Prometheus metrics

[jevonearth]

Counters and histograms for;

• signing operations
• labels
  • vault
  • vault_name
  • op
  • opkind
• errors
  • backend error types (azure, use http codes? 401,403 etc...)

Create a companion grafana dashboard to help operators make the most of these metrics easily
Documentation on use, prometheus quickstart config snippets

[AbineshECAD]

@AbineshECAD Test and close if works as specified.

[jevonearth]

We have a metric in "0.3.0-beta-rc1" names vault_sign_request_duration_milliseconds_bucket

Labels present;

• vault
• vault_name

Let's evaluate switching from histogram to summaries.

TODO:
Metrics for;

• vault errors
• Signatory signing requests
  • labels;
    • signatory response code (2xx, 4xx, 5xx)
    • signatory duration round trip
    • operation type & kind
    • chain_id
    • (possibly) protocol hash (short)

[michaelkernaghan]

superseded more recent tickets. eg #306

[stephengaudet]

signatory response code (2xx, 4xx, 5xx)
--done with #306

signatory duration round trip
--to do, currently only exposing vault time, need time between signatory receipt of request and send of response

operation type & kind
--done

chain_id
-- can be added as a label on existing metrics

(possibly) protocol hash (short)
-- can be added as a label on existing metrics

[jevonearth]

Closing as done.

signatory duration round trip - we don't have a strong need for this metric, so we will remove it from the scope.