ELUNA 2015

As a member of the Orbis Cascade Alliance, Washington State University Libraries share a Primo installation with 36 other regional academic libraries. The Alliance installation, while not strictly hosted, has some similarities to a hosted Primo implementation, where there is no direct server-side access. In this environment, notifications of service outages are only available through Ex Libris cloud status notifications. These notifications, while a good baseline, aren’t granular enough to indicate service slowdowns (in addition to full-scale outages).

In order to proactively identify reportable issues, and also to get baseline quality of service statistics, we implemented a local alerts system using Nagios and CasperJS to automate and replicate user tasks. These tasks are monitored and logged, and when performance thresholds are exceeded library systems personnel are notified. Additionally, this data can be used to monitor historical performance. More importantly, this is an end-to-end test that replicates typical user behavior across our web interfaces: the Libraries website, the initial login to PDS, the time to see a results set in Primo, through to resource access. The full stack is tested and any weak link in that chain can be identified rapidly.

The presenters will demonstrate the tools, discuss statistics collected (including identifying any trends), and share key findings post-implementation.

General/Strategic (Not Product Specific)
Primo