Network monitoring and Analytics.

Network monitoring:

• SME for Network data collection using SNMP/System logging/API polling and post processing the metrics and pushing into a time-series database such as Prometheus.

• Set up alerting based on network metrics

Network Analytics:

• Setup data streaming pipelines for correlating network events using Apache Flink. Engineering network event correlation algorithms to reduce the signal to noise ratio, anomaly detection and analyze the root cause of large scale network events.

• Combined Alerting and correlation of large scale events to implement auto remediation and self healing network use-cases.

ACI (SDN) Controller Team (Insieme BU)

• Working on policy design and implementation for Cisco SDN Controller (Application policy Infrastructure Control)

https://www.cisco.com/c/en/us/products/switches/data-center-aci-anywhere.html#/

Insieme - Intent based networking

● Increased the scale of routing stack in the SDN fabric by 4x to 128k by adding optimizations in thread processing, improved flow controlled programming, making processes lightweight with optimized memory management and API design.

● Cross domain routing with high domain scale: designed and developed a feature which provides separation of content publishers and subscribers across 300 different domains (departments); targeted for financial and stock trading customers. Also developed the data models for controller integration.

● Boundary ACL: Designed the controller model and implemented the backend to restrict user subscription and content distribution in a data center fabric across multiple geographical areas.

Insieme - Core Software Group

● IPV4 and IPV6 inception for multicast in ACI SDN fabric: Responsible for complete distributed routing stack design and development for IPV4 and IPV6 modules from controller models to route programming; also worked on process bring-up and integration across multiple teams.

● RP for multi-site SDN fabrics: Designed and developed a feature which provides flexibility in placement of Rendezvous points in a multicast network spanning across different geographical locations in order to reduce latency in source discovery.

● Designed controller models for RESTful interaction between controller and switch for configurations and exported object models to be queried by network insights.

● Developed a python based CLI parser and querier to retrieve the operational states of a routing protocol which serve as an interface for network analytics.

(Network Operating Systems)

● Resilient RPF: Worked with the platform team to develop configurations to handle quick updates to load balance updates occurring at a single node.

● XML modeling and backend for run-time state attributes.

● L3 IGMP and L2 IGMP snooping enhancements with distributed switch model.

● Implemented a way to safely clear states in software and push down the updates to the platform by introducing appropriate locking mechanisms and ACK handling.