Troubleshooting with Windows Azure Diagnostics Viewer: Best Practices

Quick Start: Installing and Reading Data with Windows Azure Diagnostics Viewer

What it is

Windows Azure Diagnostics Viewer (WAD Viewer) is a tool for opening and inspecting Azure diagnostics data (logs, performance counters, crash dumps, IIS logs) collected by the Azure Diagnostics pipeline from cloud services and virtual machines.

Prerequisites

• An Azure subscription and access to the storage account that holds diagnostics data.
• Azure Cloud Service or VM configured to emit diagnostics (diagnostics.wadcfg or Azure Diagnostics extension).
• Windows machine with .NET Framework (per tool requirements).

Install

1. Download the latest Azure Diagnostics Viewer (or use the Azure SDK tools that include it).
2. Run the installer and follow prompts.
3. If using a standalone executable, unzip and place the executable in a suitable folder.

Connect to diagnostics storage

1. Open the Diagnostics Viewer.
2. Click File → Connect (or Add Storage Account).
3. Enter the storage account name and key (or connection string).
4. Select the diagnostics container (typically named “wad-control-container” and “wad-iis-logfiles”, “wad-table”, or similar).

Select and load data

1. Choose the storage table or blob container you want to inspect:
   • Tables for trace logs and structured diagnostics (WADLogsTable).
   • Blobs for crash dumps, IIS logs, performance counters.
2. Pick a time range and role instance if applicable.
3. Click Refresh or Load to retrieve the records.

Reading and interpreting data

• Trace logs: view timestamp, level (Information/Warning/Error), and message. Filter by level, role, or text.
• Performance counters: inspect counter name, instance, value over time. Use built-in graphing to spot spikes.
• Crash dumps: download the blob and open with WinDbg or Visual Studio for postmortem analysis.
• IIS logs: open as text or export for parsing; look for 500s and high-latency entries.

Common tasks

• Filter by role instance to isolate a specific VM or cloud role.
• Export selected records to CSV for external analysis.
• Use time-range narrowing to correlate events with deployments or incidents.
• Save connections for repeated use.

Troubleshooting

• “No data” — verify diagnostics configuration on the role/VM and that the storage account credentials are correct.
• Slow loading — narrow time range or increase table partition filters.
• Access denied — check storage account key permissions and firewall/network rules.

Next steps

• Configure Azure Monitor and Application Insights for richer, integrated telemetry (recommended for new deployments).
• Automate diagnostics collection and retention policies via ARM templates or Azure Policy.