Post Upgrade to 1.25.1, some Prometheus/Grafana metrics not showing

For my main dashboard

https://charity-pool.ro/new_configuration_1.25.1

But I am using cntools.

2 Likes

No i mean your node config

1 Like

Hi. i got this working, the fix is to enable all these Trace configs.

It was one of

  "TraceBlockFetchClient": true,
  "TraceBlockFetchDecisions": true,
  "TraceBlockFetchProtocol": true,
  "TraceBlockFetchProtocolSerialised": true,
  "TraceBlockFetchServer": true,
4 Likes

Great, l’m glad u fixed :wink:

Just on this - is there a dictionary of definitions/descriptions of each metric published anywhere?

Legend thankyou!

1 Like

FYI, this will not work if “minSeverity” is set to anything higher than “Info” in the node config. I had mine set to “Error” and couldn’t make my node expose cardano_node_metrics_connectedPeers_int until I changed “minSeverity” back to “Info”.

Apparently this has been going on since 1.18:

1 Like

The needed tracer for cardano_node_metrics_connectedPeers_int is TraceBlockFetchDecisions

I don’t really understand haskell but i think the information is included here:

1 Like

cardano_node_metrics_connectedPeers_int is only available if you modify your mainnet-config.json file to include:

“TraceBlockFetchClient”: true,
“TraceBlockFetchDecisions”: true,

Does anyone have any documentation as to what metrics all the other “TraceBlockFetch” parameters in the config.json file control? I have activated quite a few of them just to check and all I get is a bloated syslog that needs to be deleted regularly, but no other useful/interesting Prometheus metrics, it seems.

4 Likes