1 / 5

Large-Scale Time-Sharing Workqueue Management System

A system built from an array of computers managing job scheduling, resource allocation, status reporting, and result collection. Utilizes distributed technologies like Condor. Machines serving GFS cells also contribute to workqueue tasks. System Health Infrastructure collects server data for analysis.

eddiebarber
Download Presentation

Large-Scale Time-Sharing Workqueue Management System

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Other Google Technologies

  2. Workqueue: Scheduling many jobs on many machines • A large scale time sharing system built out of an array of computers and their ram, cpus and disks. • Schedules jobs, allocates resources, reports status, and collects the results. • Similar to other distributed systems described in the literature, such as Condor.

  3. Workqueue: Scheduling many jobs on many machines • The same machines that serve as chunkservers for a GFS cell can also contribute CPU and memory resources to a workqueue because the computational requirements of a GFS storage cell are negligible. • So a machine A may contribute to storage cluster (GFS cell) B and also to compute cluster (workqueue) C -- the compute and storage cells can (and will) overlap.

  4. Dogfood: The System Health Infrastructure Talks to every server frequently Collects Health signals Activity information Configuration data Stores time-series data forever Enable highly parallel analysis of repository data Ref: System Health Work, by Wolf Weber, Xiabo Fan, Don Yang, Greg Slaughter, Eduardo Pinheiro and others.

  5. Next Topic... Amazon

More Related