National systems

From Alliance Doc
Revision as of 19:24, 20 October 2016 by Pjmann (talk | contribs) (→‎Compute)
Jump to navigation Jump to search
Sharing: PublicEarly DraftAuthor:CC Migration team

Introduction

TBD - detailed docs for the new systems are being written by the Research Support National Team.

Compute

NameDescriptionApproximate CapacityAvailability
CC-Cloud Resources (GP1)cloud7,640 coresIn production (integrated with west.cloud)
GP2general-purpose cluster (serial and small parallel jobs)25,000 coresFebruary, 2017
GP3general-purpose cluster (serial and small parallel jobs)25,000 coresMay, 2017
LPa cluster designed for large parallel jobs60,000 coresSummer, 2017
Available software

National Storage Infrastructure

The National Storage Infrastructure will have 2 components:

Object Storage

A fully distributed object storage system spanning all 4 sites, with geo-redundancy, replication and universal access.

Long-term File Storage

"Nearline"
Long-term file storage sites (similar to Silo) at SFU and the University of Waterloo. Each site will have tape backup.

  • Patrick Mann (talk) 16:43, 20 October 2016 (UTC) Storage Building Blocks have been delivered and are being installed.
  • Patrick Mann (talk) 15:26, 23 September 2016 (UTC) Note that due to Silo decommissioning Dec.31/2016 it will be necessary to provide interim storage while the NSI is developed. See Migration2016:Silo for details.