Configuring Peer Repositories

As introduced in the previous section, peer repositories enable you to set up one (or more) Watson Explorer Engine installations as servers for repository data. You can then configure other Watson Explorer Engine installations as clients of those servers. Clients can pull data from the server and optionally push data to the server; peer repository servers cannot initiate any action themselves.

Peer repository clients authenticate themselves to peer repository server systems using a Watson Explorer Engine user that need only exist on the server (though it can also exist on the client) and must have been granted one or both of the following permissions in the Watson Explorer Engine administration tool:

• Peer Repository Download: enables the client to download/pull data from the server
• Peer Repository Upload: enables the client to upload/push data to the server

In most cases, you will want to create a unique user account that clients can use to access the repository on your peer repository server. Creating a specialized account for this purpose is good security and administrative practice because it enables you to limit the permissions that you grant to this account to those that are required with peer repository usage, and also because it increases the visibility of the fact that special privileges have been assigned to an account on your system. If you do not want to create a specific account for this purpose, you can assign one or both of these permissions to an existing account, but that can be easy to overlook and may cause other system administrators to assume that administrative users on all Watson Explorer Engine installations require one or both of these privileges.