MSCS Failover when Nimble Controler is switching

[dominic_david]

Hello,

I have setup a two node Windows 2008 R2 Cluster in vSphere 5.5 Update 3. The settings are taken form the VMware "Setup for Failover Clustering and Microsoft Cluster Service" dokumentation.

The Disk are VMDK-Files on the same vSphere Server.

Everytime the Nimble controllers are switching (install a Nimble Update) I get a MSCS Failover. The Cluster is loosing the connection to the cluster disk.

Nimble CS300

Nimble OS Version: 2.3.14.0-325711-opt

NCM for vSphere: 2.3.1

Is there a special setting to use a MSCS with Nimble and vSphere to prevent the failover?

Regards

Dominic

[Nick_Dyer · Accepted solution]

Hi Dominic,

Sounds like the VMware timeouts aren't set correctly, so when paths are disconnected/reconnected VMware automatically creates a failover.

You can read the correct timeout settings for VMware in this KB article on Infosight here.

Engineering have confirmed that in NimbleOS 3.0, the VMware plugin and NCM will automatically correct any incorrect timeout settings within the ESX server - which is good news!