Uploaded image for project: 'Fabric'
  1. Fabric
  2. FAB-3388

Serviceability - Operational Metrics for Fabric runtime components

    XMLWordPrintable

    Details

    • Epic Name:
      Serviceability - Operational Metrics for Fabric runtime components
    • SDK Impact:
      No
    • Design Status:
      Reviewed
    • Function Test Status:
      Done
    • Documentation Status:
      Done
    • Sample/Tutorial:
      Not Required
    • Current Status:
      Metrics added in v1.4. Gossip, endorser, CA metrics added in v1.4.1 and v2.0.
    • Test Plan:
      Hide
      System Tests:

      * Configure the peer and orderer to send statsd metrics to statsd and use something like graphite to record and graph the associated data.
      - Run a workload.
      - Verify that the metrics make sense for the workload and that the metrics have appropriate names.
      * Configure the peer and orderer to use prometheus for metrics. Set up an instance of prometheus with the appropriate service scrape configuration to collect metrics from the peers and orderers.
      - Run a workload.
      - Verify that the metrics make sense for the workload and that the metrics have appropriate names, labels, and help text.
      Show
      System Tests: * Configure the peer and orderer to send statsd metrics to statsd and use something like graphite to record and graph the associated data. - Run a workload. - Verify that the metrics make sense for the workload and that the metrics have appropriate names. * Configure the peer and orderer to use prometheus for metrics. Set up an instance of prometheus with the appropriate service scrape configuration to collect metrics from the peers and orderers. - Run a workload. - Verify that the metrics make sense for the workload and that the metrics have appropriate names, labels, and help text.
    • Design:
      Show
      Design doc for Fabric peers ( FAB-3389 ): https://docs.google.com/document/d/1xveVQM6e2XeAZBisYRTSfElmeBeYKntnghGyWPtsMSs/edit

      Description

      When Fabric runtime components are deployed in real-world environments, it will be necessary to provide runtime metrics so that operators can actively monitor the health and status of each component.

      Additionally, operators will need a way to integrate these operational metrics into their existing monitoring solutions and/or ask for recommendations for which tools to use.

        Attachments

          Issue Links

            Activity

              People

              Assignee:
              sykesm Matthew Sykes
              Reporter:
              mastersingh24 Gari Singh
              Votes:
              2 Vote for this issue
              Watchers:
              22 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved:

                  Git Integration