This page describes how you can monitor your Cloud Firestore usage and spot potential problems in your app. Looking at your app's usage can also give you an idea of what to expect on your bill. Additionally, if something seems off, getting a clear picture of your database's operations can be a helpful troubleshooting tool.
The Google Cloud Platform Console and the Firebase console include a Cloud Firestore usage dashboard that shows document reads, writes, and deletes over time.
The usage dashboard requires the
monitoring.timeSeries.list Cloud IAM permission.
The Project Owner, Editor, and Viewer roles grant this permission. You can also grant this permission
through a Cloud Monitoring role
or a custom role.
Usage dashboard and billing reports
The Cloud Firestore usage dashboards in the Firebase and Cloud consoles provide an estimate of usage. They can help you identify spikes in usage. However, the dashboard is not an exact view of billed operations. Billed usage is likely higher. In all cases of discrepancy, the billing report takes precedence over the usage dashboard.
Operations that cause discrepancies between the usage dashboard and billed usage include:
- Import and export operations. Reads and writes performed by these operations do not show up in the usage dashboard.
- No-op verify-only writes. Writes that only verify the existence or non-existence of a document contribute to billed read operations, but they show as `UPDATE_NOOP` and `DELETE_NOOP` respectively in the write usage dashboard.
- No-op writes. Operations that do not result in a change to the database, such as an update that does not change field values or a write to a deleted document may show in the usage dashboard as `UPDATE_NOOP` or `DELETE_NOOP`. Even though they show as `NOOP`, they still contribute to billed operations.
Collapsed writes. In cases with multiple writes to the same document in quick succession, the usage dashboard might collapse multiple writes together and count them as one. When billing usage, each write is still counted separately.
The usage dashboard also collapses writes for field transforms like server timestamps, numeric increments, and array union operations. For field transforms, the usage dashboard might count multiple operations as a single operation.
- Queries that return zero results. Queries with zero results incur a cost of one read operation. This usage is billed but does not appear in the usage dashboard.
- Read operations associated with aggregation queries such as COUNT. This usage is billed but does not appear in the usage dashboard.
Security rule usage
Additionally, the Firebase console provides a Security Rules evaluation dashboard, a useful, at-a-glance view of rules invocations. You can supplement this dashboard with detailed analysis in Cloud Monitoring.
Daily quotasWhen you create a Firebase project, you're also creating a project in the Google Cloud. The App Engine Quotas page in the GCP Console tracks daily Cloud Firestore usage information including reads, writes, index writes, deletes, stored data, and network egress.
Cloud Monitoring collects metrics, events, and metadata from Google Cloud products. The usage dashboard in the Cloud Firestore console reports the same metrics data. To set up custom dashboards and usage alerts, use Cloud Monitoring.
Cloud Monitoring includes the following Cloud Firestore metrics:
The number of successful document reads. You can break this metric down by the type of read: LOOKUP or QUERY.
This metric does not include reads from managed export operations.
The number of successful document writes. You can break the metric down by the type of write: CREATE or UPDATE.
This metric does not include writes from managed import operations.
|Document Deletes||The number of successful document deletes.|
The number of active connections to your database.
The number of snapshot listeners across all active connections.
|Time-to-live deletion count||
Total count of documents deleted by Time-to-live (TTL) policies.
|Time-to-live expiration to deletion delays||
Time elapsed between when a document expired under a Time-to-live (TTL) policy and when it was actually deleted.
- Realtime updates usage
Use the active connections and snapshot listeners metrics to measure your usage of realtime updates.
Let's say a user opens your app on their phone. The app then connects to Cloud Firestore and subscribes to 10 queries. This increases your metrics by 1 active connection and 10 snapshot listeners.
- Sampling rate
The Cloud Firestore metrics are sampled every minute, but updates may take up to 4 minutes to show up in your dashboards.