After a successful installation and configuration of iRODS server, it is possible to create an instance of Safe Replication Service. As explained above Safe Replication Service is a service which replicates the data ingested by the community partner according to the data management policies agreed on (in particular the number and location of replicas need to be defined). The information about the location of the replicas of the data object is available in a PID record created for the ingested object.
First step when creating an instance of Safe Replication Service is to install an iRODS server and configure its resources (especially the integration with back-end storage must be assured). More information about the configuration of storage resources can be found in Resources Section.
Second step of creating an instance of Safe Replication Service is to federate with other iRODS servers. A connection to other data centers that will store replicas must be established. Please refer to Creating Federation section for details. Furthermore a collection where the data will be ingested must be created and an access to the data manager on that collection must be granted. The ingest of the data can be done either by using normal icommands or by federating with iRODS installation on the community side (if such an installation is available).
All the data objects available in Safe Replication Service must get a PID. For that a PID manipulation script has to be installed. The process of the installation is described in previous section. To add the PIDs in EPIC system a prefix has to be registered (do we provide a documentation on that??). …where the credentials need to be stored…
The subsequent step of creating an instance of Sefe Replication Service is to configure data management policies in form of rules. This step is very community and data set specific. Thus there is no generic documentation available. The Safe Replication Task Force, however, prepared an example of simple rules to replicate the data and register PIDs. This can be used as a basis for defining more specific rules and data management policies. Please refer to the documentation prepared by Willem Elbers.