SDS:June 9 2009 Meeting

Agenda, June 9, 2009

Patrick, Stephan

plan for backups
bco-dmo ontology, populate the individuals, and get them into jspace
rni packages to install for Patrick
trac for admin
- setup at https://scm.escience.rpi.edu/admin
- currently only admin users can access (httpd.conf auth setting)
- anonymous users still have permissions, these to be removed.
- currently can't log out once logged in.
- svn integration not setup, will create an admin directory at https://scm.escience.rpi.edu/svn/private/admin
escience, tetherless, provenance NFS
- Purpose of NFS mounts on production systems
  - backup storage
    - /backup
    - each machine will have a backup directory mounted from another machine. If one machine goes down; we lose one running system and one system's archived backups (assuming they have not been archived to tape). We can rebuild that system from its backup and will need to start a new archive of backups for the systems whose backups were lost.
  - data storage
    - /data
    - one single data directory which all production systems mount. keeping all data in one location will make archiving it more straightforward. This data location should be a LARGE RAID array.
  - home directories
    - ....not sure on this one...
- Patrick and I are absolutely NOT setting this up
NcML Aggregation
- prior ESG module for BES
  - NCESGCommit - on commit netcdf aggregation is performed
  - NCESGAggregationServer - during a get, aggregation is performed on DAP objects