Understand Grafana Cloud usage limits

Grafana Cloud continuously provides your account limits in the following metrics:

grafanacloud_instance_metrics_limits
grafanacloud_logs_instance_limits
grafanacloud_traces_instance_limits

To find these limits in your Grafana Cloud instance:

Click Explore (compass icon).
Select grafanacloud-usage from the drop-down menu.
In the Metrics browser query field, enter the metric, for example grafanacloud_instance_metrics_limits.
Click Run query.

The label limit_name provides the specific limit. See below for a complete reference.

Metric usage limits

The following limits apply to Grafana Cloud Metrics.

Ingestion limits

Limit name	Limit description	Default limit
`ingestion_rate`	The maximum number of ingested samples per second for a series.	10,000
`ingestion_burst_size`	The maximum allowed ingestion burst size in number of samples.	200,000
`max_global_series_per_user`	The total maximum series for each user in an instance.	150,000
`max_global_exemplars_per_user`	The total maximum exemplars for each user in an instance.	100,000
`max_aggregation_ingestion_rate`	The maximum ingestion rate for aggregated metrics.	250,000
`max_aggregated_tenant_series`	The maximum series for aggregated tenant metrics.	3,000,000
`max_global_metadata_per_user`	The maximum number of different sets of help text and unit in an instance.	30,000
`max_global_metadata_per_metric`	The maximum number of different sets of help text and unit per metric.	10

Query limits

Limit name	Limit description	Default limit
`max_partial_query_length`	The maximum time range for partial queries at the querier level (exported in seconds).	768h
`max_fetched_chunks_per_query`	The maximum number of chunks with compressed samples that each query can fetched from the storage.	2,000,000

Rules limits

Limit name	Limit description	Default limit
`ruler_max_rule_groups_per_tenant`	The maximum number of rule groups for an instance.	35
`ruler_max_rules_per_rule_group`	The maximum number of rules per rule group for an instance.	20

About limit autoscaling

Grafana Cloud automatically scales some metric usage limits based on your consumption. Limit autoscaling occurs every three hours.

The following limits are autoscaled:

ingestion_rate
ingestion_burst_size
max_global_series_per_user
max_aggregation_ingestion_rate
max_global_metadata_per_user
max_aggregated_tenant_series
ruler_max_rule_groups_per_tenant
ruler_max_rules_per_rule_group

To adjust a limit that isn’t autoscaled, contact Grafana Support.

Logs usage limits

Ingestion limits

Limit name	Limit description	Default limit
`ingestion_rate_mb`	Per-user ingestion rate in MB.	5
`max_global_streams_per_user`	The maximum number of active streams per user, across ingesters.	5,000
`retention_period`	Compactor retention for storage, in hours.	744
`max_line_size`	The maximum size of a log line. This limit can’t be modified.	256kb
`max_label_names_per_series`	The maximum number of labels per stream.	15
`max_label_name_length`	The maximum size label name.	1024
`max_label_value_length`	The maximum size label value.	2048

Query limits

Limit name	Limit description	Default limit
`max_query_length`	The maximum time range for partial queries at the querier level.	721h
`max_query_lookback`	Querier limit on how far back data can be queried.	0 (disabled)
`max_query_series`	The maximum unique series that can be returned by a metric query.	500

Traces usage limits

Use these metrics to help track your traces usage limits.

Limit name	Limit description	Default limit
`max_bytes_per_trace`	The maximum size of a trace in bytes.	3,000,000
`ingestion_rate_limit_bytes`	Per-tenant ingestion rate limit per second in bytes.	500,000
`ingestion_burst_size_bytes`	Per ingestion burst size in bytes.	500,000
`block_retention`	The duration to keep blocks/traces in seconds.	2,592,000

Discarded Spans panel in the Billing dashboard

In the Billing dashboard, there is a panel that shows Discarded Spans and a color-coded legend with a reason why the spans were discarded. The table that follows defines each of those reasons.

Reason for discarded span	Description
`internal error`	A general error that should be further explained in the logs.
`live_traces_exceeded`	Exceeded the max number of traces allowed in the ingester per tenant.
`rate limited`	Exceeded the `ingestion_rate_limit_bytes` limit.
`trace_too_large`	Exceeded the `max_bytes_per_trace` limit.

You can use grafanacloud_traces_instance_percentage_complete_traces_flushed metric. This metric has a value between 0 and 1, indicating the percentage of traces without any orphaned spans when the trace is flushed to storage.

In addition, you can use the grafanacloud_traces_instance_discarded_spans_total:rate5m metric to review the number of discarded spans, including the reasons for discarding, in 5 minute periods.

Let’s say that you want more information about traces that are discarded because they were too large. You can query for the grafanacloud_traces_instance_discarded_spans_total:rate5m metric using the grafanacloud-usage data source. To narrow the results down to the specific “trace too large” reason, modify the query to be grafanacloud_traces_instance_discarded_spans_total:rate5m{reason="trace_too_large"}. While thegrafanacloud_traces_instance_discarded_spans_total:rate5m metric shows the number of discarded spans, it doesn’t provide any information on where the large traces originate.

Grafana Cloud usage

Grafana Cloud continuously provides your account usage in the following metrics:

grafanacloud_instance_metrics_*
grafanacloud_logs_instance_*
grafanacloud_traces_instance_*

For example, to monitor log stream count, you can use grafanacloud_logs_instance_active_streams.

Cloud Traces usage

If you need to have a limit raised, then contact Grafana Support for assistance.

Metric	Description
`grafanacloud_traces_instance_bytes_received_per_second`	Number of bytes received by each Grafana Cloud Traces stack for the organization.
`grafanacloud_traces_instance_created_date`	Date (milliseconds since epoch) when the instance was created.
`grafanacloud_traces_instance_discarded_spans_total:rate5m`	Number of discarded spans, including the reasons for discarding, in 5 minute periods.
`grafanacloud_traces_instance_percentage_complete_traces_flushed`	A value between 0 and 1 indicating the percentage of traces without any orphaned spans when the trace is flushed to storage.
`grafanacloud_traces_instance_info`	Information about the instance such as org ID and customer ID.
`grafanacloud_traces_instance_limits`	Span volume limits for the organization, on a per-stack basis.
`grafanacloud_traces_instance_queries_per_second`	Number of queries that are carried out on a per-second rate basis.
`grafanacloud_traces_instance_spans_received_total:rate5m`	Per second rate of spans received by each Grafana Cloud Traces stack for the organization, over a 5 minute period.
`grafanacloud_traces_instance_usage`	Billable trace usage over the last month for stacks in the organization.
`grafanacloud_traces_instance_metrics_generator_active_series`	Number of active series being generated via ingested spans.
`grafanacloud_traces_instance_percentage_traces_with_root_spans_flushed`	A value between `0` and `1` indicating the percentage of traces with root spans when the trace is flushed to storage.