/paths
  • Fronthaul network EMSs data for three different vendors in varying XML formats
  • Backhaul network EMS data for two different vendors in RDBMS tables
  • Location data from two different sources via API

Drools Rule Engine, with over 100 custom-defined rules, is used to generate a comprehensive list of inconsistencies between Network and Inventory system with accompanying corrective actions. Some examples are:

  • Identifying equipment active in network, but not existing in Inventory System. Discovering the missing equipment location from another system and auto-creating them in the Inventory System
  • Identifying missing circuits/paths in Inventory System which are exiting and active in the network and auto-creating them with standard naming
  • Identifying equipment in Inventory System which are not named exactly as their network ID and Auto-Fixing them
  • Identifying equipment in Inventory System which have been long disabled or removed from network and Recommending deletion
  • 1Data Data Management Workflow is used to bulk-load corrections into the Inventory System and assign/track Recommended and Manual data fixes

Benefits:

  • Within 2 weeks of deployment, Network vs. Inventory inconsistencies were reduced by 70%
  • Service Assurance troubleshooting processes and MTTR were significantly improved because of data accuracy