Custom Query (121 matches)

Some aggregates report ephemeral interfaces which were created to support experiments, and physical interfaces on resources may change over time.

The database should record last_updated for interface entries, and gmoc.py lookup should ignore old interfaces when reporting data, the same way it ignores old records of other types.

At the moment, when data is downloaded using gmoc.py, the API searches for all matching data regardless of last_update time. This isn't sustainable in the long-term because of data size, and, in the short term, may get us into trouble with badly-formed data from the past.

Modify gmoc.py to allow an optional threshold argument to GMOCClient.load(), and only load values whose last_update time is more recent than that threshold.

When new GENI sites start submitting monitoring data to GMOC, they follow the procedure at ​http://groups.geni.net/geni/wiki/GENIMetaOps/SiteCredentials to report their credentials and contact info to the GMOC service desk. The service desk then follows an internal procedure to add submission credentials to gmoc-db, and also to add organizational metadata to gmoc-db tables.

Currently, the data which is added to gmoc-db via this procedure:

• Is not sufficient for site data submission via gmoc.py (so sites need to submit some separate metadata using gmoc.py itself before they can submit data)
• Is not sufficient for Organization object download via gmoc.py (so download of metadata for organizations and aggregates which have been setup this way, fail)
• Does not contain all supported metadata which the service desk would like (e.g. escalation contact, organization type, which are in the database, but don't get populated via this procedure)

Make any needed changes so that the service desk's procedure leads to gmoc-db data which contains the fields the Service Desk needs, and which is compatible with gmoc.py upload and download.