Project Ideas

I realize each of you are in different stages of your graduate career, and are looking for different things out of a course project. While all of the ideas below likely could result in a suitable class project, they vary in their ambitiousness and scope. Those of you with aspirations of possibly publishing your work might think about tackling something more open-ended.

Some of the projects require access to resources in my research group. Plese contact me if you're interesting in persuing one of them so I can make sure we have enough resources to go around.

• Add group membership and proper view-change functionality to our Distributed Rate Limiting system to gracefully handle limiter failure and network partitioning.
• Implement and extend the channel assignment procedure from our SoftSpeak system. In particular, gracefully handle situations where VoIP clients dynamically arrive and depart at neighboring access points, yet cannot hear each other directly.
• Extend our Address Hiding Protocol system to support multiple, coordinated proxies. In particular, communicate about the used addresses and key rotation schedules in a synchronized fashion.
• Instrument the Hadoop scheduler to determine the bottlenecks in Map/Reduce jobs. In particular, identify jobs where CPU, disk, or network bandwidth are limiting factors, and implement extensions to the scheduler to improve performance in these problematic cases.
• Implement simple policies for saving power in our cluster. Most of the machines are idle, turn off during idle periods. Could use OS hibernation, or, since we're using VMs, Xen save and restore.
• Add gossip-based reconciliation to LBFS. LBFS is almost exactly what you want, except that it has a strict client-server relationship. Consider a Bayou-like approach.
• Suppose you mount a NFS volume remotely and do computation on the data. At a certain point it becomes clear that you're doing a lot of computation on it and data shipping is the bottleneck. Then checkpoint the VM and ship it to where the data is stored to compute on it locally (need to translate the NFS mount into a local mount...). when done, ship the VM back.
• The XL routing algorithm provides a glimpse at the interplay between network structure and the communication complexity of routing. However we do not yet have a deep understanding of this connection. We know that in realistic networks with short detours it is possible to suppress some updates, and in cases of the same link failing repeatedly, it is possible to do asymptotically better than the current approach of complete flooding. Beyond these special cases, however, the question of which networks require complete flooding is open.
• Extend any of the papers we've read to fix problems they left danging.