SDH Network Monitoring
A Case Study on topology-driven network event correlation for root cause analysis in a multi-vendor SDH network
Operational Requirements
Customer’s network consisted of several multi-vendor EMS instances to monitor SDH network in different regions of country. Locating the root cause of a fault required scanning through EMS screens and manual correlation. Customer wanted to reduce the fault location time from hours to almost real time
Approach
- Analysis of fault events from multi-vendor EMS and present mode of fault management operations
- Analysis of network topology and its representation in inventory systems
- Analysis of various SDH fault scenarios
Solution Delivered
- Normalization of alarm objects at NMS level across multi-vendor EMS that sent alarms over TCP/IP, SNMP or CORBA
- Single SDH network alarm view for the whole multi-vendor network
- Automation of inventory interface for topology information
- Development of correlation rules to pinpoint root cause alarm and suppress redundant alarms
- End-to-end testing to demonstrate successful functioning of rules
Value Proposition
- Accurate modeling of rules due to domain knowledge
- Hundreds of alarms from the network due to a fiber cut were suppressed within 30 seconds to display the root cause problem
Technologies
- SDH, PSTN, SNMP (v1 and v2c), CORBA TMF814, TCP/IP
- HP-UX, Solaris, MS-Windows 2000
- Shell, Perl
- Navis NFM, VitalSuite NFM, DNA, TNMS, T2000, WS-NMS, OMS
