GIPSE (formerly AMDS) Store

This page collects the requirements, design and development artifacts for building the AMDS Store that will hold BioSense data extracts. The BioSense AMDS services will access this store to return counts from Real-Time, DoD and VA data sources.

This store will be hosted within the CDC network and control is limited to only approved personnel by the BioSense business steward. While the structure of the store is open (as the other project artifacts), the contents of the store are restricted to CDC users granted access. The BioSense AMDS service will provide access through the PHGrid based on permissions granted by the BioSense business steward.

The current plan for the AMDS Store is that it is populated daily with reports generated from the BioSense data team (Roseanne's team). This is done as an interim measure until the BioSense system can be modified to provide a system interconnect. We're doing this specifically to support a demonstration in time for the PHIN conference. As such we are using the existing reporting framework and existing reporting servers. To move this to full-fledged production requires an evaluation of the load during the demonstration to identify hardware and software required to support the daily report generation, extract and load. Additionally, the BioSense NextGen plan includes an evaluation and tuning task for the entire system. This evaluation and tuning needs to be performed on the reporting servers to optimize Unix and Oracle configurations to minimize the processor, disk and memory loads to only necessary tasks.

The daily report will include counts by day, syndrome, zip code and state. Roseanne and Jerry are investigating if we will also use an extract that includes the top 10-20 sub syndromes. This daily report will include the previous 40 days worth of data. The daily report will include the creation day in the file name. Two days worth of reports will be stored on the reporting server (current day and previous day). The daily report will be created in a comma separated format.

A load process will retrieve the daily report from the reporting server; process it and insert new rows and update any existing rows in the AMDS Store with the data contained within the report.

The actual report generation time will depend on decisions made by Roseanne and Jerry as to what is the most appropriate given data and usability requirements. The AMDS Store load time will be dependent on when the extract report is created.