Amazon Managed Service for Prometheus service quotas

Amazon Managed Service for Prometheus has the following quotas. Amazon Managed Service for Prometheus vends CloudWatch usage metrics to monitor Prometheus resource usage. Using the CloudWatch usage metrics alarm feature, you can monitor Prometheus resources and usage to prevent limit errors.

Note

Unless otherwise noted, these quotas are per workspace.

Service quotas

Amazon Managed Service for Prometheus has the following quotas for series, labels, and API requests. The Possible error message column shows what error message you might see if your Prometheus data exceeds a limit. If you see one of these error messages, you may want to request an increase to the corresponding limit.

Resource	Default quota	Possible error message
Active series (metrics that have reported data in the past 2 hours) per workspace	2,000,000. You can request a quota increase.	per-user series limit of 2000000 exceeded, please contact administrator to raise it
Active series per metric name	2,000,000. You can request a quota increase.	per-metric series limit of 2000000 exceeded, please contact administrator to raise it
Ingestion rate	140,000 samples per second. You can request a quota increase.	ingestion rate limit (...) exceeded
Ingestion burst size	1,000,000 samples per second. You can request a quota increase.	ingestion burst size limit (...) exceeded
Labels per metric series	70. You can request a quota increase.	series has too many labels (...) series: '%s'
Maximum HA clusters	500 HA clusters can be used to ingest samples into a workspace. This quota can't be changed.	too many HA clusters (limit: 500)
Query bytes for range queries	5GB that can be scanned per 24-hour interval in a single range query. This quota can't be changed.	the query hit the aggregated data size limit
Query chunks fetched	20,000,000 chunks can be scanned during a single query. This quota can't be changed.	expanding series: The query hit the max number of chunks limit. A chunk stores raw samples of series for a certain time span.
Query series fetched	12,000,000 series can be scanned during a single query. This quota can't be changed.	expanding series: The query hit the max number of series limit.
Query samples	50,000,000 samples that can be scanned during a single query. This quota can't be changed.	query processing would load too many samples into memory in query execution
Query time range	32 days between the start time and the end time of a PromQL query. This quota cannot be changed.	the query time range exceeds the limit (query length: xxx, limit: yyy)
Retention time for ingested data	150 days. Data older than this is deleted from the workspace. You can request a quota increase or decrease.
Workspaces per Region per account	25. You can request a quota increase.	Limit exceeded. Maximum workspaces per region
Workspace management API operations per Region per account, including CreateLoggingConfiguration, CreateWorkspace, DeleteLoggingConfiguration, DeleteWorkspace, DescribeLoggingConfiguration, DescribeWorkspaces, ListWorkspaces, UpdateLoggingConfiguration, and UpdateWorkspaceAlias	10 transactions per second (TPS), with a burst limit of 10 TPS. You can request a quota increase.

Rules and alertmanager quotas

Resource	Default quota
Active alerts	1000. You can request a quota increase.
Alert aggregation group size	1000. You can request a quota increase.
Alert manager definition file size	1 MB. This quota can't be changed.
Alerts per workspace, in size	20 MB. You can request a quota increase.
API operations for Alert manager, including CreateAlertManagerDefinition, DeleteAlertManagerDefinition, DescribeAlertManagerDefinition, and PutAlertManagerDefinition	10 transactions per second (TPS), with a burst limit of 10 TPS. You can request a quota increase.
API operations for Rule group, including CreateRuleGroupsNamespace, DeleteRuleGroupsNamespace, DescribeRuleGroupsNamespace, ListRuleGroupNamespaces, and PutRuleGroupsNamespace	10 transactions per second (TPS), with a burst limit of 10 TPS. You can request a quota increase.
Inhibition rules	100. You can request a quota increase.
Nodes in the alert manager routing tree	100. You can request a quota increase.
Rule group evaluation interval	30 seconds. You can request a quota decrease to permit setting the rule group evaluation to a lower value.
Rules per workspace. This limit is for all rules in the workspace, across rule groups and namespaces.	2000. You can request a quota increase.
Rules definition file size	1 MB. This quota can't be changed.
Templates in the alert manager definition file	100. You can request a quota increase.

Additional quotas on ingested data

Amazon Managed Service for Prometheus also has additional quotas for data ingested into the workspace.

• Metric samples older than 1 hour are refused from being ingested.

• Every sample and metadata must have a metric name.

• Maximum combined size of all label names and values for one series in an ingestion request: 7168 bytes

• Maximum number of metadata per metric: 10

• Maximum size of ingestion or query request: 1 MB

• Maximum length accepted for metric metadata, which includes metric name, HELP, and UNIT: 1024 bytes

• Maximum number of active metrics with metadata per workspace: 20,000

• Maximum retention time for ingested metrics: 150 days. Data older than this is deleted from the workspace.