Custom Dashboard

custom

Got it! Here’s a detailed step-by-step guide on creating a custom Grafana dashboard using Prometheus as a data source, covering every aspect like selecting visualization, adding queries, resizing panels, and configuring alerts.

---

🎯 Step-by-Step Guide to Creating a Custom Grafana Dashboard

1️⃣ Log in to Grafana

• Open Grafana in a browser:

  http://<your-grafana-server>:3000

• Log in with your credentials (admin/admin by default for a fresh installation).

---

2️⃣ Add Prometheus as a Data Source

1. Click on the ⚙️ "Configuration" in the left sidebar.

2. Select "Data Sources".

3. Click "Add Data Source".

4. Choose "Prometheus" from the list.

5. Set the URL to your Prometheus server:

   http://localhost:9090

6. Click "Save & Test" (you should see a success message).

---

3️⃣ Create a New Dashboard

1. Go to "Dashboards" (📊 icon) → Click "New Dashboard".

2. Click "Add a new panel".

---

4️⃣ Add Panels & Queries (Metric Visualizations)

📌 Step 1: Choose the Data Source

• In the Query Editor, make sure Prometheus is selected as the data source.

📌 Step 2: Add a Query Using the Query Builder or Manual PromQL

• Option 1: Use Query Builder

  • Click on "Builder".

  • Choose a metric, e.g., node_cpu_seconds_total.

  • Set filters (e.g., mode="idle", instance="server1", etc.).

• Option 2: Use PromQL Directly

  • Switch to "Code" and enter:

    100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

  • This calculates CPU usage over the last 5 minutes.

📌 Step 3: Select a Visualization Type

• Click the "Visualization" tab.

• Choose from:

  • Time Series (for trends over time)

  • Gauge (for single values like CPU %)

  • Bar Gauge (for disk space comparison)

  • Stat (for single numeric values)

  • Table (for raw data)

For CPU Usage, select Time Series.

---

5️⃣ Customize Panel Settings

📌 Step 1: Rename the Panel

• Click on the Panel Title → Rename it to "CPU Usage".

📌 Step 2: Adjust the Panel Size & Position

• Drag edges to resize.

• Click "Save" when done.

📌 Step 3: Customize Display Options

• Change Legend settings (e.g., show per-instance breakdown).

• Adjust Y-axis range (set 0-100% for CPU).

• Enable Thresholds (e.g., mark 80% as a warning and 90% as critical).

---

6️⃣ Add More Panels

Repeat the above steps for Memory Usage and Disk Usage:

📌 Add Memory Usage Panel

1. Click "Add new panel".

2. Enter PromQL:

   (node_memory_MemTotal_bytes - node_memory_MemAvailable_bytes) / node_memory_MemTotal_bytes * 100

3. Choose Gauge as the visualization.

4. Set the title to "Memory Usage".

5. Click "Apply".

📌 Add Disk Usage Panel

1. Click "Add new panel".

2. Use PromQL:

   100 - ((node_filesystem_avail_bytes{mountpoint="/"} * 100) / node_filesystem_size_bytes{mountpoint="/"})

3. Select Bar Gauge.

4. Set the title to "Disk Usage".

5. Click "Apply".

---

7️⃣ Configure Alerts

📌 Step 1: Add an Alert for High CPU Usage

1. Select the CPU Usage panel.

2. Click "Alert" → "Create Alert Rule".

3. Define the alert:

   • Condition:

     100 - (avg by(instance) (rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100) > 80

   • For: 5 minutes.

   • Action: Send notifications.

4. Click "Save".

---

8️⃣ Save & Share Dashboard

📌 Step 1: Save Dashboard

1. Click "Save dashboard".

2. Give it a name, e.g., "Server Monitoring".

📌 Step 2: Share Dashboard

1. Click "Share".

2. Choose:

   • Public link (for viewing)

   • JSON Export (for backup)

---

9️⃣ Export & Import Dashboard

📌 Export Dashboard

1. Go to "Dashboard settings" → "JSON model".

2. Click "Copy JSON" → Save it as a file.

📌 Import Dashboard

1. Go to "Dashboards" → Click "Import".

2. Upload the JSON file.

---

🎯 Summary

✅ Added Prometheus as a data source ✅ Created CPU, memory, and disk usage panels ✅ Configured alerts for high CPU usage ✅ Customized panel size, layout, and visualization ✅ Saved, shared, and exported the dashboard

Would you like me to generate a JSON template for this dashboard? 🚀