MemC3: Compact and Concurrent MemCache with Dumber Caching and Smarter Hashing

By Bin Fan, D. Andersen, Michael Kaminsky. Symposium on Networked Systems Design and Implementation, 2013

425 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/a73ce2aab2b4501bdffa3cd66a1b5045b9a452f4

How researchers see this paper

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

1. Anuj Kalia
2. Michael Kaminsky
3. D. Andersen
4. Alex D. Breslow
5. Xin Jin
6. Bin Fan
7. Ion Stoica
8. Hao Zhang

Citations in the corpus

• FaRM: Fast Remote Memory (2014, unknown): “For example, the results in [16] show a state-of-the-art key-value store performing 7x worse in a client-server setup using TCP/IP than in a single-machine setup despite extensive request batching.”
• The Design and Implementation of Open vSwitch (2015, unknown): “Drawing inspiration from CuckooSwitch [42], we adopted optimistic concurrent cuckoo hashing [6] and RCU [23] techniques to implement nonblocking multiple-reader, single-writer flow tables.”
• WiscKey (2016, methodology): “Many proposed techniques seek to overcome the scalability bottlenecks of in-memory key-value stores, such as Masstree [Mao et al. 2012], MemC3 [Fan et al. 2013], Mem-cache [Nishtala et al. 2013], MICA [Lim et al. 2014], and cLSM [Golan-Gueta et al. 2015].”
• Horton Tables: Fast Hash Tables for In-Memory Data-Intensive Computing (2016, methodology): “…although BCHTs are higher-performing than other open addressing schemes [60, 64], we find that as the performance of modern computing systems becomes increasingly constrained by memory bandwidth [1, 11, 52, 70], they too suffer from a number of inefficiencies that originate from how data is…”
• Cuckoo Filter: Practically Better Than Bloom (2014, methodology): “Our previous study in building high-speed and memory-e ffi cient key-value stores [17, 11] and software-based Ethernet switches [26] all applied cuckoo hash tables as internal data structures.”
• IX: A Protected Dataplane Operating System for High Throughput and Low Latency (2014, background): “We did yet not attempt to tune the internal scalability of memcached [20] or to support zero-copy I/O operations.”
• NetCache: Balancing Key-Value Stores with Fast In-Network Caching (2017, background): “Given the high-throughput and low-latency requirements of large-scale Internet services, key-value storage systems are shifting to in-memory designs [1, 15, 16, 22, 23, 26, 27, 29, 30, 34, 36, 38, 42, 43].”
• In-Memory Big Data Management and Processing: A Survey (2015, methodology): “, Memcached [61], MemC3 [63], TxCache [267], HashCache [268]), etc.”
• Using RDMA efficiently for key-value services (2014, methodology): “Cuckoo hash tables are an attractive choice for building fast key-value systems [9, 31, 17] because, with K hash functions (usually, K is 2 or 3), they require only K memory lookups for GET operations, plus an additional pointer dereference if the values are not stored in the table itself.”