Cluster failover in Sense doesn't work.

[Jio]

Hallo All,

we have a multi node sense deployemnt.

Everything is registers well in QMC. The central en Rim node works fiine.

but when we shut  down the central node, the Rim node Doesn’t take over the funtion of the central node even after 15 minutes ( waittime is 10 minutes).

the is no firewall Between the systems. And every tcp Port is accessible booth sides.

Any ons aan Kees wherry to check and how to solve this.

Extra info: the postgresql is ruiming om the Central node (by  default)

[Benoit_C · Accepted solution]

Hello @Jio,

If you completely shut down the central, the failover won't work.

Failover works if Engines, Proxy, Scheduler, Printing or Repository services are stopped.

But if you stop the Repository DATABASE service, then nothing will work anymore.

For such scenario, you would need to install the Repository Database in a dedicated server.

In the below article we mention it:

https://community.qlik.com/t5/Knowledge/Failover-Central-node-requirements-and-considerations/ta-p/1...

Regards,

Benoit

[Jio]

Hello Benoit_C,

Thanks for the response.

Now i understand it. To use Qliksense for Failover (Cluster services), you need to put the fileshare and the Reporsitory database on a different dedicated server. 

Many Thanks,

Regards 

Jio

[Benoit_C]

Hello Jio,

I'm glad it has been an useful answer. 

Regards,

Benoit