Contemporary Function-as-a-Service (FaaS) platforms leverage lightweight container-based virtualization to host instances of applications on-demand. This on-demand model is a good fit for computing environments where resource efficiency is desirable (such as the Cloud) or necessary (such as the Edge). However, the use of containers introduces significant delays to end-to-end service times in the form of cold start latency. As a workaround, FaaS platforms employ techniques like pre-warming containers or keeping idle containers alive to reduce the number of cold starts required. These techniques are effective, but greatly reduce platform resource efficiency, in turn making them less suitable for target computing deployments. Alternatives to containers like WebAssembly offer significantly lower cold start time, but have slower execution speeds than native code and cannot serve as a replacement to containers. We argue that combining the strengths of both containers and WebAssembly into a single hybrid runtime for FaaS platforms can improve both the end-to-end response time latencies and resource efficiency. We demonstrate this principle through proof-by-construction and introduce RUNE: a hybrid RUNtime for FaaS at the Edge. We evaluate RUNE empirically and show that its hybrid approach reduces tail latency response times at the 99th percentile by more than half while also improving memory utilization of the underlying platform by reducing the number of idle containers.

Container-based virtualization provides lightweight mechanisms for process isolation and resource control that are essential for maintaining a high degree of multi-tenancy in Function-as-a-Service (FaaS) platforms, where compute functions are instantiated on-demand and exist only as long as their exe-cution is active. This model is especially advantageous for Edge computing environments, where hardware resources are limited due to physical space constraints. Despite their many advantages, state-of-the-art container runtimes still suffer from startup delays of several hundred milliseconds. This delay adversely impacts user experience for existing human-in-the-loop applications and quickly erodes the low latency response times required by emerging machine-in-the-loop IoT and Edge computing applications utilizing FaaS. In turn, it causes developers of these applications to employ unsanctioned workarounds that artificially extend the lifetime of their functions, resulting in wasted platform resources. In this paper, we provide an exploration of the cause of this startup delay and insight on how container-based virtualization might be made more efficient for FaaS scenarios at the Edge. Our results show that a small number of container startup operations account for the majority of cold start time, that several of these operations have room for improvement, and that startup time is largely bound by the underlying operating system mechanisms that are the building blocks for containers. We draw on our detailed analysis to provide guidance toward developing a container runtime for Edge computing environments and demonstrate how making a few key improvements to the container creation process can lead to a 20 % reduction in cold start time.

BRIGHT-CS (Building Student Retention through Individuated Guided coHort Training in Computer Science) is a research and development project that 1) creates a computer science learning ecosystem for middle school Black girls and other girls of color and 2) researches the merits of the ecosystem in supporting persistence in CS to determine best practices for broadening participation to other marginalized student groups in computing. First, this paper describes the BRIGHT-CS program, from the structural, instructional, and curricular designs of the program to partnerships with local and community organizations that make up the ecosystem. Second, it presents the initial findings of research on the program and its impacts on student outcomes such as social-emotional attributes associated with persistence. The study employs a multi-method descriptive design. Data includes student surveys, interviews (from students, parents, instructors, teachers, and mentors), artifact reviews, and student observations. The study includes 46 students across four middle schools in two states. At the start of the program, 37% of the students reported being very interested in CS, and 72% reported being very confident in learning CS. This is much higher than a national benchmark of students. After four months of program implementation, the qualitative results show a more nuanced picture of the value of a learning ecosystem. First, the ecosystem offers implicit messaging about equity and success. Second, the ecosystem offers explicit messaging about personal challenges and improvement. Third, following the implicit and explicit messaging to students, students went from naïve confidence to authentic self-efficacy in CS.

Enabling next generation technologies such as self-driving cars or smart cities requires us to rethink the way we support their applications. The emergence of these technologies is fueled by the proliferation of a large number of devices in the Internet of Things. These devices have the potential to generate massive amounts of data, and applications supporting them often require this data to be processed in a timely manner. Because of these requirements, we must augment and extend the Cloud computing model to better serve such applications. The backhaul links connecting clients to Cloud data centers could quickly become overwhelmed by such data, and the physical distance of these data centers from clients prevents low-latency response times. To meet the challenges posed by emerging IoT applications, we must provide Cloudlike functionality closer to the edge of the network, where clients and their data live. We propose to elevate the Edge to be a peer of the Cloud for addressing these challenges.

UNLABELLED: Gemma is a database, analysis software system and web site for genomics data re-use and meta-analysis. Currently, Gemma contains analyzed data from over 3300 expression profiling studies, yielding hundreds of millions of differential expression results and coexpression patterns (correlated expression) for retrieval and visualization. With optional registration users can save their own data and securely share it with other users. Web services and integration with third-party resources further increase the scope of the tools, which include a Cytoscape plugin. AVAILABILITY: http://chibi.ubc.ca/Gemma, Apache 2.0 license.