We definitely need a crisis cleanup fest. There are many, many things that are in a state of disarray for this system.
Here is a start of a problem list.
- PBS job accounting information collection
- Linux accounting record collection
- MPI job state and timing record collection
- cleanup of old database - keep current good data and migrate to new database
- removal of all old system checkout and install areas
- shutdown of all old system daemons and cleanup of all system startup files for old system (and cron)
- removal of all old system worker and head node collectors
- install of new system on lqcdbackup
- auditing / cleanup procedures for database
- install and configuration of new framework on LQCD clusters
Beyond that we have the the user interface to deal this for this data...