Skip to content

Latest commit

 

History

History
26 lines (19 loc) · 2.38 KB

mssql-disaster-recovery.adoc

File metadata and controls

26 lines (19 loc) · 2.38 KB
Raw
sidebar permalink keywords summary
sidebar
mssql/mssql-disaster-recovery.html
MSSQL,SQL Server, disaster recovery, snapmirror
Microsoft SQL Server disaster recovery with ONTAP

Microsoft SQL Server disaster recovery with ONTAP

Enterprise databases and application infrastructures often require replication to protect from natural disaster or unexpected business disruption with minimal downtime.

The SQL Server Always-On availability group replication feature can be an excellent option, and NetApp offers optiosn to integrate data protection with Always-On. In some cases, however, you might want to consider ONTAP replication technology. ONTAP replication options, including MetroCluster and SnapMirror, can scale better with minimal performance impact, protect non-SQL data, and generally provide a full-infrastructure replication and DR solution.

SnapMirror asynchronous

SnapMirror technology offers a fast and flexible asynchronous enterprise solution for replicating data over LANs and WANs. SnapMirror technology transfers only changed data blocks to the destination after the initial mirror is created, significantly reducing network bandwidth requirements.

The following are recommendations for SnapMirror for SQL Server:

  • If CIFS is used, the destination SVM must be a member of the same Active Directory domain of which the source SVM is a member so that the access control lists (ACLs) stored within NAS files are not broken during recovery from a disaster.

  • Using destination volume names that are the same as the source volume names is not required but can make the process of mounting destination volumes into the destination simpler to manage. If CIFS is used, you must make the destination NAS namespace identical in paths and directory structure to the source namespace.

  • For consistency purposes, do not schedule SnapMirror updates from the controllers. Instead, enable SnapMirror updates from SnapCenter to update SnapMirror after either full or log backup is completed.

  • Distribute volumes that contain SQL Server data across different nodes in the cluster to allow all cluster nodes to share SnapMirror replication activity. This distribution optimizes the use of node resources.