Apply Changes operation in a Azure environment shows 500 Interval Server in the Operations Manager UI

Products

• Operations Manager

Issue/Introduction

When running an Apply Changes operation using the Operations Manager (Ops Manager)
UI for an Azure environment, you see the following error message:





The Operations Manager production.log file shows the following stack trace about the 500 Internal Server Error:

E, [2021-05-03T16:29:49.146510 #679] ERROR -- :         Exception caught via rescue_from (code 500):
        (See app/controllers/application_controller.rb)
        undefined method `location' for nil:NilClass
    /home/tempest-web/tempest/web/lib/karl/azure.rb:113:in `block in check_availability_zones'
/home/tempest-web/tempest/web/lib/karl/azure.rb:113:in `select'
/home/tempest-web/tempest/web/lib/karl/azure.rb:113:in `check_availability_zones'
/home/tempest-web/tempest/web/lib/karl/azure.rb:16:in `check_configuration!'
/home/tempest-web/tempest/web/app/models/tempest/verifiers/internal/iaas_configuration_verifier.rb:23:in `block in verify_iaas_configuration'

<LINES-REDACTED>

I, [2021-05-03T16:29:49.146899 #679]  INFO -- : Completed 500 Internal Server Error in 11376ms (Views: 0.2ms | ActiveRecord: 1.5ms)

Cause

From the error message, the piece of code (iaas_configuration_verifier) that invokes calls for Azure API is getting unexpected or missing data when looking up resource SKUs. Ops Manager looks up a resource SKUs information to make sure availability zones are being used. It also uses that information to filter location information. When a resource SKU has an empty locationInfo field in the response, the iaas_configuration_verifier code, as a safety measure, fails and reports the 500 Internal Server Error.

Environment

Product Version: 2.8
OS: Linux

Resolution

The piece of code that was identified to have the behavior described in the Cause section was fixed to better handle situations when an SKU reports empty responses for the locationInfo field. The Operations Manager versions that uses the new piece of code are: 2.7.24, 2.8.14, 2.9.11, and 2.10.2

There are two resolutions that can be implemented to fix this:

• The recommend approach is to upgrade or Install to one of these Operations Manager version's: 2.7.24, 2.8.14, 2.9.11, and 2.10.2. It's recommended because most of the scenarios where this has been reported, was during the initial installation of the Operations Manager.
• To workaround this issue, if you want to stay on a specific version of Operations Manager that does not have a fix for this issue, disable the IaaS Configuration Verifier to proceed with the Apply Changes:
  • Instructions for Disabling Operations Manager Verifiers.
  • Please make sure that you have read the document carefully before disabling the verifier.