May 2019 – Mar 2020 · 10 mos

• Technical Lead Engineer for the product called Big Cloud Fabric which enables virtual private cloud (VPC) based logical networking, delivering network automation and visibility to both on-premises enterprise clouds and multiple public clouds. I have been involved with the design and development of features for the SDN controller that manages a data center network. The application development for the controller involves Java for the backend and python for CLI development.
• Fabric Management
• Discovery and management of the life cycle of physical switches, virtual switches, interfaces and interface groups
• LLDP or CDP based host discovery
• LAG management
• Fabric Upgrade
• Network Monitoring and Analytics
• I was responsible for conceptualizing and delivering the integration of Big Switch's Analytics Node with the Big Cloud Fabric product. https://www.bigswitch.com/big-mon-analytics-node
• Co-ordinated collaboration between various teams for the development of an analytics agent in the switches and orchestration of the agents by the SDN controller, to provide analytics for the traffic in the fabric, to help customers visualize and troubleshoot their network.
• Responsible for features such as generation of sflow, spanning traffic from the fabric, endpoint discovery using ARP, Operating system fingerprinting using DHCP, DNS analytics to name a few.
• Big Cloud Fabric for Public Cloud:
• Provides the ability for networking teams to manage and analyze VPCs (Virtual Private Clouds) deployed in AWS.
• My blogs about the launch of the product:
• https://www.linkedin.com/pulse/big-cloud-fabric-aws-onkar-bhat/
• https://www.bigswitch.com/blog/2019/07/17/big-cloud-fabric-for-aws
• Deployed in the AWS marketplace - https://aws.amazon.com/marketplace/pp/B07TK21Z79

May 2017 – May 2019 · 2 yrs

May 2017 – May 2017 · 0 mo

Apr 2015 – May 2017 · 2 yrs 1 mo

• Disk to Cloud backup
• I joined a new team within the AltaVault group that was tasked with the job of integrating with the other products at NetApp based on the SnapMirror protocol. My experience with SnapMirror in the Disk to Disk backup team was useful for helping this team. This team grew to a size of 5 engineers and a technical leader. We wrote a new module from scratch that represents a SnapMirror endpoint in AltaVault.
• We delivered the following features:
• Backup of volume snapshots from disk to cloud.
• Restore of volume snapshots from the cloud to disk.
• Restore specific files form a specific snapshot from the cloud to disk.
• rocksdb for storing the metadata for snapshots.
• c++ poco for managing network connections.
• Cloud connectors for data backup to S3/ Azure blob storage / Google cloud storage etc.
• I worked on enhancing the AWS S3 storage cloud connector of the AltaVault appliance to support the use of the STANDARD INFREQUENT ACCESS tier that helps reduce the cost of storing data significantly in exchange for reduced SLAs compared to the STANDARD tier of S3.

Jun 2012 – Jun 2015 · 3 yrs

• Software engineer in the transfer engine team which is responsible for the design and development of the data path of the SnapMirror and SnapVault products.
• SnapVault(Disk to Disk Backup):
• http://www.netapp.com/us/products/protection-software/snapvault.aspx
• SnapMirror(Data replication and Disaster Recovery):
• http://www.netapp.com/us/products/protection-software/snapmirror.aspx
• Languages : C++ and C.
• Worked on two types of replication protocols:
• Logical Replication with Storage Efficiency (LRSE)
• Block Replication Engine (BRE)
• WAFL filesystem related work:
• Optimization of reads by batching contiguous blocks using the
• physical volume block numbers, and RAID topology maps.
• Design and development of buftree agnostic APIs for finding the differences between two
• filesystem snapshots.
• Replication Service Level Objectives:
• Worked on the infrastructure for assigning service level objectives for replication workloads
• and using the objectives for prioritizing the replication workloads.
• Patent issued by the United States Patent and Trademark Office - http://www.google.com/patents/US20160026535
• Pull based model for replication:
• Worked on a model of replication where the destination asks for the data that is missing at
• the destination. This is in comparison to the other more mature model where the source finds
• the differences between two snapshots, and then sends the differences to the destination.
• New filesystem semantics , required a re design of the pull based modules at the
• destination end point of replication. I designed a solution that would allows this model to be used
• for replication based on the new filesystem changes.

Nov 2011 – May 2012 · 6 mos

• Graduate Teaching Assistant for the course titled "Linux and Open Source"

Aug 2010 – May 2012 · 1 yr 9 mos

• Master of Science in Information Networking at the Information Networking Institute(INI)

May 2011 – Aug 2011 · 3 mos

• Distributed Stack project
• Primary developer for the project, involving making IOS applications like BGP, FTP and TTCP run over the Linux TCP/IP stack instead of the classic Cisco IOS TCP/IP stack.
• Created new infrastructure that mapped IOS transport API's onto Linux sockets. Preserved IOS API behavior and scheduling semantics, so that application code is unchanged.
• Allowed coexistence of the IOS and Linux TCP/IP stacks, so that applications may choose when, and if, to migrate to Linux TCP/IP. Carried out performance benchmarking with promising preliminary results.

Aug 2008 – Aug 2010 · 2 yrs

• Developed and tested the Network Based Application Recognition (NBAR) SDK, using C, which is used for enabling a customer to add a custom protocol to the NBAR engine to enable the recognition of that protocol on Cisco Cat6k switches.
• Implemented parsers, using C, for protocols like Oracle,CIFS and TIBCO in order to parse packets captured from the network, for network monitoring as part of the Application Visibility Monitoring (AVM) project.

Feb 2008 – Jul 2008 · 5 mos

• Project titled "Tool Tracker : A toolkit ensembling useful online networking tools for efficient management and operation of a network"
• Published by the World Academy of Science Engineering and Technology in the 42nd volume of their journal - ( http://www.waset.org/journals/waset/v42/v42-126.pdf )

Jul 2006 – Aug 2006 · 1 mo

• Implementation of a secure communication system using chaotic cryptography.
• Implementation of simple and mutual chaotic synchronization between two communicating entitites
• Implementation of an attack on this system to decrypt the super key used by the communicating entities