ClusterIQ v2.0: From Inventory Tool to Automation Platform — and It’s Kind of a Big Deal
So I’ve been minding my own business managing Hyper-V and Azure Local clusters the usual way — a mix of PowerShell one-liners, Failover Cluster Manager, Windows Admin Center, and the occasional prayer — when I realized the gap in my toolbelt wasn’t visibility. I had plenty of visibility. What I didn’t have was a single place to act on what I was seeing, without writing yet another maintenance script at 11pm on a Thursday.
That’s the gap ClusterIQ v2.0 is going after. And honestly? This release is a doooooozy.
—
What ClusterIQ Actually Is (Quick Primer)
If you haven’t seen it before: ClusterIQ is a Windows GUI app built specifically for Hyper-V and Azure Local cluster operations. It’s not a portal replacement, it’s not trying to be Windows Admin Center, and it’s definitely not another monitoring dashboard you’ll set up and forget. It’s an ops tool — the kind of thing you actually have open on a second monitor while you’re doing real work.
Version 1.x was solid for inventory and real-time visibility: cluster node health, VM state, live migration, S2D storage insights, performance graphs. Useful stuff. But v2.0 is a different animal. The jump from v1 to v2 is essentially the jump from “I can see what’s happening” to “I can do something about it — automatically, on a schedule, with a full audit trail.”
Azure Local VMs are designed to be managed through the Azure control plane, but on-premises tools still offer a broader range of capabilities — including Failover Cluster Manager, Hyper-V Manager, and Windows Admin Center. ClusterIQ slots into that on-premises tooling layer and fills in the operational automation gap those tools don’t really cover.
—
The Big New Thing: Autopilot Playbooks
The headline feature in v2.0 is hvAutopilot — a playbook engine that lets you chain maintenance steps together, schedule them, and run them headlessly. We’re talking things like snapshot cleanup, patching readiness checks, live-migrate-to-balance, orphaned VHD detection, and Integration Services verification. You build a playbook from a step library, set a schedule (daily or weekly), and walk away.
Every run produces a per-step audit trail and auto-opens a Markdown summary. That last part matters more than it sounds — when something goes sideways at 2am, you want a readable summary waiting for you, not a wall of raw log output.
There’s also a Preview Mode before anything executes, which shows you the actual PowerShell that will run. No black boxes. I appreciate that a lot. The number of “automation” tools that hide what they’re actually doing on your cluster is… too many.
The Headless Service
This is the part that makes the playbook engine actually useful in production. ClusterIQ ships a ClusterIQ.Headless.exe that runs as a Windows service, executing scheduled playbooks even when the GUI is closed. The install is one line from an elevated prompt:
# Install the headless service — run this from an elevated prompt on your management machine
ClusterIQ.Headless.exe --install
# When you're done with it
ClusterIQ.Headless.exe --uninstallThe GUI detects whether the headless service is running and yields scheduling control to it automatically. That’s a thoughtful design choice — you don’t end up with two things fighting over the same schedule. There are five total CLI modes documented in docs/HEADLESS.md, including connection setup and service account configuration. Worth reading before you deploy this in production.
One gotcha worth flagging: the service needs network access to your cluster hosts and admin privileges. Make sure your service account has the right delegation configured, especially if you’re running Kerberos auth (which you should be).
—
Ten Built-In Reports, One Click
hvReports ships with ten operational report templates ready on first launch:
- Cluster Health Overview
- VM Inventory
- Host Capacity
- Right-sizing Candidates
- Storage/CSV Health
- Snapshot Hygiene
- Patching Compliance
- Drift History
- Network Audit
- Executive Roll-up
Hit generate, and it saves both a .md and .html file and opens it in your browser. You can clone any template, customize it, and schedule it on a cadence. The Executive Roll-up one in particular is something I’d have killed for on previous engagements — a single-pane summary you can drop into a Teams message or email without manually cobbling it together from five different sources.
One thing to be aware of: report inventory sections are empty until the first cluster refresh. So if you generate a report immediately after launching the app and connecting to a cluster for the first time, don’t panic when sections look blank. Refresh first, then generate.
