Uploaded image for project: 'VOLTHA'
  1. VOLTHA
  2. VOL-4507

Test Disable and Enable ONU for DT reliably fails

    XMLWordPrintable

    Details

      Description

      I am looking at an issue where ONU does not move to omci-flows-pushed state after ONU reboot. The subscriber was provisioned for the ONU when it was rebooted. This happens on the XGSPON OLTs in test Test Disable and Enable ONU for DT . The ONU is disabled->enabled->rebooted. At each stage ONU states and port states at ONOS are checked. After ONU reboot the omci state is supposed to move to omci-flows-pushed state following flow being pushed down for the subscriber.
      When the port went down due to reboot I see this at ONOS
      infra/voltha-infra-onos-classic-0[onos-classic]: 15:15:32.076 INFO [McastHandler] Processing port down of:0000a82bb536786c/655616
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:32.129 INFO [Olt] Adding subscriber to queue: of:0000a82bb536786c/655616[ISKT71e80080-1] with status REMOVED and subscriber true
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:32.206 DEBUG [OltFlowService] Not dealing with DP flows remove on as subscriber flow status is false
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:32.206 INFO [OltFlowService] Removal of subscriber on of:0000a82bb536786c/655616[ISKT71e80080-1] completed
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:32.207 DEBUG [Olt] Subscriber of:0000a82bb536786c/655616[ISKT71e80080-1] (status: REMOVED, provisionSubscriber: true) has been removed from the queue
      When the port was enabled again I see this at ONOS
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.062 INFO [Olt] Adding subscriber to queue: of:0000a82bb536786c/655616[ISKT71e80080-1] with status ADDED and subscriber true
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.365 DEBUG [OltMeterService] Meter found for BandwidthProfile Default on device of:0000a82bb536786c
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.366 DEBUG [OltMeterService] Meter found for BandwidthProfile Default on device of:0000a82bb536786c
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.367 DEBUG [OltMeterService] Meter found for BandwidthProfile Default on device of:0000a82bb536786c
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.367 DEBUG [OltMeterService] Meter found for BandwidthProfile Default on device of:0000a82bb536786c
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.368 DEBUG [OltFlowService] Not dealing with DP flows add on as subscriber flow status is true
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.369 INFO [OltFlowService] Provisioning of subscriber on of:0000a82bb536786c/655616[ISKT71e80080-1] completed
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.369 DEBUG [Olt] Subscriber of:0000a82bb536786c/655616[ISKT71e80080-1] (status: ADDED, provisionSubscriber: true) has been removed from the queue
      I see this in this job https://jenkins.opencord.org/view/VOLTHA-2.X-Tests/job/build_onf-demo-pod_1T8GEM_voltha_DT_master_test/647. Hardik mentioned this is also seen in Soak pod and that is quite possible as we XGSPON OLT there as well.


      The surprising part is that the app logs:
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:32.206 DEBUG [OltFlowService] Not dealing with DP flows remove on as subscriber flow status is false
      and then
      infra/voltha-infra-onos-classic-1[onos-classic]: 15:15:56.368 DEBUG [OltFlowService] Not dealing with DP flows add on as subscriber flow status is true
      as the first indicates that flow removal is not needed as flows are not present, while the second indicates that flow installation is not needed as they are already present.One thing that I noticed is that the log message is:
      log.debug("Not dealing with DP flows {} on {} as subscriber flow status is {}", action, uti.getServiceName(), hasFlows);
      and the serviceName is not present in the sadis config, while it's present in the one expose via BBSim... in theory that should be fine, but I'll double-check later today

      This issue is very consistent with the DT job
       

        Attachments

        # Subject Branch Project Status CR V

          Activity

            People

            Assignee:
            teone Matteo Scandolo
            Reporter:
            ggowdra Girish Gowdra
            Watchers:
            1 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved:

                Gerrit Reviews

                There are no open Gerrit changes