Frontend vs backend processes

Broadly the work of the front end and back end processes is broken out as follows:

Backend

1. Harvest new, updated and deleted records from the LMS hosts within the consortium

2. Ingest imported records, update the cluster records with any changes and update the search index

3. Poll the LMS systems for any changes to the records that have been borrowed between one LMS and another for circulation updates.

The backend is responsible for making calls to third party systems and then processing the results that are received. Beyond harvest polling and search index updates, the volume of external connections made every hour is relative to the number of in-flight requests: see polling intervals.

Front end

1. Accept new requests for items

2. Authentication for patrons, for sites not using SSO

3. Obtaining item availability from sites that are known to hold it

4. Handling requests to perform support type jobs using the api as we do not have access to the backend (normally these are quite small, but may occasionally be significant)

5. Handling requests from the Admin UI to make life easier for the administrators, including

   1. request queries and operations

   2. catalog queries

   3. shared index queries

   4. library and consortium configuration queries and operations