TarMK Farm Automatic Failover

[nsvsrk]

Hi,

Five Deployment Scenarios were discussed for AEM 6.0.

In the TarMK Farm, (3rd scenario) is there an Automatic Fail over or it is manual?

Please refer to http://docs.adobe.com/docs/en/aem/6-0/deploy/recommended-deploys.html#Microkernels

Appreciate your response.

Thanks,

Rama.

[Jörg_Hoh]

Hi,

In case of the TarMK farming approach, you have a number of independent instances, so no inherent high availability. You normally don't do failover, but you run multiple instances in parallel, and you have a loadbalancer in front of it.

Kind regards,
Jörg

[Sham_HC]

You can configure lb for automatic failover. 

[nsvsrk]

Hi,

Thanks for your reply.

I am sorry, what is lb?

Is it that the default it is manual, but could be configured for automatic fail over?

Thanks,

Rama.

[Sham_HC]

lb means load balancer. This is not inbuilt with aem & discuss with your system admin team they will help you out.

[Opkar_Gill · Accepted solution]

Hi Rama,

as has been mentioned, with the TarMK farm, you are running multiple separate Publish instances all independently kept in synch by the author instance.

In front of you publish instances, you have your web server with the dispatcher module(usually in a 1 to 1 relationship with a publish instance) 

In front of the web server you would have a load balancer, which redirects traffic to each publish instance based on load. If the load balancer detects a publish instance is down, it will redirect traffic to another live instance in the farm. This is the failover scenario that is mentioned in the link you gave, however the section could be more descriptive.

Automatic failover usually applies to clustered systems[0], which the TarMK farm is not.

Regards,

Opkar

[0]https://en.wikipedia.org/wiki/High-availability_cluster