Accounting data is processed and stored in a database in order to display eye-candy charts grouped by user, queue, full cluster or historical (yearly). Moreover, you can use several filter options to customize the results.
An example of data charts
At RDlab we have been running an Oracle Grid Engine cluster for several years and, after realizing that available solutions did not fit our needs, we decided to develop a good one.
How does it work?
On one hand, s-gae runs in POSIX environments using a standard PHP script executed by a CRON job (or Windows Task Scheduler). The previous named script calls the qacct tool to gather the accounting information and, after being processed and summarized, it is stored in a database. On the other hand, the application has a web frontend that displays this information and lets the user choose among several filters.
s-gae's work diagram
Who is currently using s-gae?
Several investigation departments/institutions are currently using s-gae to display the statistics of their live clusters. These are some of them:
The Queensland Brain Institute, The University of Queensland
French agricultural research organization for development in the South and the French overseas regions.