* telemetry: tidy the server module after the protobuf bump
Claude-Session: https://claude.ai/code/session_01VGiphDxpsKMwu9XpUFhybQ
* telemetry: keep only clusters that store at least 10 GiB
Fresh weed server runs, CI jobs and throwaway containers each mint their
own cluster id. They came in at tens of thousands a day, were most of
the counted clusters and held almost none of the bytes, and the state
file and the metrics page grew with every one of them. Reports under
the floor are counted and dropped, and a state file written before the
floor sheds them on the first restart.
Claude-Session: https://claude.ai/code/session_01VGiphDxpsKMwu9XpUFhybQ
* master: report telemetry only once the cluster stores 10 GiB
A throwaway cluster no longer registers itself with its first report a
minute after start; a real one begins reporting at the first daily tick
after it crosses the floor.
Claude-Session: https://claude.ai/code/session_01VGiphDxpsKMwu9XpUFhybQ
collectVolumeStats walked only DataNode.GetVolumes(), which returns the
regular volumes on each disk. An encoded volume leaves that set and is
reported through GetEcShards instead, so total_disk_bytes and
total_volume_count silently excluded every erasure-coded volume: a cluster
that encoded everything reported zero bytes and zero volumes while still
counting as a volume server.
Sum each holder's shard sizes into the byte total, parity and extra copies
included, matching how a replicated volume's used size counts every replica
and how CollectionEcVolumeStats already reports EC footprint. Count volume
ids rather than shard entries, since one volume's shards are spread over
many nodes and would otherwise multiply the volume count by the number of
holders.
Chris Lu