Monitoring in Spanner Omni works similarly to monitoring in Spanner, with some exceptions. This document provides explanations of those exceptions and points you to related information in Spanner monitoring documentation. As you read the Spanner documentation, refer back to this page to confirm how Spanner Omni differs.
Spanner features that aren't supported
Spanner Omni supports all Spanner monitoring features except end-to-end tracing.
Capture telemetry signals
To gain insights into performance, usage, and potential issues, you can capture signals from your Spanner Omni database.
Signal capture overview
Telemetry signals help you understand how your database is performing.
The following signals are available for Spanner Omni:
| Signal type | Description |
|---|---|
| Metrics | Numeric data about database health or performance. |
| Logs | Generated records of system activity over time. |
| Traces | Request paths through your application. |
| Statistics tables | Built-in tables for querying database information. |
In Spanner Omni, client-side metrics aren't supported. Client-side metrics include latencies for GFE metrics, client-side metric collection, and custom client-side metric collection. For more information, see Signal capture overview in the Spanner documentation.
Trace collection
Traces represent the path of a request through your application. They let you follow the flow of a request and help you identify the root cause of an issue. For example, tracing can provide information for every request from a client, such as spans with timestamps of when the client sent and received RPC requests. This helps you monitor and debug database requests.
For more information, see Trace collection overview in the Spanner documentation.
Set up trace collection using OpenTelemetry
You can configure Spanner client libraries to export traces using OpenTelemetry APIs. This process involves configuring the OpenTelemetry SDK and using an exporter to send trace data to an observability backend.
Client-side tracing isn't available for Spanner Omni.
For more information, see Set up trace collection using OpenTelemetry in the Spanner documentation.
Audit logs
Audit logs track administrative changes and data access events for security and compliance. You can use audit logs to monitor request latencies by referring to the processing duration fields. For more information, see Audit logs in the Spanner documentation.
Monitor instances
Monitor your instances to ensure they are performing as expected and to help you troubleshoot potential issues.
Monitor instance performance using insights
The system insights dashboard provides charts and metrics for latency, CPU utilization, storage, throughput, and other performance statistics. You can use this dashboard to monitor Spanner Omni instances and databases. For more information, see Monitor instances with system insights in the Spanner documentation.