0

MY understanding of using WID is only one server is the primary ADFS and this has write access to the DB only. Therefore, why is an internal load balancer required? The fail over is a manual process. I can se why this would be useful when using SQL as this works differently.

we have a client and I can see they have followed the Microsoft design to a T because Microsoft does recommend both an internal and external load balancer.

The environment is in Azure. Slightly off topic but still relating to load balancers is for health probe that specified HTTPS/ 443 which is fine but they used the Basic SKU. I did not think 443 was supported in the basic SKU and this was standard only?

Could be I am missing something?

Thank in advance.

Improve this question

1 Answer 1

Reset to default
0

yes, you use an internal load balancer for availability.

With Windows Internal Database (WID), each server maintains its own WID, and non-primary servers replicate from the primary server every few minutes.

Replication is at an 'application' level, not a database level (in other words, the AD FS service pulls changes from the primary and writes to its local WID.

If the primary server goes down, other servers can continue to answer (at least in the short term) and can become a primary server.

Improve this answer
1
  • Thanks for the response. My understanding is the DB config is replicated from the primary which has write access to the DB. The others in the ADFS farm only have read access. No changes are made to the Federated servers until the primary is back up. Making another server the primary is a manual process using Powershell.
    – IT_Makes_you_old
    Feb 1, 2021 at 21:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .