Manage high availability and disaster recovery

Chapter summary

• High availability is not equivalent to disaster recovery.

• Log Shipping is not a high availability technology.

• Log Shipping supports multiple secondary servers.

• With Log Shipping, users cannot access the secondary database when a log backup is being restored.

• Failover clustering and Availability Groups rely on the Windows Server Failover Cluster (WSFC) feature.

• Failover clustering and Availability Groups support automatic failover.

• The scope of protection in Availability Groups is at the database level.

• Availability Groups support three failover partners with SQL Server 2016.

• Each replica in an Availability Group maintains its own version of the database.

• Availability Groups support asynchronous and synchronous communication between the replicas.

• With an Availability Groups you can only perform a manual fail over to a synchronous replica.

• A forced failover in an Availability Group to an asynchronous replica can result in data loss.

• Availability Groups support readable secondaries.

• Availability Groups are the only high-availability technology that allows you to scale out your database solution.

• You can offload reports, read operations, database consistency checks and backup operations to a readable secondary.

• SQL Server 2016 Standard Edition supports Basic Availability Groups, which is intended to replace Database Mirroring.

• Distributed Availability Groups are designed to be used between data centers where you want to minimize the network usage across the WAN link between your data centers.

• SQL Server 2016 provides limited support for DTC in Availability Groups.

• Availability Groups are more complicated to maintain and administer as they do not automatically synchronize logins, SQL Server Agent jobs and other external database dependencies between replicas.

• You can install a number of SQL Server Failover cluster instances on a Windows failover cluster.

• The scope of protection in a failover cluster is at the instance level.

• WSFC no longer require a domain.

• Failover clustering uses shared storage which represents a single point of failure.

• Failover clustering will work with SMB 3.0 file shares as a location for your database files.

• Cluster Shared Volumes (CSVs) represent a new clustered file system in Windows Server.

• SQL Server 2014 added support for (CSVs).

• CSV remove the dependency on physical disks, which reduces the fail over time in a failover cluster.