My research revisits fundamental design assumptions made in widely-deployed networks with a goal of improving their operational and deployment complexity. Ongoing projects include:

• Towards reliable and predictable Ethernet network stacks with Gazelle. This work revisits foundational design assumptions baked into end-host network stacks and their interactions with the network fabric with the lens of removing ambiguity around application errors that can be traced back to networks. Our work on How to diagnose nanosecond network latencies in rich end-host stacks was presented at NSDI 2022. It was showcased in the VMware OCTO blog Tackling Timely Problems that Affect Business Outcomes.
• Towards manageable networks with Netivus. This work revisits fundamental design assumptions baked into tcp/ip network models employed by network vendors with the lens of simplifying network deployment and operations for network owners. Our work on Understanding Lifecycle Management Complexity of Datacenter Topologies won the best paper award in NSDI 2019.


Service:
TPC: SIGCOMM 2021, HotNets 2020, SIGCOMM 2020, Usenix ATC 2020, SOSR 2020, ANCS 2018, SOSR 2018, SIGCOMM 2016
OC: SIGCOMM 2017

I can be reached at rniranjan at vmware.com