Sunday, August 04, 2013

70462 My Virtual server Spec

My computer has the below spec
Intel(R) Core (TM) i7-3612QM CPU @2.10GHZ (8 CPUs)
Memory : 8192MB RAM
Harddisk : 1 TB
OS : Windows 8 Pro 64-bit

My Virtual server setting
Domain Controller (DC) - 1 Virtual Processor, RAM 512, Virtual harddisk - size 127gb (It will expand when disk needed); I DC file size (3 gb)
SQL-A - 1 Virtual Processor, RAM 1 GB, Virtual Harddisk - auto expand as well (11 GB when checking)
SQL-B - 1 Virtual Processor, RAM 1 GB, Virtual Harddisk - auto expand as well (11 GB when checking)
SQL-C - 1 Virtual Processor, RAM 1 GB, Virtual Harddisk - auto expand as well (11 GB when checking)
SQL-D - 1 Virtual Processor, RAM 1 GB, Virtual Harddisk - auto expand as well (11 GB when checking)

Most of the time I only turn On DC, SQL-A, SQL-B and it more then enough for Testing.

Saturday, August 03, 2013

Quorum - Node and File Share Majority; Shutdown one node

Cluster service, Change Quorum to Node and File Share Majority; Shutdown one node

SQL-B is Primary Replicas


Shutdown SQL-B server, simulate Primary replicas SQL-B fail and see if the database will auto-failover to SQL-A

Cluster service - SQL-ALWAYSONCL.Contso.com still available with warning
SQL-B Node is down




Failover database Happen and SQL-A become Primary Replicas and Availability Database and Availability Group Listeners still can be access



SQL Availability Dashboard
- SQL-A Replicas is online
- SQL-B Replicas is offline


SQL Error log

Availability replica
'SECONDARY_NORMAL' to 'RESOLVING_NORMAL'; to 'PRIMARY_PENDING'; to 'PRIMARY_NORMAL'
Availability database
'SECONDARY' to 'RESOLVING'; Changed to 'PRIMARY'




Start SQL-B server
SQL-B node is back online
Current Host server still SQL-A and failover didn't happen



SQL Server


SQL Availability Dashboard


SQL Error Log (SQL-A)

Change Quorum to Node and File Share Majority


70-462 Cluster service, Change Quorum to Node and File Share Majority
Can sustain failure of 1 node(s) if the file share witness remains available.
Can sustain failure of 0 Node(s) if the file share witness becomes unavailable

Compare to Default Quorum - Node Majority (not recommended for your current number of nodes
can sustain failures of 0 node(s)










Friday, August 02, 2013

2 nodes with Default Quorum (Node Majority) setup; Shutdown Secondary node

Cluster service, 2 nodes with Default Quorum (Node Majority) setup;
Shutdown server node with Secondary Replicas host (SQL-A); What will happen to SQLServer?



Shutdown Server node SQL-A where Secondary Replica sit.

After SQL-A server node shutdown,
- SQL-B SQLServer availability database testing in "Recovery Pending"
- AlwaysOn High Availability groups - ALWAYSONAG in "Resolving"
- Availability Replicas, SQL-B in "Resolving" mode
- Availability Group Listeners is missing


Availability Dashboard


SQL Error log


why this happen ? We expect the Sqlserver (SQL-B) for Primary should not affected if Secondary server down (SQL-A).

System Error log (SQL-B)


Remember in the cluster quorum configuration, we use default quorum with node majority whereby cluster service will sustain with 0 node down?
When on node down, cluster service can't sustain and SQL Availability database is down as well
2 nodes with Default Quorum (Node Majority) setup; Cluster service will sustain with 0 node down

 Restart SQL-A server
SQLServer up

SQL Error Log

Thursday, August 01, 2013

Shutdown Primary sqlserver service; 2 nodes with Default Quorum (Node Majority) setup;

SQL Cluster : 2 nodes with Default Quorum (Node Majority) setup;
Testing with Shutdown Primary sqlserver service and verify if the database failover to secondary Availability Replicas and secondary availability promote to Primary Replicas

Availability Group Properties
Synchronous commit , Failover mode = Automatics
SQL-A primary, SQL-B secondary




Shutdown SQL-A, SQLService to simulate Primary SQLServer (SQL-A) fail and see if the database will auto-failover to SQL-B

SQLServer SQL-B become Primary Replicas; Auto-Failover is happen when SQL-A fail.
From Replicas, we can see SQL-A become secondary and down



