The recommendation get stuck in waiting status
We only allow either Continuous Monitoring job or Recommendation spark job to run at the same time. The Continuous Monitoring job is scheduled to run hourly and checks the status of the previous job before scheduling the next one. It will also delete the old job if it encounters an error while checking the status of the old job. In this customer setup, The status of the continuous monitoring job spark application was not updated by the spark operator for some reason. As a result, there was no error, but the state was empty so old job was not deleted and new Recommendation job could also not be submitted.
This is a known issue affecting NAPP
The permanent fix will be available starting NAPP 4.2
Workaround:
1. use SSH command to login into the NSX manager
2. use napp-k get sparkapp -n nsxi-platform command to get all the current spark applications in the cluster.
3. delete the continuousmonitoringjob spark application by using napp-k delete sparkapp continuousmonitoringjob -n nsxi-platform command
4. make sure the continuousmonitoringjob spark application is deleted and create a new recommendation, it will start running.