• Accueil
  • Accueil
  • Accueil
  • Accueil

CNRS

Rechercher




Accueil >

[PDS/HPDA Seminar] 18/2/2022 from 10:00 to 11:30 at 4A312 - Jana Ismail (reading group) and Timothée Zerbib (reading group)

[PDS/HPDA Seminar] 18/2/2022 from 10:00 to 11:30 at 4A312 - Jana Ismail (reading group) and Timothée Zerbib (reading group)

Rabia : Simplifying State-Machine Replication Through Randomization

Reading group : Jana Ismail will present "Rabia : Simplifying State-Machine Replication Through Randomization" (SOSP’21) at 4A312 (visio link) the 18/2/2022 at 10h00.
Abstract

We introduce Rabia, a simple and high performance framework for implementing state-machine replication (SMR) within a datacenter. The main innovation of Rabia is in using randomization to simplify the design. Rabia provides the following two features : (i) It does not need any fail-over protocol and supports trivial auxiliary protocols like log compaction, snapshotting, and reconfiguration, components that are often considered the most challenging when developing SMR systems ; and (ii) It provides high performance, up to 1.5x higher throughput than the closest competitor (i.e., EPaxos) in a favorable setup (same availability zone with three replicas) and is comparable with a larger number of replicas or when deployed in multiple availability zones.


Corey : An operating system for many cores

Reading group : Timothée Zerbib will present "Corey : An operating system for many cores" (OSDI’08) at 4A312 (visio link) the 18/2/2022 at 10h30.
Abstract

Multiprocessor application performance can be limited by the operating system when the application uses the operating system frequently and the operating system services use data structures shared and modified by multiple processing cores. If the application does not need the sharing, then the operating system will become an unnecessary bottleneck to the application’s performance. This paper argues that applications should control sharing : the kernel should arrange each data structure so that only a single processor need update it, unless directed otherwise by the application. Guided by this design principle, this paper proposes three operating system abstractions (address ranges, kernel cores, and shares) that allow applications to control inter-core sharing and to take advantage of the likely abundance of cores by dedicating cores to specific operating system functions. Measurements of microbenchmarks on the Corey prototype operating system, which embodies the new abstractions, show how control over sharing can improve performance. Application benchmarks, using MapReduce and a Web server, show that the improvements can be significant for overall performance : MapReduce on Corey performs 25% faster than on Linux when using 16 cores. Hardware event counters confirm that these improvements are due to avoiding operations that are expensive on multicore machines.