Uploaded image for project: 'ONAP Operations Manager'
  1. ONAP Operations Manager
  2. OOM-496

Readiness check is marking full availability of some components like SDC and SDNC before they would pass healthcheck

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: High High
    • Beijing Release
    • Beijing Release
    • None
    • None
    • R3 Sprint 1

      We need to verify that the readiness probe is the full REST level readiness (either on our end or pushing for better /heathcheck endpoints on the components)

      for example if i run robot healthcheck too early I get random failures on SDNC or SDC - If I wait 4 min after the last pod up (usually aai-service) and do 3 health checks to warm the system - the 3rd is better but SDC still fails 35% of the time (uncovered this by using shanedaniel s new dashboard

      see SDC-739

      and the short analysis on 

      https://wiki.onap.org/display/DW/Auto+Continuous+Deployment+via+Jenkins+and+Kibana#AutoContinuousDeploymentviaJenkinsandKibana-KibanaDashboardofCDsystemdiagnosinghealthcheckissuesinanHourlyONAPOOMDeploy

       

       

            atulan atulan
            michaelobrien michaelobrien
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated:
              Resolved: