Docs

Torch Object Storage

Metric monitoring

Creating alert policies

You can create alert policies for TOS-related metrics according to your business needs. The system will push notifications via the specified channel when a monitored metric meets the trigger conditions, helping you promptly identify abnormal data. This topic explains how to do that.

Notes

TOS imposes limits on bandwidth and QPS performance. We recommend setting appropriate alert policies based on your business conditions and these limits. Otherwise, you may trigger throttling that could disrupt your business. For more information, see Constraints and limitations.

Prerequisites

Before you begin, see Creating an alert policy.

Steps

In the left navigation pane, select Alert Center > Alert policy.

On the Alert policy page, click Create alert policy.

On the Create alert policy page, configure the following parameters.

The main parameters are explained below. For descriptions of other parameters, see Creating an alert policy.

Section	Parameter	Description
Alert object	Resource type	Select Storage > TOS.
		Dimension	You can create an alert policy based on Bandwidth, QPS, Storage, Bucket Status Code, Account Dimension, and Time Delay.
		Resources	Select the bucket(s) for which you want to set an alert policy.
Alert rules	Metric type	Select the metric type. The options are Single-metric and Multi-metric. Select "Multi-metric" if you want to monitor more than one metric.
Alert rules		Trigger condition	For a list of monitoring metrics supported by TOS, see Monitoring metrics. When the dimension is set to Bucket Status Code, common trigger conditions include: 2xx Ratio < 90%: Indicates that the request success rate may be below 90%. 5xx Ratio > 0%: Indicates that server errors may have occurred. 429 Number of status code requests > 0: Indicates that throttling may be triggered. Configure trigger conditions based on your business scenarios.

After completing the settings, click Create.

Last updated: 2026.08.18 18:07:13