The CERN Beam Instrumentation Group Offline Analysis Framework

Beam instrumentation (BI) systems at CERN require periodic verifications of both their state and condition. An instrument's condition can be diagnosed by looking for outliers in the logged data which can indicate the malfunction of a device. Presently, experts have no generic solution to observe and analyse an instrument's condition and as a result, many ad-hoc Python scripts have been developed to extract historical data from CERN's logging service. Clearly, ad-hoc developments are not desirable for medium/long term maintenance reasons and therefore a generic solution has been developed. In this paper we present the Offline Analysis Framework (OAF), used for automatic report generation based on data from the central logging service. OAF is a Java / Python based tool which allows generic analysis of any instrument's data extracted from the database. In addition to the generic analysis, advanced analysis can also be performed by providing custom Python code. This paper will explain the steps of the analysis, its scope and present the kind of reports that are generated and how instrumentation experts can benefit from them. It will subsequently demonstrate how this approach simplifies debugging, allows code re-use and optimises database and CPU resource usage.