In Newton, while replacing controller node with ID 0 : StackValidationFailed: resources.AllNodesValidationConfig: Property error: AllNodesValidationConfig.Properties.PingTestIps: The Referenced Attribute (Controller resource.0.external_ip_address) is incorrect.

[Luca Lorenzetto]

While replacing the controller node with id 0 stack validation fails with error:

StackValidationFailed: resources.AllNodesValidationConfig: Property error: AllNodesValidationConfig.Properties.PingTestIps: The Referenced Attribute (Controller resource.0.external_ip_address) is incorrect.

This because in overcloud.j2.yaml there is this explicit reference to node 0:

https://github.com/openstack/tripleo-heat-templates/blob/stable/newton/overcloud.j2.yaml#L500

Same reference is done also in Ocata, so i suppose also this release is affected (not tested). Maybe the node uses for validation should be selected looking at existing nodes.

In my case node 0 is no more existing since is object of replacement, so using 1 could be a good option. I suppose then that since node 0 will no more exist, all future validations will fail.

In RH documentation, up to Mitaka there were an indication saying that replace resource.0 with a valid ID was required.

https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/9/html/director_installation_and_usage/sect-scaling_the_overcloud#sect-Replacing_Controller_Nodes-Node_Replacement

[Luca Lorenzetto (lorenzetto-luca)]

At the moment replaced resource.0 with resource.3, since is a valid node that will persist.

[Steven Hardy (shardy)]

I think this is because of this hard-coded reference here:

https://github.com/openstack/tripleo-heat-templates/blob/stable/newton/overcloud.j2.yaml#L506

Probably we should do this similar to:

https://github.com/openstack/tripleo-heat-templates/blob/master/overcloud.j2.yaml#L299

Which does a yaql lookup to get the first item in the list not the node named "0"

[Steven Hardy (shardy)]

To clarify we need a change like:

  # All Nodes Validations
  AllNodesValidationConfig:
    type: OS::TripleO::AllNodes::Validation
    properties:
      PingTestIps:
        list_join:
        - ' '
        - - {get_attr: [{{primary_role_name}}, resource.0.external_ip_address]}
          - {get_attr: [{{primary_role_name}}, resource.0.internal_api_ip_address]}
          - {get_attr: [{{primary_role_name}}, resource.0.storage_ip_address]}
          - {get_attr: [{{primary_role_name}}, resource.0.storage_mgmt_ip_address]}
          - {get_attr: [{{primary_role_name}}, resource.0.tenant_ip_address]}
          - {get_attr: [{{primary_role_name}}, resource.0.management_ip_address]}

Should be:
  # All Nodes Validations
  AllNodesValidationConfig:
    type: OS::TripleO::AllNodes::Validation
    properties:
      PingTestIps:
        list_join:
        - ' '
        - - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, external_ip_address]}
          - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, internal_api_ip_address]}
          - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, storage_ip_address]}
          - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, storage_mgmt_ip_address]}
          - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, tenant_ip_address]}
          - yaql:
              expression: coalesce($.data, []).first(null)
              data: {get_attr: [{{primary_role_name}}, management_ip_address]}

Untested - let me know if you'd prefer me to submit the patch, or feel free to copy the above, test and push it to gerrit yourself :)

[Luca Lorenzetto (lorenzetto-luca)]

Thanks shardy, i'll test by myself and then report.

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/446887

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/ocata
Review: https://review.openstack.org/447416

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: stable/newton
Review: https://review.openstack.org/447419

[OpenStack Infra (hudson-openstack)]

Change abandoned by Emilien Macchi (<email address hidden>) on branch: stable/newton
Review: https://review.openstack.org/447419

[OpenStack Infra (hudson-openstack)]

Change abandoned by Emilien Macchi (<email address hidden>) on branch: stable/ocata
Review: https://review.openstack.org/447416

[OpenStack Infra (hudson-openstack)]

Change abandoned by Emilien Macchi (<email address hidden>) on branch: stable/newton
Review: https://review.openstack.org/447419

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/447419
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=9106ddee3276f0fa0c5d196ded2fa20fb9cd950d
Submitter: Jenkins
Branch: stable/newton

commit 9106ddee3276f0fa0c5d196ded2fa20fb9cd950d
Author: Luca Lorenzetto <email address hidden>
Date: Fri Mar 17 09:29:41 2017 +0100

    Pick dynamically the first node for stack validation

    When replacing the controller node with resource id 0,
    AllNodesValidation will fail because there is an hardcoded reference
    to resource.0. With this commit the id for validation is extracted
    dynamically with yaql query, picking the first available.
    Thanks to Steven Hardy for pointing to the right direction.

    Change-Id: I8f2bfacbc005d948bd31ebd51c3d3df3182d5a3c
    Closes-Bug: #1673439
    (cherry picked from commit d385fc38b22b204414e04ba1962d7c6add262f1d)

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/447416
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=d385fc38b22b204414e04ba1962d7c6add262f1d
Submitter: Jenkins
Branch: stable/ocata

commit d385fc38b22b204414e04ba1962d7c6add262f1d
Author: Luca Lorenzetto <email address hidden>
Date: Fri Mar 17 09:29:41 2017 +0100

    Pick dynamically the first node for stack validation

    When replacing the controller node with resource id 0,
    AllNodesValidation will fail because there is an hardcoded reference
    to resource.0. With this commit the id for validation is extracted
    dynamically with yaql query, picking the first available.
    Thanks to Steven Hardy for pointing to the right direction.

    Change-Id: I8f2bfacbc005d948bd31ebd51c3d3df3182d5a3c
    Closes-Bug: #1673439

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/446887
Committed: https://git.openstack.org/cgit/openstack/tripleo-heat-templates/commit/?id=fd15a091f7ab6927833275df17b96ecacc2b1827
Submitter: Jenkins
Branch: master

commit fd15a091f7ab6927833275df17b96ecacc2b1827
Author: Luca Lorenzetto <email address hidden>
Date: Fri Mar 17 09:29:41 2017 +0100

    Pick dynamically the first node for stack validation

    When replacing the controller node with resource id 0,
    AllNodesValidation will fail because there is an hardcoded reference
    to resource.0. With this commit the id for validation is extracted
    dynamically with yaql query, picking the first available.
    Thanks to Steven Hardy for pointing to the right direction.

    Change-Id: I8f2bfacbc005d948bd31ebd51c3d3df3182d5a3c
    Closes-Bug: #1673439

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tripleo-heat-templates 7.0.0.0b1 development milestone.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tripleo-heat-templates 5.3.0 release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/tripleo-heat-templates 6.1.0 release.