∫ntegrabℓε ∂iﬀerentiαℓs · unorganised notes, code, and writings of random topics

Kamp (2010) You're doing it wrong (CACM)

July 17, 2010 paper

This is written by a FreeBSD kernel developer. It discussed the effect of OS-level caching to algorithm performance. The story is that, one has to do store/retrieve operation on a heap. If the heap is made of a balanced binary tree, usually it is implemented as an array such that... [more]

Clévenot et al (2005) Stochastic fluid models for cache clusters (PE)

June 25, 2010 paper

This paper models the hit rate of a cluster of cache servers. [more]

Abu-Libdeh et al (2010) RACS: A Case for Cloud Storage Diversity (SoCC)

June 22, 2010 paper

This paper proposed to use multiple cloud services for storage in a RAID-like style, so that the user can prevent vendor lock-in, and provide a higher redundancy. Software is created and published. Performance evaluation is presented. [more]

Costa et al (2009) Why Should We Integrate Services, Servers, and Networking in a Data Center? (WREN)

June 22, 2010 paper

This paper challenges the idea of using, say, fat-tree in DCN because in a DCN, network bisection bandwidth and low diameter are just part of the criteria. We still have to consider resilience to failure, wiring complexity, and energy consumption. [more]

Vigfusson et al (2010) Dr. Multicast: Rx for Data Center Communication Scalability (EuroSys)

June 22, 2010 paper

IP Multicast is not used in data center but it should be beneficial. This is because of several reasons: [more]