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 dropdown 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

To increase metrics usage limits, contact Grafana Support.

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_metric	The total maximum series for each metric across the entire instance.	20,000
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

Query limits

Limit name	Limit description	Default limit
max_partial_query_length	The maximum time range for partial queries at the querier level.	768h
max_series_per_query	The maximum number of unique series for which a query can fetch samples	100,000

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

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 cannot 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

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-user ingestion rate limit 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	Definition
`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 to monitor the percentage of traces that have no orphaned spans when the trace is flushed in Tempo. The value ranges from 0 to 1.