Unified Holistic Memory Management Supporting Multiple Big Data Processing Frameworks over Hybrid Memories

Abstract

To process real-world datasets, modern data-parallel systems often require extremely large amounts of memory, which are both costly and energy-inefficient. Emerging non-volatile memory (NVM) technologies offer high capacity compared to DRAM and low energy compared to SSDs. Hence, NVMs have the potential to fundamentally change the dichotomy between DRAM and durable storage in Big Data processing. However, most Big Data applications are written in managed languages and executed on top of a managed runtime that already performs various dimensions of memory management. Supporting hybrid physical memories adds in a new dimension, creating unique challenges in data replacement. This paper proposes Panthera, a semantics-aware, fully automated memory management technique for Big Data processing over hybrid memories. Panthera analyzes user programs on a Big Data system to infer their coarse-grained access patterns, which are then passed to the Panthera runtime for efficient data placement and migration. For Big Data applications, the coarse-grained data division information is accurate enough to guide the GC for data layout, which hardly incurs overhead in data monitoring and moving. We implemented Panthera in OpenJDK and Apache Spark. Based on Big Data applications’ memory access pattern, we also implemented a new profiling-guided optimization strategy, which is transparent to applications. With this optimization, our extensive evaluation demonstrates that Panthera reduces energy by 32 – 53% at less than 1% time overhead on average. To show Panthera’s applicability, we extend it to QuickCached, a pure Java implementation of Memcached. Our evaluation results show that Panthera reduces energy by 28.7% at 5.2% time overhead on average.

Performance Evaluation Analysis of Spark Streaming Backpressure for Data-Intensive Pipelines

“…The authors in [20] mention that a Big Data system such as Spark is highly memoryintensive. This work reinforces that a lack of memory can lead to several functional and performance issues, including OOM crashes, a significantly degraded efficiency, or even a loss of data upon node failures.…”

“…In addition, the complexity behind the configuration of each approach is hard to manage. Furthermore, it is noticeable that data pipelines for SP could produce data faster than the downstream operators can consume, requiring large amounts of memory [20]. In such a case, backpressure [21] mechanisms have been widely adopted in the most varied domains of SP systems.…”

Challenges and future directions for energy, latency, and lifetime improvements in NVMs

“…So, researchers come up with different solutions to deal with these limitations, from redesigning the conventional data structures to proposing hardware-level methods, to deploy these new technologies in their systems. Among all the challenges that Energy efficiency [4,9,13,18,[23][24][25][26][27]35] [41, [65][66][67][68][69][70][71] [ [72][73][74][75][76][77][78] NVMs face, in this paper, we focus on (1) low endurance, high energy consumption, and asymmetric read/write related problems and (2) how researchers in different communities, from databases to storage systems to embedded systems and distributed systems, overcome these limitations. Table 3 classifies the research studies from Table 2 based on their memory technologies.…”