To ensure efficient operation, Cloud Monitoring imposes the following usage limits.
Custom metrics
| Category | Maximum value |
|---|---|
| Custom metric descriptors per project1 | 10,000 |
| Labels per metric descriptor | 30 |
| String length for label key | 100 |
| String length for label value | 1024 |
| Time series included in a write request2 | 200 |
| Rate at which data can be written to a single time series3 | one point each 5 seconds |
| Histogram buckets per custom distribution metric | 200 |
| Workload, Prometheus, and external4 metric descriptors per project | 25,000 |
| Active time series from custom metrics per monitored resource5 | 200,000 |
| Active time series from workload metrics per monitored resource5 | 200,000 |
| Active time series from Prometheus per monitored resource5 | 1,000,000 |
| Active time series from external metrics per monitored resource5 | 200,000 |
1
This limit is imposed by Cloud Monitoring. Other services
might impose lower maximum values.
2
You can write only one data point for each time series in a request, so this
limit also functions as the maximum number of points that can be written per
request.
3
The Cloud Monitoring API requires that the end times of points written to
a time series be at least 5 seconds apart. You can batch write points to
a time series, provided that the data points are written in order.
4
External metrics are those written to external.googleapis.com.
5
A time series is active if you have written data points to it
within the previous 24 hours.
The limit specified in the row is the total number of active time series for a
single monitored resource (for example, a single gce_instance VM
or a single k8s_container container) across all user-defined
metrics within that row (custom, workload, Prometheus, or external). An
exception is the global monitored resource, for which the limit applies to
each user-defined metric separately. This is a system-wide safety limit and
isn't customizable.
Monitoring API quotas and limits
| Category | Maximum value |
|---|---|
| Limits to API usage | See the Quotas dashboard. For an API, click All Quotas to see your quota. |
| Lifetime of API page tokens | 24 hours |
About Monitoring API quotas
The Monitoring API has quota limits for the rates of time-series ingestion requests and time-series queries. Ingestion requests are calls that write time-series data, and queries are calls that retrieve time-series data. There are also internal limits on other Monitoring API endpoints; these endpoints aren't intended to handle high rates of requests.
To reduce the number of API requests you issue when your services write time-series data, use one API request to write data for multiple time series. We recommend that you write at least 10 objects per request.
For more information about batching API requests, see timeSeries.create.
If, after batching your API requests, you still require a higher Monitoring API quota limits, contact Google Cloud Support.
The other limits are fixed and as detailed on this page.
For more information, go to Working with quotas.
Data retention
Metric data points older than the retention period are deleted from time series.
| Category | Value |
|---|---|
Retention of data points from custom, external, and agent metric
types, including:
|
24 months1 |
Retention of data points from process-health metric
types: agent.googleapis.com/processes,except for count_by_state and fork_state, as
noted in the previous entry. |
24 hours |
| Retention of data points from all other metric types, including: | 6 weeks |
| Lifetime of API page tokens | 24 hours |
1 Metric data is stored for
6 weeks at its original sampling frequency,
then it is down-sampled to 10-minute intervals for extended storage.
2
Google Cloud Managed Service for Prometheus metric data is stored for 1 week at its original
sampling frequency, then it is down-sampled to 1-minute intervals for the next
5 weeks, then it is down-sampled to 10-minute intervals for extended storage.
Resource groups
| Category | Value |
|---|---|
| Number of resource groups per metrics scope | 500 |
| Maximum number of groups included in an email report1 | 10 |
1 When you configure Cloud Monitoring email reports, you can request information on utilization of your resource groups. Due to a limitation in the email reporter, the generated reports include information for only 10 groups.
Monitored project limits
Cloud Monitoring officially supports up to 375 Google Cloud projects per metrics scope .
You can add up to 1,000 Google Cloud projects per metrics scope , but you might experience performance issues, especially when querying custom metrics or historical data. Cloud Monitoring guarantees performant queries and charts only for 375 Google Cloud projects per metrics scope .
To raise your Cloud projects per metrics scope quota, you can request an increase of the "Monitored Projects / Monitoring Metrics Scope" quota. See the documentation about managing your quota for more details.
Limits for alerting and uptime checks
| Category | Value | Policy type1 |
|---|---|---|
| Alerting policies (sum of metric and log) per metrics scope 2 | 500 | Metric, Log |
| Conditions per alerting policy | 6 | Metric |
| Maximum time period that a metric-absence condition evaluates3 |
1 day | Metric |
| Maximum time period that a metric-threshold condition evaluates3 |
23 hours 30 minutes | Metric |
| Notification channels per alerting policy | 16 | Metric, Log |
| Maximum rate of notifications | 1 notification every 5 minutes for each log-based alert | Log |
| Maximum number of notifications | 20 notifications a day for each log-based alert | Log |
| Maximum number of simultaneously open incidents per alerting policy |
5000 | Metric |
| Period after which an incident with no new data is automatically closed |
7 days | Metric |
| Maximum duration of an incident if not manually closed | 7 days | Log |
| Retention of closed incidents | 13 months | Not applicable |
| Retention of open incidents | Indefinite | Not applicable |
| Notification channels per metrics scope | 4000 | Not applicable |
| Uptime checks per metrics scope 4 | 100 | Not applicable |
| Maximum number of ICMP pings per public uptime check | 3 | Not applicable |
2Apigee and Apigee hybrid are deeply integrated with Cloud Monitoring. The alerting limit for all Apigee subscription levels—Standard, Enterprise, and Enterprise Plus—is the same as for Cloud Monitoring: 500 per metrics scope .
3The maximum time period that a condition evaluates is the sum of the alignment period and the duration window values. For example, if the alignment period is set to 15 hours, and the duration window is set 15 hours, then 30 hours of data is required to evaluate the condition.
4This limit applies to the number of uptime-check configurations. Each uptime-check configuration includes the time interval between testing the status of the specified resource. See Managing uptime checks for more information.
Limits for charting
| Category | Value |
|---|---|
| Dashboards per metrics scope | 1000 |
| Charts on a dashboard | 40 |
| Lines on a chart | 300 |
Service-level objectives
| Category | Value |
|---|---|
| Number of SLOs per service | 500 |