Skip to main content

Asynchronous Replication

This article applies to: Shared File Services


Asynchronous replication is a method by which data is written to a second on-campus Data Center.

The goal of asynchronous replication is to ensure that critical data may be made available again from a secondary location, if the primary location becomes inaccessible. Data which is not replicated will be unavailable, and restoration may require substantially more time.

  • The PRIMARY site is normally in Rhodes Hall.
    • Data is accessible Read-Write from the PRIMARY location

    • Accessible via Windows Distributed File System (DFS)
      \\files.cornell.edu\OU\Share

    • Accessible via a DNS C-Name
      \\OU-Share.files.cornell.edu\OU-Share

  • The SECONDARY site is normally in the Computing and Communication Center (CCC).
    • Data is accessible Read-Only from the SECONDARY location

    • Accessible via a different DNS C-Name
      \\OU-Share-dr.files.cornell.edu\OU-Share-dr

  • Selection of the Replication option requires the selection of the snapshot option.

SFS shares with the optional Replication option have a Recovery Point Objective (RPO) of 4 hours, because replication by default will be performed asynchronously, at 4-hour intervals.  This means that up to the last 4 hours of updates to the share could be lost in the event that the replicated share is promoted to primary status during a failover after a disaster at the primary site.  In the event of a planned failover, the data is replicated first, so both sites have identical content.

In the event of an outage at the primary site (planned or emergency) CIT can make shares available read/write again from the secondary site initially via the DNS C-Names. The considerations for this promotion are described in outages affecting shared file services.

About this Article

Last updated: 

Thursday, April 18, 2019 - 9:23am

Was this page helpful?

Your feedback helps improve the site.

Comments?