• Article

Backup and Restoration Process Overview

 

Topic Last Modified: 2011-05-12

This section provides an overview of how the backup and restoration process works for Microsoft Lync Server 2010. You use the same process for all Standard Edition servers and Enterprise Edition servers, regardless of their location.

In general, backup works as follows:

  • You create a backup location as a shared folder on a stand-alone computer that is not part of any pool. The location of the backup is referenced in $Backup.

  • On a regular, scheduled basis, you back up all the Lync Server databases and all the file stores that are described in Backup and Restoration Requirements: Data by following the procedures described in this document. The Central Management store includes all the server settings and configurations.

  • Each time you run a subsequent backup, you create a new shared folder and change the path that $Backup references.

In general, restoration works as follows:

  • When a failure or outage occurs, you restore the data in the location referenced by $Backup to new or clean computers by following the procedures described in this document.

    Important

    This restoration process does not restore data onto an existing server state. That is, this process requires that the server is clean or new.

  • Lync Server 2010 supports only the Simple Recovery model for SQL Server databases. With the Simple Recovery model, databases are recovered to the point of last full backup, which means that you cannot restore a database to the point of failure or to a specific point in time.

    Note

    The Simple Recovery model is optimal, because the Lync Server 2010 SQL Server back-end (Rtc.mdf) database is actually smaller than the transaction log files, and is significantly smaller than those of typical line-of-business (LOB) database applications.

  • All Domain Name System (DNS) configuration, Dynamic Host Configuration Protocol (DHCP) configuration, domain names, computer fully qualified domain names (FQDNs), file store paths, and so on must be the same at the time of restoration that they were at the time of back up.

Restoration can be of any of the following varieties:

  • A Standard Edition server fails.

  • The Central Management Server fails.

  • An Enterprise Edition Back End Server fails.

  • A server that is a member of a Lync Server pool fails.

  • A Lync Server pool fails.

    Note

    Lync Server pools include Front End, Director, Mediation, A/V Conferencing, Archving, Monitoring, and Edge pools.

  • A File Store fails.

  • An Archiving Server or a Monitoring Server with a collocated database fails.

  • A stand-alone Archiving database or Monitoring database fails.

If a server running Lync Server 2010 fails, recovery includes the following steps:

  • Install the operating system on a new or clean computer with the same FQDN as the failed computer.

  • Reinstall certificates.

  • If the server hosted a database, install SQL Server 2008 R2, SQL Server 2008, or SQL Server 2005.

  • In general, if the server hosted a database, run Topology Builder to create and install the database and set up access control lists (ACLs).

  • In general, if the server hosted a server role, run Step 1 through Step 4 of the Lync Server 2010 Deployment Wizard to install the local configuration files, install the server role components, assign certificates, and start the services.

    Note

    If the server hosted a database collocated with the server role, running Step 2 of the Lync Server Deployment Wizard recreates the database.

  • If the server hosted a database, restore backed up data.