PhD candidate Ahmed Alquraan is one of 16 recipients internationally and the only recipient from Canada to receive a prestigious 2021 IBM PhD Fellowship. Established in 1951, these intensely competitive fellowships have supported thousands of PhD students over seven decades.
IBM PhD Fellowships are awarded for two years to doctoral candidates who have three years remaining in their graduate program. In addition to financial support for each of two years, fellowship recipients have the opportunity to intern with IBM. Though not mandatory, these internships are strongly encouraged as they improve the fellowship holder’s understanding of industrial research, broaden their range of technical contacts, and strengthen their technical experience.
Ahmed’s PhD research focuses on building distributed systems that leverage modern networking technology efficiently while tolerating complex network faults. He is supervised by Professor Samer Al-Kiswany, who leads the Waterloo Advanced Systems Lab.
“Network partitioning faults are among the hardest infrastructure faults to deal with,” Ahmed said. “To understand their impact, I conducted the first comprehensive study of system failures caused by networking partitioning in 25 production systems. An unexpected conclusion was that network partitions lead to catastrophic systems failure, yet they are easy to manifest. In fact, the majority of failures could be triggered reliably by isolating a single node. The most surprising finding was identifying a type of network fault known as a partial network partition — a kind of partition that disrupts communication between a subset of nodes of a computer cluster.”
This research was presented at the USENIX Symposium on Operating Systems Design and Implementation in 2018 and 2020. The 2018 paper has not only contributed to the curriculum of computer science graduate courses at four universities — University of Illinois at Urbana-Champaign, Harvard University, Johns Hopkins University and the University of Waterloo — but it also received coverage in the technical press and was awarded the 2019 Huawei Prize for Best Research Paper by a Mathematics Graduate Student.
Ahmed was a member of a team of Professor Al-Kiswany’s graduate students to develop a novel software-based approach, which they dubbed Nifty — an abbreviation for network partitioning fault-tolerance layer — to prevent these system failures. Nifty overcomes both the shortcomings of current fault tolerance approaches and effectively masks partial partitions, while imposing negligible overhead.
Ahmed’s other area of research is on network-accelerated systems, which led to a novel approach that significantly improves the storage efficiency and output speed of computer systems.
“Current data storage systems use only one storage server to process information, making them slow to retrieve information,” Professor Al-Kiswany said. “Ahmed was a key member of my team of graduate students who developed a new approach we called FLAIR, short for fast, linearizable, network-accelerated client reads. FLAIR optimizes data storage systems by using all the servers within a given network. When a user makes a data request, if the main server is full, another server automatically activates to fill it. We tested FLAIR using real workloads and it increases retrieval speeds from 35 to 97 percent.”
The other 15 doctoral candidates who received a 2021 IBM PhD Fellowship are from universities in Brazil, Japan, Israel, The Netherlands, South Africa, Switzerland and the United States.
The IBM PhD Fellowship Awards Program is an intensely competitive worldwide program, which honours exceptional PhD students who want to make their mark in promising and disruptive technologies. Focus areas include the following topics of particular interest — hybrid cloud, quantum computing and quantum systems, artificial intelligence, cloud and open source technologies, security and cyber security, data science, and systems.