All Products
Search
Document Center

:Service limits (legacy)

Last Updated:Jul 03, 2025

This topic describes the service limits of Managed Service for Prometheus.

Data query

Item

Limit

Adjustable

Time range

Unlimited. However, a single query for one time series can return up to 11,000 data points ((End Time − Start Time)/Time Interval ≤ 11,000).

No

Single query timeout

10 minutes

No

Time series scanned per query

40,000

Yes. Submit a ticket for quota increase. Expansion does not affect existing data.

Data points scanned per query

50 million

No

Queries per second (QPS)

  • China (Beijing), China (Hangzhou), China (Shanghai), China (Shenzhen), and China (Zhangjiakou) regions: 2,000

  • Other regions: 500

No

Data collection and reporting

Item

Limit

Adjustable

Data point size

2 KB

No

Time series per user

Automatically scaled (dynamically increased as usage approaches limit).

N/A

Data size per reporting request

1 MB. Excess data is discarded.

No

Label length per metric

256 characters per label key/value. Excess labels use the MAX_TAG_VALUE_LIMITED parameter value.

No

Labels per metric

256. Excess labels may result in data loss.

No

Data points per collection for an agent replica

3.5 million (3-core CPU, 4 GB memory)

Yes. Scale out replicas.

Targets per agent replica

5,000 (3-core CPU, 4 GB memory)

Yes. Scale out replicas.

Data points per target (30s window)

600,000

Yes. Scale out replicas.

Replicas in agent HPA settings

20

Yes. Manually scale out the replicas in the Managed Service for Prometheus console.

Container Service for Kubernetes (ACK) cluster compatibility

Agents V3.2.1 or later require Kubernetes V1.16 or later.

Yes. Update the cluster version.

Data volume per target

512 MB. Excess data is discarded.

No

Others

Item

Limit

Adjustable

Recording rules per cluster

1,000

No hard limit, but processing delays may occur.

Folders per tenant in Managed Service for Grafana Shared Edition

250

No. Upgrade it to Pro Edition.