- Rates continue to look good
- Failures that we do see are consistent with expectations
- Graphs look useful: https://gracc.opensciencegrid.org/dashboard/db/perfsonar-collector?orgId=1
- Interval is every 5 minutes (polling 10-12x per hour)
- Discuss what analytics are needed at a UChicago meeting?
- Action items:
- Don’t send summary info to bus - Done (Derek)
- Project organization: moved to Docker Hub - Done (Derek)
- Goal is to have all the SAND pieces show up in the OSG check_mk
- How to separate high-level vs low-level pieces
- When is a service like GRACC “up” or “down”
- Business intelligence?
- Make a “poor-man’s” version with a clever probe?
- Service status format changes (Huijun)
- Two tables:
- Table of SLA services vs non-SLA services?
- Comma-separated list of those meeting the SLAs
- All the services in a table sorted by uptime
- Pick the four most important services and have them captured by check_mk
- Link presentations from the SAND website (Shawn)
- Find a week in May that would work for a Chicago meeting (Derek)