Presentation is loading. Please wait.

Presentation is loading. Please wait.

Designing Database Solutions for SQL Server

Similar presentations


Presentation on theme: "Designing Database Solutions for SQL Server"— Presentation transcript:

1 Designing Database Solutions for SQL Server
Christian Bolton | Technical Director, Coeo Graeme Malcolm | Microsoft

2 Course Topics Designing Database Solutions for SQL Server
01 | Managing a SQL Server Environment 02 | Designing Database Security 03 | Designing a Backup & Recovery Solution 04 | Designing a High-Availability Solution 05 | Troubleshooting & Maintaining a database

3 Christian Bolton | Technical Director, Coeo Graeme Malcolm | Microsoft
04 | Designing a High-Availability Solution Christian Bolton | Technical Director, Coeo Graeme Malcolm | Microsoft

4 Module Overview Log Shipping Windows Server Failover Clustering
AlwaysOn Failover Clustering AlwaysOn Availability Groups Advanced deployment scenarios

5 Log Shipping Log shipping is a three-phase process
Backup log on primary server Copy log to secondary server Restore log on secondary server Secondary servers can be readable but be aware of caveats Manual failover to secondary servers

6 Windows Server Failover Clustering
SQL Server AlwaysOn technologies rely on Windows Server Failover Clustering A Windows Server Failover Cluster includes Cluster nodes Cluster resources Cluster roles Failover Cluster Instance Each node maintains a copy of the cluster configuration During failover and failback, ownership of resources passes from one node to another Windows Cluster

7 Quorum Quorum defines the minimum number of votes required for the cluster to remain online More than 50% must be counted to achieve quorum Quorum configurations Node majority Node and disk majority Node and file share majority Disk only

8 AlwaysOn Failover Clustering
SQL Server is a cluster-aware application Instance-level protection – databases, logins, Scheduled Jobs Transparent to end-user applications Requires shared storage SQL Server installations are Failover Cluster Instances (FCI)

9 Enhancements to AlwaysOn Failover Clustering
Enhancements in SQL Server 2012 Detailed health monitoring Flexible failover policies Improved troubleshooting Support for multi-subnet clustering Local tempdb Enhancements in SQL Server 2014 Support for Cluster Shared Volumes

10 Enhancements to AlwaysOn Failover Clustering
The sp_server_diagnostics stored procedure provides health monitoring and flexible failover policies 5 – Failover/restart on any qualified failure conditions 4 – Failover/restart on moderate SQL Server errors 3 – Failover/restart on critical SQL Server errors 2 – Failover/restart on server unresponsive 1 – Failover/restart on server down 0 – No automatic failover

11 AlwaysOn Availability Groups
Listener Clustered Network Name Clustered IP Address ERP CRM Availability Group ERP CRM Availability Group ERP CRM Availability Group Asynchronous Synchronous Windows Cluster Node

12 AlwaysOn Availability Groups
Synchronous-commit mode Primary replica send transaction log changes to a secondary replica Secondary replica applies the changes in the log to its copy of the database and sends an acknowledgement to the primary replica Primary replica commits the transaction

13 AlwaysOn Availability Groups
Asynchronous-commit mode Primary replica send transaction log changes to a secondary replica and then commits the transaction Secondary replica applies the changes in the log to its copy of the database

14 AlwaysOn Availability Groups
Offload workload with Active Secondary Replicas Near-real-time reporting from replicas Read-only routing to readable replicas Offload Full and Log Backups to replicas No differential and copy_only full backups

15 AlwaysOn Availability Groups
Failover options Automatic failover No data loss Planned manual failover Forced manual failover Possible data loss Use the Start Failover Wizard to initiate manual failover Don’t use Failover Cluster Manager to initiate failover

16 AlwaysOn Technologies

17 Solution Example 1 Existing architecture FCI for HA
Failover Clustering SAN User and Application Access Production Environment Production Server Passive Node MPLS Disaster Recovery Environment DR Server Log Shipping Existing architecture FCI for HA Log Shipping for DR Requirement to test DR every 6 months

18 Solution Example 1 New architecture AG with auto-failover for HA
Synchronous Replica User and Application Access Production Environment Primary Replicas Sync Replicas MPLS Disaster Recovery Environment DR Server Async Replicas Asynchronous Replica Automatic Failover New architecture AG with auto-failover for HA Async replicas for DR

19 Solution Example 1 DR can be tested easily
Synchronous Replica User and Application Access Production Environment Primary Replicas Sync Replicas MPLS Disaster Recovery Environment DR Server Async Replicas Asynchronous Replica Automatic Failover DR can be tested easily Make async replicas sync and initiate failover Configure new auto- failover partner and remote async Solution can run from DR and still have DR capabilities

20 Peer-to-Peer Replication
Solution Example 2 Failover Clustering SAN Website Users Web Production Environment (Hosted) Web Production Server Passive Node MPLS HQ Production Environment (On-Premises) HQ Production Server Peer-to-Peer Replication User and Applications Scale out solution with P2P replication FCIs for HA Requirement to simplify by removing multi-site write capability Still need the data in HQ for reporting

21 Async Replica for Reporting
Solution Example 2 Synchronous Replica User, Web, and Application Access Web Production Environment (Hosted) Primary Replicas Sync Replicas MPLS HQ Production Environment (On-Premises) Data Warehouse Asynchronous Replica Automatic Failover Async Replica for Reporting SSIS File Share Witness AG with autofailover for HA Async replica back to HQ for reporting Node vote for async replica removed File share witness added

22 Solution Example 3 Failover Cluster Instances for HA SAN replication
Databases attached during DR Requirement to remove dependency on SAN replication due to cost Failover Clustering SAN Production SQL Server SAN Replication DC 2 DC 1 Production Site ( Network) Disaster Recovery Site ( Network)

23 Availability Group Replica
Solution Example 3 Failover Cluster Instances for HA Asymmetric storage Asynchronous Availability Group Replica for DR File share witness added Votes removed from DR Production Site ( Network) DC 1 SAN Production SQL Server File Share Witness Availability Group Replica Failover Clustering Disaster Recovery Site ( Network) DC 2 SAN Production SQL Server File Share Witness

24 Designing a High-Availability Solution
Summary Log Shipping Windows Server Failover Clustering AlwaysOn Failover Clustering AlwaysOn Availability Groups Advanced deployment scenarios

25


Download ppt "Designing Database Solutions for SQL Server"

Similar presentations


Ads by Google