Cloud services have recently started undergoing a major shift from monolithic applications, to graphs of hundreds of loosely-coupled microservices. Microservices fundamentally change a lot of assumptions current cloud systems are designed with, and present both opportunities and challenges when optimizing for quality of service (QoS) and utilization. In this paper we explore the implications microservices have across the cloud system stack. We first present Death-StarBench, a novel, open-source benchmark suite built with microservices that is representative of large end-to-end services, modular and extensible. DeathStarBench includes a

Performance unpredictability is a major roadblock towards cloud adoption, and has performance, cost, and revenue ramifications. Predictable performance is even more critical as cloud services transition from monolithic designs to microservices. Detecting QoS violations after they occur in systems with microservices results in long recovery times, as hotspots propagate and amplify across dependent services.We present Seer, an online cloud performance debugging system that leverages deep learning and the massive amount of tracing data cloud systems collect to learn spatial and temporal patterns that translate to QoS violations. Seer combines lightweight distributed RPC-level tracing, with detailed low-level hardware monitoring to signal an upcoming QoS violation, and diagnose the source of unpredictable performance. Once an imminent QoS violation is detected, Seer notifies the cluster manager to take action to avoid performance degradation altogether. We evaluate Seer both in local clusters, and in large-scale deployments of end-to-end applications built with microservices with hundreds of users. We show that Seer correctly anticipates QoS violations 91% of the time, and avoids the QoS violation to begin with in 84% of cases. Finally, we show that Seer can identify applicationlevel design bugs, and provide insights on how to better architect microservices to achieve predictable performance.

No abstract

Performance unpredictability is a major roadblock towards cloud adoption, and has performance, cost, and revenue ramifications. Predictable performance is even more critical as cloud services transition from monolithic designs to microservices. Detecting QoS violations after they occur in systems with microservices results in long recovery times, as hotspots propagate and amplify across dependent services. We present Seer, an online cloud performance debugging system that leverages deep learning and the massive amount of tracing data cloud systems collect to learn spatial and temporal patterns that translate to QoS violations. Seer combines lightweight distributed RPC-level tracing, with detailed low-level hardware monitoring to signal an upcoming QoS violation, and diagnose the source of unpredictable performance. Once an imminent QoS violation is detected, Seer notifies the cluster manager to take action to avoid performance degradation altogether. We evaluate Seer both in local clusters, and in large-scale deployments of end-to-end applications built with microservices with hundreds of users. We show that Seer correctly anticipates QoS violations 91% of the time, and avoids the QoS violation to begin with in 84% of cases. Finally, we show that Seer can identify applicationlevel design bugs, and provide insights on how to better architect microservices to achieve predictable performance.

In this work we present a new set of error correcting codes -Multi-Use Error Correcting Codes (MUSE ECC) -that have the ability to match reliability guarantees of all commodity, conventional state-of-the-art ECC with fewer bits of storage. MUSE ECC derives its power by building on arithmetic coding methods (first used in an experimental system in 1960s). We show that our MUSE construction can be used as a "drop in" replacement within error correction frameworks used widely today. Further, we show how MUSE is a promising fit for emerging technologies such as a DDR5 memories. Concretely, all instantiations of MUSE we show in this paper offer 100% Single Error Correction, and multi-bit error detection between 70% and 95% while using fewer check bits. MUSE ECC corrects failure of a single chip on a DIMM with check bit space savings of 12.5% compared to conventional techniques. The performance overheads, if any, are negligible. Our results open the possibility of reusing ECC storage for things beyond reliability without compromising reliability, thus solving a 40-year-old puzzle.