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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

In data centers, up to dozens of tasks are colocated on a single physicalmachine. Machines are used more efficiently, but tasks' performancedeteriorates, as colocated tasks compete for shared resources. As tasks areheterogeneous, the resulting performance dependencies are complex. In ourprevious work [18] we proposed a new combinatorial optimization model that usestwo parameters of a task - its size and its type - to characterize how a taskinfluences the performance of other tasks allocated to the same machine.  In this paper, we study the egalitarian optimization goal: maximizing theworst-off performance. This problem generalizes the classic makespanminimization on multiple processors (P||Cmax). We prove thatpolynomially-solvable variants of multiprocessor scheduling are NP-hard andhard to approximate when the number of types is not constant. For a constantnumber of types, we propose a PTAS, a fast approximation algorithm, and aseries of heuristics. We simulate the algorithms on instances derived from atrace of one of Google clusters. Algorithms aware of jobs' types lead to betterperformance compared with algorithms solving P||Cmax.  The notion of type enables us to model degeneration of performance caused byusing standard combinatorial optimization methods. Types add a layer ofadditional complexity. However, our results - approximation algorithms and goodaverage-case performance - show that types can be handled efficiently.

Optimizing egalitarian performance in the side-effects model of colocation for data center resource management