Access Based Enumeration (ABE) and High Availability Clustering
We had a good question come back through the blog the other day - "How do you support ABE on clustered file servers?"
Although our plans for this integrating this feature are geared more toward the Longhorn timeframe, one of our Program Managers for ABE, Sundar Subbarayan, was able to provide this solution for the current version of ABE to work on clustered file servers:
1. Install ABE tool from the download location and install it on each of the nodes in the cluster.
2. From the cluster administrator, for each share in the cluster, create a resource of type “application”. In the command to execute for this resource (under parameters), the following command should be entered “cmd /k abecmd /enable <sharename>”. Additionally, if you don’t want that cmd.exe will appear on the desktop, ensure that the option “Allow interaction with the desktop” isn’t checked.
3. Make this resource depend on the share resource. (Make sure that both the application resource and the share resource are part of the same group)
4. Take the group offline and back online to ensure that the shares that need to be ABE enabled are actually so.
From this point on, you are good and whenever the share fails over to another node, then ABE setting follows.
Cheers,
Ward Ralston
