VCAP-CMA Deploy – Objective 4.2/4.3

Disclaimer: These are my notes from studying for the 3V0-31.18 exam. If something doesn’t make sense, please feel free to reach out.

The main goal for the whole of section 4 is to know where to start with troubleshooting various aspects of vRealize Automation.

Objective 4.2 – Troubleshoot Tenant Operations

References

What kind of issues might you run into with respect to tenant operations?

  • Users without access to the correct role
  • Approval policies not firing
  • Creating new tenants

Following my blog post from the VCP-CMA, I’d want to be very familiar with what access a single role grants. What does an IaaS administrator get different to a tenant administrator etc.

To check on an approval policy, first I’d double check it was linked to the correct item within the entitlement.

Approval_Pol_Assigned

Then I’d just sanity check any conditions you’ve attached to the policy. I often see the less than/greater than the wrong way round.

When creating new tenants there are a number of items that will need to be created for it to be useful:

  • New directory connector to AD
  • New business groups
  • New reservations
  • Roles will need (re)allocating

Exactly the same as creating them for the first tenant.

Objective 4.3 – Troubleshoot Provisioning Issues

In addition to the references at the top of the post, I want to point out this awesome post over at orchestration.io on troubleshooting provisioning – https://orchestration.io/2017/05/01/troubleshooting-vrealize-automation-provisioning/

Well worth a read if you want in depth detail on where to look for any issues with provisioning.

Before you get to that level of trawling logs, it’s worth checking on the reservations (as mentioned in the previous post):

  • Does the reservation have enough free capacity to accommodate the request?
  • Does the reservation have access to the capabilities required by the blueprint?

It’s also always worth checking the endpoint is configured correctly and that data collection is occurring.

That covers the IaaS requests, but what if an XaaS request is failing?

  • Is vRO running?
  • Did the workflow start?
  • Is there any logs on the failed workflow?

Check within vRO, under the task to see the instances that have run, check the date. The icon next to the runtime will tell you if it’s been successful or otherwise. If you select the run itself, you can then browse the logs to find out what’s happening during execution.

vRO_Success_Workflow

I’d expect you to need to be able to read enough Javascript to understand what the workflow is doing. Nothing more than a basic understanding.

 

VCAP-CMA Deploy – Objective 4.1

Disclaimer: These are my notes from studying for the 3V0-31.18 exam. If something doesn’t make sense, please feel free to reach out.

The main goal for the whole of section 4 is to know where to start with troubleshooting various aspects of vRealize Automation.

Objective 4.1 – Troubleshoot Infrastructure Issues

References

I’m going to break this down into user troubleshooting (think permissions, blueprints, business group etc.) & compute troubleshooting (Endpoints, SQL, IaaS etc.). For both of these sections you’ll need to know where to find what the issue is and how to interpret that information. I suspect this will be fixing stuff that was working and now isn’t, rather than a failed deployment. Although it could be either.

Infrastructure – Compute

For infrastructure issues think along these lines:

  • Endpoints
  • DEM Orchestrator
  • DEM Agent
  • SQL

When checking for issues, the first place I go to is the Log within the vRA console –¬†Infrastructure, Monitoring, Log. This is the place where you would see if vRA can’t connect to the endpoint, and likely why. The error below is because the vCenter endpoint was offline.

vRA_Log_Example

While you’re in the monitoring section, the other useful tab is DEM Status. The error below shows that I’ve stopped the DEM service on the IaaS server.

vRA_DEM_Status

This can be verified from the vRA appliance VAMI on the Services tab. This is a good place to check for any services that aren’t responding.

vRA_VAMI_Services

While we’re in this console, it’s worth having a look on the Cluster tab. This will show you if any servers in the vRA installation haven’t contacted vRA in a while. Also on the cluster tab you can generate & download the support bundle.

Once these have all been checked, the remaining place is the IaaS node. Login and check the Windows services & IIS AppPools have started. If everything is running, check Event Viewer, this tends to be where I find if the SQL database is available or not. Or if there is generally anything wrong with the Windows box. It’s also worth running through the list of pre-requisites (the automated check when installing). These could have changed through the application of a GPO or even manually.

Users

For user type issues think along these lines:

  • User role
  • Entitlements
  • Reservations
  • Blueprints

If the user needs to perform either an administrative function or a design function but isn’t able to, this is often down to the role of the user. Do they have the relevant tab available? The tabs each role should have available is detailed on my blog post from the VCP7-CMA, linked at the top.

If the user is attempting to provision an item but it isn’t available in the catalogue you might want to think of these:

  • Are they a member of the correct business group?
  • Is the blueprint published?
  • Is the blueprint part of a service?
  • Is the service and/or blueprint entitled correctly?

If a user is a member of multiple business groups, the business group the catalogue item is assigned from is shown in the catalogue.

Catalog_BG

If provisioning is failing, the request will usually give you a reason why. This can quite often be something to do with the reservation. Such as:

  • Reservation capacity
    • Is there any remaining in the allocated quota?
  • Reservation capability
    • Can the reservation support the infrastructure requirements of the blueprint?
      • Quote often network related (for me anyway!)

Provisioning_Request_Error.png