Setting Up ManageIQ Spassky for KVM and Hyper-V

ManageIQ is the upstream open-source project behind Red Hat CloudForms and Red Hat Ansible Automation Platform's infrastructure management layer. It gives you a single pane of glass across heterogeneous hypervisors, pulling inventory, capacity metrics, policy compliance, and automation into one interface. If you have KVM hosts managed through oVirt, Hyper-V machines under System Center Virtual Machine Manager (SCVMM), or a mix of both, ManageIQ can monitor and manage them together without requiring you to deploy separate tools for each platform.

This guide covers deploying the ManageIQ Spassky appliance on VMware (the most common bootstrap environment), completing initial configuration through the appliance console, then connecting both KVM/oVirt and Hyper-V/SCVMM infrastructure as providers. The Hyper-V integration requires one extra step at the gem level that catches most people off guard, so we cover that in full.

Requirements

The appliance runs CentOS Stream 9 and ships as a pre-built OVA. Kafka and Zookeeper run inside the appliance itself for the messaging layer, so the RAM requirement is real. Running under 16 GB results in services getting OOM-killed under any load.

VM hardware used in this guide: 8 vCPU, 32 GB RAM, 100 GB OS disk. A second 100 GB disk will be added for the database.

Download and Deploy the OVA

Download the latest ManageIQ appliance from the official releases page:

# OVA download (VMware target)
https://github.com/ManageIQ/manageiq/releases
# Look for: manageiq-vsphere-spassky-N.ova

The OVA deploys to any VMware environment that supports vSphere 6.5+. During deployment, you do not need to configure the network at the OVA level. The appliance console handles networking after first boot.

Before powering on, add a second virtual disk. The appliance console will initialize this disk as a dedicated PostgreSQL volume during setup. Using a separate disk keeps your database I/O isolated from the OS and makes it easier to expand later.

First Boot and Appliance Console

Power on the VM and connect to the console. The appliance boots into CentOS Stream 9 and drops to a login prompt. Log in as root with the default password smartvm. You will be prompted to change the password immediately.

First boot login. Default credentials are root / smartvm. The appliance will force a password change before proceeding.

After changing the password, launch the appliance console menu:

appliance_console

This opens a TUI with a summary screen showing the current state of all subsystems. On a fresh deploy, everything shows "not configured."

Fresh appliance state. Hostname is localhost.localdomain, database and messaging are unconfigured. This is the starting point for all configuration.

Configure Networking

Press any key to get to the main menu, then choose Configure Network. This launches nmtui, which is the standard NetworkManager text UI. Set a static IP, subnet mask, default gateway, and DNS servers.

NetworkManager TUI. Set IPv4 to Manual and enter your static IP, gateway, and DNS. Save and activate the connection before exiting.

After saving, exit nmtui and return to the appliance console menu. The summary should now show your configured IP address. The hostname at this point is still localhost.localdomain. You will set it as part of the messaging configuration step coming up.

Appliance Console: Advanced Settings

Return to the appliance console menu and select Advanced Settings (option 5 in most versions). This exposes a sub-menu with all the major configuration steps.

Advanced Settings menu. The main configuration flow runs in order: Configure Database, Configure Messaging, then Start EVM Server Processes.

Configure the Database

Select Configure Application from the Advanced Settings menu. The wizard walks you through creating a new internal PostgreSQL database.

When it asks whether to use an existing disk, select yes and choose the second virtual disk you added before powering on. On a standard VMware deployment this will be /dev/sdb.

Disk selection for the PostgreSQL database volume. Select the second disk (/dev/sdb). The appliance formats and mounts it automatically.

The wizard then asks for a database region number. This is an integer that uniquely identifies this ManageIQ region. If this is your first deployment, use 01. The region number affects replication topology if you later add remote appliances to the same region.

Each database region number must be unique.
Enter the database region number: 01
Checking for connections to the database...

Create region starting
Create region complete
Database reset successfully

Configure Messaging

Back at the Advanced Settings menu, select Configure Messaging.

Configure Messaging submenu. For a standalone deployment, choose "Configure this appliance as a messaging server."

Choose to configure this appliance as its own messaging server. The wizard prompts for the hostname, a keystore username, and a keystore password. Enter the fully qualified hostname of this appliance for the hostname field.

Message server parameter entry. Use the appliance's FQDN as the hostname. The keystore username defaults to admin. Use a strong password for the keystore.

The console will verify that the hostname resolves and is reachable, then configure Kafka and Zookeeper. This takes about 30 to 60 seconds. When complete, you will see:

Message Server configured successfully.
Starting ManageIQ Server...

Start EVM Server Processes

After the messaging configuration completes, return to the Advanced Settings menu and choose Start EVM Server Processes. The appliance starts the ManageIQ Rails application stack, worker processes, and supporting services. This takes several minutes on first start.