SQL AG (alwaysonag)Dashboard


SQL-B error log



SQL-B:
Availability Replica status : SECONDARY_NORMAL -- RESOLVING_NORMAL  -- PRIMARY_PENDING -- PRIMARY_NORMAL

Availability Group database status : SECONDARY - RESOLVING --PRIMARY

Bring back SQLService - SQL-A
SQL-B still act as Primary Availability Replicas
SQL-A is back online and show as Secondary Availability Replicas





SQL-B : SQL Error log




Shutdown secondary sqlserver service; 2 nodes with Default Quorum (Node Majority) setup;

Cluster service, 2 nodes (SQL-A,SQL-B) with default Quorum (Node Majority) setup;
SQL Server AG group (alwaysonag) with 2 node (SQL-A, SQL-B)
Test shutdown secondary sqlserver service (SQL-B), will it affect primary replicas (SQL-A)



Shutdown SQL-B, SQLService to simulate SQL-B SQLServer fail

From SQL-A, sql-error log:
- A connection timeout has occurred on a previously established connection to availability replica 'SQL-B' with 'DAF385F5-FE27-4A48-A5DC-7E21214498C7].
Either a networking or a firewall issue exists or the availability replica has transitioned to the resolving role
- AlwaysOn Availability Groups connection with secondary database terminated for primary database 'testing' on the availability replica with Replica ID : {daf385f5-fe27-4a48-a5dc-7e21214498c7}

SQL-A database still ONLINE, from
AlwaysOn High Availability  --Availability Group -- AlwaysONAG --  Availability Replicas
we can see SQL-B replicas is down



SQL AG (alwaysonag) Dashboard



From Availability group state -- Critical --- Critical (1), Warning (3)


Some synchronous replicas are not synchronized
In this availability group, at least one synchronous replica is not currently synchronized. The replica synchronization state could be either SYNCHONIZING or NOT SYNCHRONIZING

Some availability replicas are not synchronizing data
In this availability group, at least one secondary replica has a NOT SYNCHRONIZING synchronization state and is not receiving data from the primary replica

Some availability replicas are disconnected
In this availability group, at least one secondary replica is not connected to the primary replica. The connected state is DISCONNECTED

Availability group is not ready for automatic failover
The availability group is not ready for automatic failover. The primary replica and a secondary replica are configured for automatic failover. however, the secondary replica is not ready for an automatic failover. Possibly the secondary replica is unavailable, or its data synchronization state is currently not in the SYNCHRONIZED synchronization state.

From Availability replica -- Critical (1), Warnings (1)



Start SQLService --SQL-B
From SQL-A, sql-error log:
A connection for availability group 'ALWAYSONAG' from availability replica 'SQL-A' with id [C8B53A5D-75CE-40CC-BB27-50CFBA48CB32] to 'SQL-B' with id [DAF385F5-FE27-4A48-A5DC-7E21214498C7] has been successfully established. This is an informational message only. No user action is required.

AlwaysOn Availability Groups connection with secondary database established for primary database 'testag' on the availability replica with Replica ID :{DAF385F5-FE27-4A48-A5DC-7E21214498C7}. This is an informational message only. No user action is required

AlwaysON Availability Groups connection with secondary database established for primary database 'testag' on the availability replica with Replica ID :{DAF385F5-FE27-4A48-A5DC-7E21214498C7}. This is an informational message only. No user action is required

SQL-B back online in availability Replicas


SQL AG (alwaysonag) dashboard

2 nodes with Default Quorum (Node Majority) setup; Cluster service will sustain with 0 node down

Test cluster service, 2 node with Default Quorum (Node Majority);
What will happen if one of the node down;

Create a cluster – clustertest.contso.com with 2 node : SQL-C and SQL-D

Default Quorum : Node Majority









More information about quorum
WSFC Quorum Modes and Voting Configuration (SQL Server)http://msdn.microsoft.com/en-us/library/hh270280.aspx#QuorumModes


With Default Quorum (Node Majority) and 2 nodes cluster


Quorum Configuration warning
Node Majority - Warning: Failure of a  node will cause the cluster to fail. Check the status of the nodes


Shutdown SQL-D, simulate SQL-D server fail
System event log at SQL-C


SQL Cluster is missing from Failover Cluster Manager

Bring SQL-D back online
Cluster service - Clustertest.Contso.com back online.



Cluster service will offline, if other node down.
Custer service will sustain with 0 node down.