Uploaded image for project: 'Red Hat OpenShift Data Science'
  1. Red Hat OpenShift Data Science
  2. RHODS-2160

Uninstall is stuck when RHODS and RHOAM are installed

    XMLWordPrintable

Details

    • 1
    • False
    • False
    • Release Notes
    • Yes
    • 1.7.0-4
    • No
    • Hide
      When OpenShift Data Science and OpenShift API Management are installed together on the same cluster, they use the same Virtual Private Cluster (VPC). The uninstall process for these Add-ons attempts to delete the VPC. Previously, when both Add-ons are installed, the uninstall process for one service was blocked because the other service still had resources in the VPC. The cleanup process has been updated so that this conflict does not occur.
      Show
      When OpenShift Data Science and OpenShift API Management are installed together on the same cluster, they use the same Virtual Private Cluster (VPC). The uninstall process for these Add-ons attempts to delete the VPC. Previously, when both Add-ons are installed, the uninstall process for one service was blocked because the other service still had resources in the VPC. The cleanup process has been updated so that this conflict does not occur.
    • Documented as Resolved Issue
    • No
    • Yes
    • None
    • MODH Sprint 34, MODH Sprint 1.7

    Description

      Description of problem:

      When RHODS and RHOAM are installed together on the same cluster, the uninstall process of one of them can be stuck because CRO fails to delete AWS VPC. 

      The issue has been faced uninstalling RHOAM, but it is reasonable (and i will be tested) that it may happen uninstalling RHODS too. 

      Talking to RHOAM developers, they said the cause should be that the current CRO implementation only support the creation of one CRO network instance for each cluster. So, when you install the first Add-on (e.g., RHODS) it will create the CRO network. When the second Add-on (e.g., RHOAM) is installed, it uses the CRO network created by the first one. When you try to uninstall one of the two Add-on, the CRO fails to delete VPC because there are still resources installed by the other Add-on. 

      We have not been able to test every scenario so far, but the one where the issue came out was the following:

      1. Install RHODS
      2. Install RHOAM
      3. Uninstall RHOAM
      4. Uninstallation is stuck

      Prerequisites (if any, like setup, operators/versions):

      Install RHODS and RHOAM

      Steps to Reproduce

      1. Install one of the two add on
      2. install the second one
      3. uninstall one of them
      4. check if the uninstallation is stuck

      Actual results:

      uninstall blocked because there are resources created by the other Addon in the same VPC

      Expected results:

      uninstallation goes smoothly

      Reproducibility (Always/Intermittent/Only Once):

      Build Details:

      Workaround:

      1. In case of RHOAM is stuck while uninstalling, you can try doing this: oc edit postgres.integreatly.org -n redhat-rhoam-operator and remove the lines related to the finalizers. After that the uninstallation will progress [TESTED]
      2. Delete the cluster from OCM. If there are problems uninstalling addons, It forces the cluster deletion after 1 hour. After that you may want to delete resources from AWS console to clean up.

      Additional info:

      looking at the created resources on AWS, you can see that on a cluster there are two different VPCs: one has the cluster name inside, while the other one is named "RHMI Cloud Resource VPC". Basically the latter is the VPC created by the CRO and this is the one used by both the Addons.

      Attachments

        Issue Links

          Activity

            People

              vhire Vaishnavi Hire
              rhn-support-bdattoma Berto D'Attoma
              Votes:
              0 Vote for this issue
              Watchers:
              12 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: