Bhavana V S

I am interested in the fields of Computer Networks, Distributed Systems, and Operating Systems. I am pursuing my PhD, advised by Prof. Srinivas Narayana and Prof. Badri Nath, in the Department of Computer Science at Rutgers University. My current work focuses on developing techniques to accurately measure client-visible application response latency from a middle box, such as a load balancer. Additionally, I explore how middleboxes can use these measurements to enhance feedback control, thereby improving the overall performance of distributed systems. A significant aspect of my work is implemented in the eBPF framework, as it offers inherent security and observability across various layers of the stack.

Also, I am involved in developing online streaming algorithms for network applications, including intrusion detection systems and load balancers deployed in middleboxes. I am collaborating with Prof. Shir Landau Feibish and Jonathan Diamant from the University of Haifa on an ongoing research project to identify the heaviest intervals within a given data stream.

Previously, I gained experience working with Citrix Systems in Bangalore, India, and I completed my M. Tech. Degree at the Indian Institute of Technology Guwahati, under the mentorship of Prof. Pradip K. Das. I did my B. Tech from Marian Engineering College, Kerala.

Publications

Preprint

Measuring Round-Trip Response Latencies Under Asymmetric Routing

Bhavana Vannarth Shobhana, Yen-lin Chien, Jonathan Diamant, and 3 more authors

2025

arXiv
HotNets ’22

Load Balancers Need In-Band Feedback Control

Bhavana Vannarth Shobhana, Srinivas Narayana Ganapathy, and Badri Nath

2022

Abs PDF

Load balancers (LBs) are critical components of interactive services today, distributing client requests over a server pool to improve performance and availability. LBs enable application logic to scale out to a pool of replicated servers, improving application performance by avoiding hot spots. LBs may run as frontends, routing client requests arriving from the Internet to the server pool. LBs may also run as tier-to-tier load balancers, scaling out a single application tier (e.g., an in-memory database) of a complex application, routing requests sent from other tiers. LBs may run at layer-4 (using connection 4-tuples) or layer-7 (e.g., using HTTP-based service identifiers) to map requests to servers. There are a wide variety of solutions available for load balancing. Majority of these solutions depends on measurements - derived via visibility of request and response on LB or out-of band server metrics or logs collected from the server, for making the decisions as where to forward client request. A key challenge for LBs when they only process requests and not responses is that it cannot directly measure server performance. We propose an in-band feedback control operating purely locally at LBs to adapt request-routing to server performance. We present an initial design of an LB that adapts to a server latency inflation of 1 ms and reduces tail latencies in milliseconds, while only observing client-to-server traffic