What if CPS on-prem will go down?
02-09-2017 06:08 AM
Hi, i am wondering is there any mechanism to garanty High Avaliability of the service? What is the best practice? Is thre a clustering? Roadmap?
Solved! Go to Solution.
02-09-2017 06:35 AM
High Availability can be provided by simply deploying more Centrify Connectors (by default the server where you are installing CPS is also acting as a Centrify Connector).
There is also other option comming in the roadmap, especially for Disaster Recovery and High Availbilty of the CPS database server (the Windows server where CPS is installed). I will let PM comment on this as I don't have release in head.
At the moment, recommendation are to build your Windows Server where CPS is installed with mirrored HDD or SAN in order to be able to quickly restore or balance a failing system.
Hope that helps,
Don't forget to mark posts as "Solution" to help other identify quickly the answers. And don't be afraid to deliver Kudos as well when you are happy with the solution ;)
02-09-2017 12:25 PM
As Fabrice stated, the Centrify connector is automatically highly-available (all you need to do is to deploy more).
The only design consideration is if you're using "Resource Subnet Mapping" (e.g. designating connectors per network segments) then you must have multiple connectors per network segment to ensure proper availability.
More availability options are coming soon, including support for any number of Web/application tier instances of CPS in a cluster, and separate database instances (the latter should always be used with highly-available filers).
09-27-2017 05:54 AM
Hello and welcome to Centrify.
(I am currently writing a Lab entry on the TechBlog to discuss this at length)
Are you using the customer-managed version of Privilege Service? What version are you using?
09-28-2017 10:50 PM
Here are a few resources for you:
Starting on 17.7 customer managed CPS (released Wednesday) the setup options have changed. We removed the stand-alone option, and we now require a highly-available Windows Failover Cluster as the server infrastructure. These articles and the documentation have additional info. Use the lab entries above to set up and test the capability.