Documentation

Requirements

Installation

Download HyperVThisSetup.exe from the downloads page and run it. The installer will:

1. Verify .NET Framework 4.8 is installed
2. Check for SQL Server and warn if not found
3. Install to C:\Program Files\HyperVThis\
4. Create Start Menu shortcuts and optional desktop icon
5. Register an uninstaller in Add/Remove Programs

Database Setup

Hyper-V This! stores all data in a local SQL Server database. Complete these steps before first use.

Step 1 — Create the database

CREATE DATABASE HyperVThis;

Step 2 — Run the schema script

Open HyperVThis_Schema.sql (in the install folder) in SSMS and execute it against your new database.

Step 3 — Configure the connection string

In Settings, update the SQL Server connection string and click Test SQL Connection:

Server=.\SQLEXPRESS;Database=HyperVThis;Integrated Security=True;

First Run

A 30-day trial starts automatically on first launch — no licence key required.

1. Configure your SQL connection in Settings
2. Go to Clusters → Add Cluster
3. Enter your cluster FQDN and click Test Connection
4. Save — then click Poll Now to discover nodes and VMs
5. Navigate to the Dashboard to see your cluster

Settings

SQL Connection

Connection string for your SQL Server database. Supports Windows Authentication (recommended) and SQL Authentication.

Poll Interval

How often the app polls your cluster. Default 30 seconds, minimum 5 seconds.

AI Settings

Anthropic API key for the Ask AI feature. Get yours at console.anthropic.com. Only your question and schema are sent — no cluster data leaves your network.

Adding Clusters

Go to Clusters → Add Cluster. Enter the cluster FQDN (e.g. hv-cluster.domain.local) and click Test Connection before saving.

If the connection test fails check that: WMI ports are open (TCP 135 + 49152-65535), your account has WMI access, and the FQDN resolves from this machine.

Email Alerts

Configure SMTP in Settings → Email Alerts.

Emails are sent for: node down (with affected VM list), VM state changes, CSV low space, and critical anomalies. Alerts are deduplicated within a 30-60 minute window.

Licensing

To activate: Settings → Licence → paste key → Activate. Licences are validated offline — no internet required.

Root Cause Analysis

Root Cause Analysis is the feature that makes Hyper-V This! fundamentally different from other monitoring tools. Rather than simply alerting you that something happened, the RCA engine analyses the event timeline, correlates historical metrics, and explains why it happened in plain English.

How it works

The RCA engine runs automatically after every poll cycle and on demand. For each detected event it:

1. Captures a 30-minute lookback window of node metrics before the event
2. Compares those metrics against the 7-day rolling baseline for that node
3. Correlates concurrent events — other VM migrations, alerts, node state changes
4. Builds a plain-English narrative with a probable cause and confidence level
5. Presents the full event timeline in chronological order

RCA scenarios

🔀 VM Migration

When a VM moves between nodes, the engine checks whether the source node was under resource pressure, whether other VMs migrated simultaneously (mass evacuation vs single move), and whether the source node went down shortly after. It distinguishes between operator-initiated migrations, cluster load-balancing, and failover events.

Example output: "VM 'SQLPROD01' moved from NODE1 to NODE2 at 14:32. In the 30 minutes prior, NODE1 CPU was at 94% — 2.3× above its 7-day average of 41%. 3 other VMs also migrated from NODE1 within the same 10-minute window, suggesting a cluster drain or load-balancing event."

🖥 Node Down

When a node goes offline, the engine checks whether memory or CPU were trending upward before the failure, whether VMs began migrating away before the node fully went down (a pre-crash signal the cluster detected instability), and whether the failure was sudden or gradual.

Example output: "NODE2 went offline at 09:14. Memory usage climbed from 67% to 98% across 3 poll cycles in the 30 minutes before failure. 2 VMs migrated away at 09:11 — 3 minutes before complete loss of connectivity — suggesting the cluster detected instability before full node failure."

💿 CSV Space Loss

When a CSV volume triggers a low-space alert, the engine checks whether VM count on the owning node has increased recently due to inbound migrations, correlates the rate of space consumption against the historical average, and identifies whether the cause is migration-driven or organic growth.

Example output: "CSV 'Volume1' has 8.2 GB free (4.1%). VM count on NODE1 increased by 3 following migrations from NODE2 last week. Additional VM storage, checkpoints, and swap files from the new VMs are likely responsible for the accelerated space consumption."

⚡ Anomaly Spike

When an anomaly is detected against the 7-day baseline, the engine checks whether inbound VM migrations can explain the increased load, or whether the spike is driven by an existing workload behaving unexpectedly.

Confidence levels

Using the RCA screen

1. Navigate to Root Cause Analysis in the sidebar
2. Select a cluster from the dropdown or leave on All Clusters
3. Click Analyse Now to run analysis against the last 24 hours of data
4. Click any finding card on the left to see the full narrative, probable cause, and event timeline on the right

Monitoring

Dashboard

Real-time overview: summary tiles (running VMs, alerts, CSV warnings), anomaly detection, per-node CPU/memory bars, recent alerts, and VM movement feed.

Anomaly Detection

After 5+ poll cycles, the engine builds a 7-day baseline per node and flags unusual CPU usage, memory usage, VM count changes, and storage depletion.

CSV Volumes

Visual space usage bars per volume. Below 10% = amber warning, below 5% = red critical.

VM Management

• Start / Stop / Pause / Save State — standard power actions via WMI
• Migrate — move VM to another node using the cluster's migration settings
• Console — connect via vmconnect.exe (requires Hyper-V Management Tools)
• Checkpoints — create, view, restore, and delete VM checkpoints

Smart Alerts

The Alert Log shows all events with severity, type, entity, and timestamp. Filter by cluster, severity, and resolved status. Acknowledge individually or in bulk. Smart alerts include contextual detail — a node-down alert lists the VMs that were running on it.

Ask AI

Query your cluster in plain English. Requires an Anthropic API key (free at console.anthropic.com).

Claude generates a safe SQL SELECT query, which runs locally against your database. No cluster data ever leaves your network.

Example questions: "Which VMs are stopped?", "Show nodes with CPU above 80%", "How many migrations happened this week?"

Reports

Available on Pro and Enterprise licences. Five report types with date range and cluster filtering: VM Inventory, Node Summary, CSV Volume Usage, Alert History, VM Movement History. Export as CSV or HTML/PDF.

FAQ

Do I need an internet connection?

No — fully offline. Only Ask AI (Anthropic API) and email alerts (your SMTP server) require network access.

Which account should I use?

An account with WMI access to your cluster nodes. Use Run as different user if this differs from your daily login.

Is SQL Server Express sufficient?

Yes. The free 10GB Express edition is more than enough for any Hyper-V environment.

Troubleshooting

Cannot connect to cluster

Verify WMI ports (TCP 135 + 49152-65535), account permissions, and that the FQDN resolves. Run as your cluster admin account.

VMs show but CPU/memory are 0

These values require running VMs with integration services active. Stopped VMs will always show 0.

Email test works but no alert emails

Check %APPDATA%\HyperVThis\smartalert_debug.log after a poll. It shows whether conditions are detected and whether email is enabled.

Ask AI blocked query

The safety filter blocks any statement that isn't a SELECT. Try rephrasing as a question about your data.