Bug 2074659

Provides False negative when searching for DNS Records. Empty String will always appear.

Hello,

I found this issue while working on a Jira Card that involved testing a function called ValidateForProvisioning for AWS. The install config was using this value. There was not currently a test for this functionality, but the way that you could hit this code was through a platform provisioning check for AWS. The zone name was always blank when trying to validate the zones.

Is this enough to go off of?

Hello Brent, I tried following tests:

install-config:
<--SNIP-->
baseDomain: qe.devcluster.openshift.com
metadata:
  creationTimestamp: null
  name: yunjiang-aaaccc
platform:
  aws:
    region: us-east-2
    subnets:
    - subnet-034f5beb385a151c3
    - subnet-0f1f876a97ae76fd2
    - subnet-0219becdbf9d5519f
    - subnet-0979c17619fc8387e
    hostedZone: Z07770582ZWZBOWLV6AQJ <- private hosed zone: yunjiang-050953a.qe.devcluster.openshift.com
publish: External
<--SNIP-->

before applied fix (4.10):
FATAL failed to fetch Terraform Variables: failed to fetch dependency of "Terraform Variables": failed to generate asset "Platform Provisioning Check": aws.hostedZone: Invalid value: "Z07770582ZWZBOWLV6AQJ": hosted zone domain "yunjiang-050953a.qe.devcluster.openshift.com." is not a parent of the cluster domain "yunjiang-aaaccc.qe.devcluster.openshift.com."

after applied fix (4.11.0-0.nightly-2022-05-08-193101):
FATAL failed to fetch Terraform Variables: failed to fetch dependency of "Terraform Variables": failed to generate asset "Platform Provisioning Check": aws.hostedZone: Internal error: could not list record sets for domain "Z07770582ZWZBOWLV6AQJ": hosted zone domain "yunjiang-050953a.qe.devcluster.openshift.com." is not a parent of the cluster domain "yunjiang-aaaccc.qe.devcluster.openshift.com."

it looks like the installer of 4.11 went into code [1], but 4.10 did not [2]

[1] https://github.com/openshift/installer/blob/release-4.11/pkg/asset/installconfig/aws/validation.go#L395
[2] https://github.com/openshift/installer/blob/release-4.10/pkg/asset/installconfig/aws/validation.go#L398


I have some questions:
1. per 4.11 log message, the zoneName is Z07770582ZWZBOWLV6AQJ, is the expected?
2. error message `could not list record sets` was not shown in 4.10, is it ，is this related to the current modifications?

1. per 4.11 log message, the zoneName is Z07770582ZWZBOWLV6AQJ, is the expected?

>>> yes, I believe that is the expected zoneName

2. error message `could not list record sets` was not shown in 4.10, is it ，is this related to the current modifications?

>>> That message would only appear on an error from getSubDomainDNSRecords. The "error" or indication of false data appears
when an error occurs in getSubDomainDNSRecords because the zoneName was not set correctly when passed to validateZoneRecords

Ok, I was able to produce the error with an install-config. 

apiVersion: v1
baseDomain: bbarbach-public-zone.com
compute:
- architecture: amd64
  hyperthreading: Enabled
  name: worker
  platform: {}
  replicas: 3
controlPlane:
  architecture: amd64
  hyperthreading: Enabled
  name: master
  platform: {}
  replicas: 3
metadata:
  creationTimestamp: null
  name: bbarbach-zone-test
networking:
  clusterNetwork:
  - cidr: 10.128.0.0/14
    hostPrefix: 23
  machineNetwork:
  - cidr: 10.0.0.0/16
  networkType: OpenShiftSDN
  serviceNetwork:
  - 172.30.0.0/16
platform:
  aws:
    region: us-east-1
    subnets:
    - subnet-0412605e3e4a5e892
    - subnet-088e713b919753182
    hostedZone: Z0909126N1BWS293HVRS
publish: External


The hosted zone and subnets are ones that I created. When running this install-config with create cluster I get the following results:

FATAL failed to fetch Terraform Variables: failed to fetch dependency of "Terraform Variables": failed to generate asset "Platform Provisioning Check": <nil>: Internal error: could not list record sets for domain "": hosted zone domain "bbarbach-pvc.com." is not a parent of the cluster domain "bbarbach-zone-test.bbarbach-public-zone.com." 

The error message:
Internal error: could not list record sets for domain ""

should be filled with the correct hostedZone not an empty string. No matter what hostedZone name is there it would produce this result as it was not set when passed to the function.

(In reply to Brent Barbachem from comment #8)
> Ok, I was able to produce the error with an install-config. 
> 
> apiVersion: v1
> baseDomain: bbarbach-public-zone.com
> compute:
> - architecture: amd64
>   hyperthreading: Enabled
>   name: worker
>   platform: {}
>   replicas: 3
> controlPlane:
>   architecture: amd64
>   hyperthreading: Enabled
>   name: master
>   platform: {}
>   replicas: 3
> metadata:
>   creationTimestamp: null
>   name: bbarbach-zone-test
> networking:
>   clusterNetwork:
>   - cidr: 10.128.0.0/14
>     hostPrefix: 23
>   machineNetwork:
>   - cidr: 10.0.0.0/16
>   networkType: OpenShiftSDN
>   serviceNetwork:
>   - 172.30.0.0/16
> platform:
>   aws:
>     region: us-east-1
>     subnets:
>     - subnet-0412605e3e4a5e892
>     - subnet-088e713b919753182
>     hostedZone: Z0909126N1BWS293HVRS
> publish: External
> 
> 
> The hosted zone and subnets are ones that I created. When running this
> install-config with create cluster I get the following results:
> 
> FATAL failed to fetch Terraform Variables: failed to fetch dependency of
> "Terraform Variables": failed to generate asset "Platform Provisioning
> Check": <nil>: Internal error: could not list record sets for domain "":
> hosted zone domain "bbarbach-pvc.com." is not a parent of the cluster domain
> "bbarbach-zone-test.bbarbach-public-zone.com." 
> 
> The error message:
> Internal error: could not list record sets for domain ""
> 
> should be filled with the correct hostedZone not an empty string. No matter
> what hostedZone name is there it would produce this result as it was not set
> when passed to the function.

This was tested against version 4.10 as 4.11 has the issue solved I believe.

Thanks Brent, I can hit the issue while setting "hosted zone domain is not a parent of the cluster domain".

> OCP 4.11.0-0.nightly-2022-05-20-213928

FATAL failed to fetch Terraform Variables: failed to fetch dependency of "Terraform Variables": failed to generate asset "Platform Provisioning Check": aws.hostedZone: Internal error: could not list record sets for domain "Z0832957144NMU2JF1TBH": hosted zone domain "yunjiang-0525r1.qe.devcluster.openshift.com." is not a parent of the cluster domain "yunjiang-0525rb.qe1.devcluster.openshift.com."

> OCP 4.10.12

FATAL failed to fetch Terraform Variables: failed to fetch dependency of "Terraform Variables": failed to generate asset "Platform Provisioning Check": <nil>: Internal error: could not list record sets for domain "": hosted zone domain "yunjiang-0525r1.qe.devcluster.openshift.com." is not a parent of the cluster domain "yunjiang-0525r2.qe1.devcluster.openshift.com."

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Important: OpenShift Container Platform 4.11.0 bug fix and security update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2022:5069