Metrics that can be learned from network observations are available
Display number of nodes discovered over discv5, by shard
Display number of nodes successfully connected to, by shard and user agent, in last hour, day, week
Display number of light protocol clients fleet nodes had a inbound connection of, by libp2p protocol, user agent, libp2p transport, and shard, in last hour, day, week; using unique peer id as identifier
Number of messages unique messages seen, by shard
Inbound and outbound bandwidth, by shard
Number of different content topics in last hour, day and week; considering full content topic and application name.
Average and max message size by shard
Messages stored by fleet store nodes: total size, number, oldest timestamp, by shard
Usability
Metrics above are available publicly for the network of major integrations (Status, RAILGUN, TWN).
Reliability
Metrics should be available 90% of the time.
Performance
The data is updated at least hourly
At least 3 months of data is available
Supportability
Grafana dashboards
Some metrics may be retrieved by a Waku monitor node with aggressive parameters, other from existing fleet nodes;
this includes running limited fleet nodes for other networks.