AlwaysOn availability doesn't fail over when power off primary

Question

AlwaysOn availability doesn't fail over when power off primary

martin.vogt 19133

SSC Veteran

Points: 216
More actions
June 28, 2017 at 9:29 am

#331338

Hi there
in a hyper-v environment I have set up two SQL server 2016 standard (on Windows Server 2016) servers and installed the fail over cluster feature on both nodes. After this, I created an availability group and configured a database as part of the availability group. The fail over to the secondary node works fine, when i fail over manually or when I perform a regular shut down of the primary.
When i just disconnect the primary from the network or do a forced power-off, the cluster (connected to the listener IP address) becomes unavailable.
After this, i connected to the secondary node via RDP and used the Mgmt Studio for checking the status. Looking to the availability replicas, the secondary node (which I'm connected to) says "Resolving", for the primary there is a question mark.
I assume, that my cluster lost quorum. From my understanding, the new feature called dynamic quorum does not need to setup a witness anymore, or are i'm wrong?
Some help is really appreciated!
Many thanks
martin

Viewing 5 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply

Perry Whittle SSC Guru Points: 233835 More actions · Answer 1

martin.vogt 19133 - Wednesday, June 28, 2017 9:29 AM
The fail over to the secondary node works fine, when i fail over manually or when I perform a regular shut down of the primary.
When i just disconnect the primary from the network or do a forced power-off, the cluster (connected to the listener IP address) becomes unavailable.
After this, i connected to the secondary node via RDP and used the Mgmt Studio for checking the status. Looking to the availability replicas, the secondary node (which I'm connected to) says "Resolving", for the primary there is a question mark.

So the availability group runs just fine on the 2nd node AVG02, have you checked the application event log and more importantly the cluster events for more info?

martin.vogt 19133 - wenesday, June 28, 2017 9:29 AMFrom my understanding, the new feature called dynamic quorum does not need to setup a witness anymore, or are i'm wrong?

Wrong, you still need to configure an appropriate quorum for your cluster design

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs" 😉

martin.vogt 19133 SSC Veteran Points: 216 More actions · Answer 2

Perry Whittle - Wednesday, June 28, 2017 9:58 AM
So the availability group runs just fine on the 2nd node AVG02, have you checked the application event log and more importantly the cluster events for more info?

Log says: Always On Availability Groups: Waiting for local Windows Server Failover Clustering node to come online. This is an informational message only. No user action is required (Event-ID 41054.

Perry Whittle SSC Guru Points: 233835 More actions · Answer 3

have you also run a cluster validation

-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs" 😉

martin.vogt 19133 SSC Veteran Points: 216 More actions · Answer 4

Thanks for the quick reply. After i used the Start-ClusterNode -ForceQuorum cmdlet, i now have the option for manual failover using SSMS (goto AG => r-click => failover). Now i got another error: "Object reference not set to an instance of an object. (Microsoft.SqlServer.Management.HadrTasks)"