—
Troubleshooting Scenarios (Read-Only, Thank Goodness)
hvTroubleshoot covers six diagnostic scenarios:
- Stuck live migrations
- CSV redirected I/O
- VMs stuck in Saved state
- Replica out of sync
- Unregistered cluster roles
- Inter-node network issues
These are read-only diagnostics — the tool explains the symptom, runs the checks, and shows you a fix script with full preview before anything touches the cluster. That’s the right approach. Cluster-Aware Updating automates software update installation on clusters, but for the kind of mid-incident “why is this live migration stuck?” troubleshooting that CAU doesn’t help with, having a guided diagnostic that doesn’t accidentally make things worse is genuinely useful.
—
Workload Insights and Recommendations
hvInsights surfaces right-sizing candidates, noisy neighbor detection, and utilization pattern analysis from collected performance data. hvRecommendations gives you a risk-rated queue of actionable items with script preview before you apply anything. There’s also an engine diagnostic panel — and here’s a nice touch — an empty recommendations queue tells you why it’s empty, rather than just showing nothing and leaving you to wonder if the feature is broken.
The drift tracking surface is also new in v2.0: severity-classified “what changed?” history across your cluster. Useful for those Monday morning “wait, who changed that?” conversations.
—
The AI Assistant (Optional, Off by Default)
ClusterIQ v2.0 adds an optional AI assistant that supports Claude, OpenAI, Azure OpenAI, or a local Ollama instance. It’s wired into drift events, recommendations, troubleshooting findings, playbook runs, and there’s an Ask AI toolbar button for ad-hoc questions.
I want to be clear: this is off by default, and you have to configure a provider explicitly. That’s the right call. Nobody wants an LLM quietly phoning home from their cluster management tool without knowing about it. The Azure OpenAI option is particularly interesting for shops with data residency requirements — you keep the inference on your own Azure tenant.
—
Requirements and the Self-Contained EXE Thing
No .NET install required. The release builds are self-contained single-file win-x64 executables with the .NET 8 Desktop Runtime bundled. You download one EXE, run it with admin privileges on a Windows machine with network access to your Hyper-V or Azure Local hosts, and you’re in.
The current Azure Local release is version 12.2609.1003.7, focused on reliability improvements. ClusterIQ’s S2D/Azure Local views are version-aware in their queries, which matters — starting with the 2604 release, Azure Local supports disaggregated deployments using SAN storage only, enabling clusters to scale beyond 16 nodes — so the tooling landscape is moving and version-aware queries are the right approach.
The update mechanism is also baked in: it queries the GitHub API, compares versions, downloads the latest EXE, replaces the binary, and restarts the app. Simple, no installer drama.
—
Screenshots
Here’s what the major views look like in v2.0:
Cluster and Node Health View
VM Management View
Autopilot Playbooks (v2.0)
Operational Reports (v2.0)
Troubleshooting Tool (v2.0)
—
My Take
The v1 release of ClusterIQ was a good visibility tool. v2.0 is something I’d actually deploy in a production ops workflow. The playbook engine with headless scheduling, the read-only troubleshooting diagnostics, and the one-click operational reports are the three features that move this from “nice to have open on a second monitor” to “this is part of how we operate clusters now.”
Local tools like PowerShell, Windows Admin Center, Hyper-V Manager, and Failover Cluster Manager support a range of VM operations that the Azure control plane doesn’t cover — and ClusterIQ lives in that same space, but adds the automation and reporting layer those tools are missing.
The AI assistant is a bonus. The dark mode is a necessity (sigh, why does every tool ship light-mode-first in 2026). And the fact that it’s a single self-contained EXE with no installer means you can actually get it in front of a customer or a new team member without a 45-minute setup conversation.
tl;dr — grab it from the GitHub releases page, connect it to your cluster, and run a cluster refresh before you touch the reports. You’ll thank me later.
—
What to Watch Next
- The headless service docs (
docs/HEADLESS.md) — specifically the service account setup section if you’re running this in a domain environment with Kerberos - The drift tracking surface over a few weeks of operation — that’s where the real operational value starts showing up once you have history
- Azure Local’s Arc Site Manager preview for multi-site fleet visibility — it’s a different layer but complementary to what ClusterIQ is doing on-prem
