The case for RAMClouds: scalable high-performance storage entirely in DRAM

By John K. Ousterhout, Parag Agrawal, David Erickson, Christos Kozyrakis, Jacob Leverich, David Mazières, Subhasish G. Mitra, Aravind Narayanan, G. Parulkar, Mendel Rosenblum and 3 more. OPSR, 2010

566 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Hardware and Architecture, Information Systems

View Paper: https://www.semanticscholar.org/paper/3dce3f430f042b20abc9e4677461842b67a1fb18

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Hardware and Architecture

In Information Systems

Citations in the corpus

In-Memory Big Data Management and Processing: A Survey (2015, methodology): “To eliminate the potential I/O bottleneck caused by logging, group commit and log coalescing [37], [127], and remote logging [2], [40] are adopted to optimize the logging efficiency.”
Resilient Distributed Datasets: A Fault-Tolerant Abstraction for In-Memory Cluster Computing (2012, methodology): “For these applications, it is more efficient to use systems that perform traditional update logging and data checkpointing, such as databases, RAMCloud [25], Percolator [26] and Piccolo [27].”
Consistent and Durable Data Structures for Non-Volatile Byte-Addressable Memory (2011, methodology): “As DRAM is volatile, these systems provide data durability using backend databases (e.g., memcached/MySQL ), on-disk logs (e.g., RAMCloud), or, for systems with relaxed durability semantics, via periodic checkpoints.”
Trinity: a distributed graph engine on a memory cloud (2013, background): “RAMCloud is a data storage system that hosts all its data permanently in memory.”
Piccolo: Building Fast, Distributed Programs with Partitioned Tables (2010, methodology): “At table creation time, the master assigns table partitions to all workers using a simple round-robin assign-5 ment for empty memory tables.”
HyPer: A hybrid OLTP&OLAP main memory database system based on virtual memory snapshots (2011, background): “[3] who proclaim the so-called RAMcloud as a main-memory storage device for the largest Internet software applications.”
DeTail: reducing the flow completion time tail in datacenter networks (2012, methodology): “For example, a page load at Facebook consists of an average of 130 data requests [28].”
CACTI 7 (2017, methodology): “There are many example platforms and workloads that rely on processing of large in-memory datasets, such as SAP HANA [SAP 2013], SAS in-memory analytics [SAS 2013], RAMCloud [Ousterhout et al. 2009], SPARK [Zaharia et al. 2010], and memory blades [Lim et al. 2009].”
PACMan: Coordinated Memory Caching for Parallel Jobs (2012, methodology): “RAMCloud [20] and prior work on databases such as MMDB [17] propose storing all data in RAM.”