Rohan Kadekodi

Rohan Kadekodi

Systems and storage researcher

About Me

I am a PhD student at University of Texas - Austin. My research advisor is Prof. Vijay Chidambaram. My interests are in systems and storage. I have worked on Persistent Memory, file systems and key-value stores. I am a part of the UT Systems and Storage Lab and the LASR research group.


  • SplitFS. File system for Persistent Memory which reduces software overhead. SOSP 2019.

  • PebblesDB. Key-value store that reduces write amplification and increases throughput compared to LevelDB and RocksDB.

Ongoing work

  • Designing persistent memory systems from the point of view of performance and scalability

  • Analyzing IO amplification in file systems


SplitFS: Reducing Software Overhead in File Systems for Persistent Memory

Rohan Kadekodi, Se Kwon Lee, Sanidhya Kashyap, Taesoo Kim, Aasheesh Kolli, Vijay Chidambaram
SOSP 2019

Analyzing IO Amplification in Linux File Systems

Jayashree Mohan, Rohan Kadekodi, Vijay Chidambaram
Best Poster Award!

PebblesDB: Building Key-Value Stores using Fragmented Log-Structured Merge Trees

Pandian Raju, Rohan Kadekodi, Vijay Chidambaram, Ittai Abraham
SOSP 2017


Sequentality Matters on Persistent Memory

Analysis project to measure the effect of prefetching and TLB to argue that sequentiality is king, even for Persistent Memory, and random access patterns suffer greatly, as opposed to sequential access patterns.

Content Based Storage on Btrfs

Introducing a new mechanism to store and access files in Btrfs, based on the hash of their content, rather than based on their path. This also implies that inherent deduplication of files can be done, if they have the same content hash, which would otherwise be impossible with path based storage of files.