Planning to release tomorrow an initial version of the plan for consolidated problem-reporting
Have dealt with a variety of filesystem-full and ran-out-of-inodes problems. Once recent case of large-inode-count was both reported and triggered by Igor Gaponenko; Unknown User (pdomagala) asked for details. Plan to discuss this in more detail at the infrastructure meeting.
Currently no monitoring on Nebula, need to do something about that. May use that as a chance to compare using Nagios to some newer alternatives (e.g. Grafana, InfluxDB)
Unknown User (pdomagala) will arrange for a report on near-term plans for monitoring at next week's infrastructure meeting.
Fritz Mueller asks that SUIT post some specifics of what they would like to have monitored in their systems
Fritz Mueller wonders if the alternatives being considered can match the real-time latency of Ganglia?
Technology decisions to be made by Michelle Butler