Skip hostname validation for monitoring manager

[zeari]

BZ: https://bugzilla.redhat.com/show_bug.cgi?id=1535941

The problem is that the monitoring manager looks for a default endpoint which it does not have.
The parent provider has that endpoint and validates it anyway so no need for the monitoring manager to validate it.

Another solution could be to delegate the default endpoint from the monitoring_manager to the parent provider or just regard the alert endpoint as the default endpoint for monitoring.
Monitoring manager has some prep work for the latter: https://github.com/manageiq/manageiq-providers-kubernetes/blob/master/app/models/manageiq/providers/kubernetes/monitoring_manager_mixin.rb#L59

but ext_management_system doesnt take it into account:
https://github.com/manageiq/manageiq/blob/master/app/models/ext_management_system.rb#L295

@cben @moolitayer Please review

@miq-bot add_label bug, gapridashvili/yes

[miq-bot]

@zeari Cannot apply the following label because they are not recognized: gapridashvili/yes

[zeari]

@miq-bot add_label gaprindashvili/yes

[miq-bot]

Checked commit zeari@d0c4d4e with ruby 2.3.3, rubocop 0.52.0, haml-lint 0.20.0, and yamllint 1.10.0
1 file checked, 0 offenses detected
Everything looks fine. 👍

[cben]

Perfect!

I'm still on the lookout for a cleaner interface to both UI and API then endpoints-with-side-effects. And I still hate the child delegates :endpoints, :to => :parent_manager lie.
But for now, this is great 👍

[cben]

cc @masayag you might want similar fix for kubevirt child manager.

[zeari]

Is it worth to add tests for this to manageiq-api?
I feel we should but this bug wouldnt have been prevented by a test. It happend due to a change in the main repo: ManageIQ/manageiq#16714

[moolitayer]

@zeari could we have caught the breakage in the main repo by having a model spec doing
ExtManagemenySystem.create!(#endpoint in a certain order)

If so, do you think it is worth doing?

[moolitayer]

Verified locally:

Without PR

cat <<EOF | curl -X POST -H "Content-Type: application/json" -u "admin:smartvm" --data @- "localhost:3000/api/providers"
{
"name": "container",
"zone_id": 1,
  "type": "ManageIQ::Providers::Openshift::ContainerManager",
"connection_configurations": [
     {
       "endpoint": {
         "role": "prometheus_alerts",
         "hostname": "alerts.example.com",
         "port": 443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "prometheus_alerts",
         "auth_key": "token"
       }
     },
     {
       "endpoint": {
         "role": "default",
         "hostname": "master.example.com",
         "port": 8443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "bearer",
         "auth_key": "token"
       }
     },
     {
       "endpoint": {
         "role": "prometheus",
         "hostname": "prometheus.example.com",
         "port": 443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "prometheus",
         "auth_key": "token"
       }
     }

   ]
}
EOF
{"error":{"kind":"bad_request","message":"Could not create the new provider - undefined method `ipaddress?' for nil:NilClass","klass":"Api::BadRequestError"}}

With PR:

$ cat <<EOF | curl -X POST -H "Content-Type: application/json" -u "admin:smartvm" --data @- "localhost:3000/api/providers"
{
"name": "container",
"zone_id": 1,
  "type": "ManageIQ::Providers::Openshift::ContainerManager",
"connection_configurations": [
     {
       "endpoint": {
         "role": "prometheus_alerts",
         "hostname": "alerts.example.com",
         "port": 443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "prometheus_alerts",
         "auth_key": "token"
       }
     },
     {
       "endpoint": {
         "role": "default",
         "hostname": "master.example.com",
         "port": 8443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "bearer",
         "auth_key": "token"
       }
     },
     {
       "endpoint": {
         "role": "prometheus",
         "hostname": "prometheus.example.com",
         "port": 443,
         "security_protocol": "ssl-without-validation"
       },
       "authentication": {
         "role": "prometheus",
         "auth_key": "token"
       }
     }

   ]
}
EOF


{"results":[{"href":"http://localhost:3000/api/providers/33","id":"33","name":"container","created_on":"2018-01-21T17:12:08Z","updated_on":"2018-01-21T17:12:08Z","guid":"dc26f7d0-ee13-43ee-8177-67e8b69de432","zone_id":"1","type":"ManageIQ::Providers::Openshift::ContainerManager","api_version":"v1","tenant_id":"1","enabled":true}]}

PS: (had to rebase the PR or I would get "message":"Could not create the new provider - uninitialized constant ManageIQ::Providers::Kubernetes::VirtualizationManagerMixin","klass":"Api::BadRequestError"`)

[moolitayer]

LGTM 👍

[simaishi]

Gaprindashvili backport details:

$ git log -1
commit 2fd717fe62f4319b5cdcd486ec3a42f2d98f9943
Author: Mooli Tayer <mtayer@redhat.com>
Date:   Sun Jan 21 19:18:24 2018 +0200

    Merge pull request #220 from zeari/monitoring_hostname_invalid
    
    Skip hostname validation for monitoring manager
    (cherry picked from commit d64480ed1bb566fe4307131846a08772991054f4)
    
    https://bugzilla.redhat.com/show_bug.cgi?id=1537137