English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In cloud storage systems with a large number of servers, files are typicallynot stored in single servers. Instead, they are split, replicated (to ensurereliability in case of server malfunction) and stored in different servers. Weanalyze the mean latency of such a split-and-replicate cloud storage systemunder general sub-exponential service time. We present a novel schedulingscheme that utilizes the load-balancing policy of the \textit{power of $d$$(\geq 2)$} choices. An alternative to split-and-replicate is to useerasure-codes, and recently, it has been observed that they can reduce latencyin data access (see \cite{longbo_delay} for details). We argue that under highredundancy (integer redundancy factor strictly greater than or equal to 2)regime, the mean latency of a coded system is upper bounded by that of asplit-and-replicate system (with same replication factor) and the gap betweenthese two is small. We validate this claim numerically under different servicedistributions such as exponential, shift plus exponential and the heavy-tailedWeibull distribution and compare the mean latency to that of anunsplit-replicated system. We observe that the coded system outperforms theunsplit-replication system by at least $20\%$. Furthermore, we consider themean latency for an erasure coded system with low redundancy (fractionalredundancy factor between 1 and 2), a scenario which is more pragmatic, giventhe storage constraints (\cite{rashmi_thesis}). However under this regime, werestrict ourselves to the special case of exponential service time distributionand use the randomized load balancing policy namely \textit{batch-sampling}. Weobtain an upper bound on mean delay that depends on the order statistics of thequeue lengths, which, we further smooth out via a discrete to continuousapproximation.

Faster Data-access in Large-scale Systems: Network-scale Latency Analysis under General Service-time Distributions