The appliance summary screen will eventually show all key services as running. You can also monitor startup progress from the Cockpit web console at https://<yourip>:9090 or watch the EVM log:

journalctl -fu evmserverd

Access the ManageIQ Web Interface

Once EVM server processes are running, browse to https://<yourip>. The default credentials are:

ManageIQ login page. Region 1, Zone: default, Appliance: EVM. Change the admin password immediately after first login.

Configure Server Roles

Navigate to Settings > Configuration > Settings > Server (gear icon at top right). This page controls which roles are active on this EVM server instance. For a standalone single-appliance deployment, most roles should be on. The roles that matter for provider connectivity are:

Server Roles configuration. Provider Inventory, Provider Operations, Event Monitor, and both Capacity and Utilization roles must be on to get full functionality from connected providers.

Enable the roles you need and click Save. Role changes take effect at the next EVM server heartbeat, which is approximately 60 seconds.

Connecting KVM (oVirt / RHV) Infrastructure

ManageIQ connects to KVM infrastructure through the oVirt provider type. This requires oVirt Engine or Red Hat Virtualization Manager running and accessible. If your KVM hosts are managed directly without oVirt, use the Red Hat Virtualization provider type and point it at the Manager URL.

Navigate to Compute > Infrastructure > Providers, then click Configuration > Add a New Infrastructure Provider.

Provider type dropdown. For KVM infrastructure, select oVirt. Red Hat Virtualization is an alias for the same provider type and connects to the same oVirt Manager API endpoint.

Fill in the provider form:

Click Validate before saving. ManageIQ will test the connection and return a success or error message. Common errors at this stage are certificate verification failures (switch to "Do not verify" to test) and credentials with insufficient permissions. The oVirt service account needs at minimum the UserRole role at the data center level to enumerate all clusters and hosts.

Once validation passes, click Add. ManageIQ will save the provider and trigger an immediate inventory refresh. After a few minutes, navigate to Compute > Infrastructure > Hosts to confirm hosts are appearing.

Connecting Hyper-V (SCVMM) Infrastructure

Hyper-V integration in ManageIQ uses Microsoft's SCVMM provider. Unlike the other provider types, the Microsoft provider gem is not included in the default ManageIQ appliance bundle. You need to install it manually before the provider type becomes available in the UI.

Step 1: Enable the Microsoft Provider Gem

SSH into the ManageIQ appliance and run the following commands. These add the Microsoft provider gem to the bundle and restart the EVM service to load the new code.

# Navigate to the ManageIQ application directory
cd /var/www/miq/vmdb

# Add the Microsoft provider gem to the Gemfile
echo "gem 'manageiq-providers-microsoft', :group => [:microsoft]" >> Gemfile.dev.rb

# Install the new gem (takes 1-3 minutes)
bundle install --with microsoft

# Restart ManageIQ services to load the new provider code
systemctl restart evmserverd

After the restart, wait two to three minutes for EVM to fully initialize, then refresh the Add New Infrastructure Provider form. You should now see Microsoft SCVMM as a provider type option.

Step 2: Configure the SCVMM Provider

Navigate again to Compute > Infrastructure > Providers > Add a New Infrastructure Provider and select Microsoft SCVMM.

ManageIQ communicates with SCVMM over WinRM (Windows Remote Management). The service account must have the SCVMM Administrator role, not just Delegated Administrator. Delegated Administrator lacks the PowerShell runspace permissions that ManageIQ requires for inventory collection.

Enable-PSRemoting -Force
winrm set winrm/config/listener?Address=*+Transport=HTTP '@{Port="5985"}'

Validate the connection, then click Add. The initial SCVMM inventory refresh collects hosts, VMs, templates, and logical networks. Depending on the size of your environment, the first refresh can take 10 to 30 minutes.

Verifying Provider Connectivity

After adding both providers, navigate to Compute > Infrastructure > Providers to see the unified list. Each provider entry shows the discovered host count, vCore count, and total memory. If these are all zero after 10 minutes, click the provider name and check the provider timeline for error events.

Common post-add issues:

To force a full re-inventory, select the provider in the list, click Configuration > Refresh Relationships and Power States. This queues an immediate inventory task regardless of the scheduled interval.

Enabling Capacity and Utilization Collection

By default, ManageIQ collects inventory (hosts, VMs, clusters) but not real-time metrics. To enable C&U data collection, you need to configure the provider to capture metrics and ensure the C&U server roles are active.

For oVirt, navigate to the provider record and click Configuration > Edit this Infrastructure Provider. On the C&U tab, enable collection and set the collection interval. oVirt exposes metrics through its own REST API; ManageIQ polls this directly.

For SCVMM/Hyper-V, performance data collection uses Windows Performance Counters over WinRM. The SCVMM service account needs read access to performance counter data on each Hyper-V host, not just on the SCVMM server itself. If C&U shows no data for Hyper-V hosts, check that the service account is in the local Performance Monitor Users group on the Hyper-V host machines.