comparison of alerting features between cloudmonitor and arms - Realtime Compute for Apache Flink

Realtime Compute for Apache Flink provides job monitoring and alerting features. You can use either CloudMonitor, a free monitoring service, or Managed Service for Prometheus, which is part of Application Real-Time Monitoring Service (ARMS). This topic compares these two services for Flink monitoring and alerting to help you choose the one that best suits your needs.

Feature comparison

Category	Features	ARMS	CloudMonitor
Service availability	-	The service availability for both Prometheus monitoring and alerting is at least 99.9%. For more information, see the Prometheus Service Level Agreement.	The availability of monitoring metrics is not covered by the Service-Level Agreement (SLA). The alerting service has an availability of at least 99.9%. For more information, see the CloudMonitor Service Level Agreement.
Costs	Monitoring and alerting fees	Metric collection and the alerting service for Prometheus are pay-as-you-go.	No subscription fees. Fully managed. Monitoring and alert management are free. For notification channels, only text messages and voice calls incur a small fee based on usage.
Costs	Data storage duration	Includes a monthly free quota of 50 GB. Data is stored for 90 days by default. For more information, see Prometheus Instance Billing.	Storage is free. Data is stored for 30 days by default.
Monitoring metrics	Metric display	Metrics are displayed in monitoring charts in the Flink development console. The ARMS console supports queries using Prometheus Query Language (PromQL) and display on Grafana dashboards.	Pre-aggregated metrics are displayed in monitoring charts in the Realtime Compute for Apache Flink console and the CloudMonitor console. This method has some limitations compared to ARMS. For more information, see What are the limitations of CloudMonitor alerting compared to ARMS?
Monitoring metrics	Number of monitoring metrics	Monitoring metrics in the Flink development console. Flink metrics on ARMS.	Monitoring metrics for Flink jobs.
Alert management	Configuration method	Flink development console. ARMS console.	Go to the CloudMonitor console from the Flink development console to configure or subscribe.
	Number of alert metrics	Configure alerts for 7 monitoring metrics, including job failure, in the Flink development console. For more information, see Configure monitoring and alerting. Configure alerts for other monitoring metrics in the ARMS console using custom PromQL syntax. Subscribe to event alerts for ECS breakdown post-processing and ECS proactive O&M impacts in the CloudMonitor console. For more information, see Configure monitoring and alerting.	Provides alerting for 20 monitoring metrics and lets you subscribe to event alerts. Event alerts include job failure alerts, ECS breakdown post-processing, and ECS proactive O&M impacts.
	Single-metric/multi-metric support	Configure single-metric alerts, including job failure, directly in the Flink development console. Configure multi-metric alerts in the ARMS console using custom PromQL syntax.	Go to the CloudMonitor console from the Flink development console to set single-metric or multi-metric alert rules and subscribe to job failure event alerts.
	Configure alert rules	Create an alert rule in Flink. Create an alert rule in Prometheus.	Create an alert rule in CloudMonitor.
	Configure alert templates	Create an alert rule template in Flink. Create an alert rule template in Prometheus.	Create an alert template in CloudMonitor.
	Event alerting	Does not support event alerting. Only job failure alerts are supported.	Supports system or custom event alerting. Supports delivering event alerts to a specified queue in Simple Message Queue (formerly MNS), a specified function, a URL accessible over the public network, or Simple Log Service.
Alert notifications	Notification methods	The Flink development console supports alert notifications by phone, text message, email, DingTalk, and webhook. On-call scheduling is not supported. The ARMS console sends alert notifications to contacts by phone, text message, email, DingTalk, WeCom, Lark, and webhook. On-call scheduling is supported. For more information, see Notification Recipient.	Sends alert notifications to contacts by phone, text message, email, DingTalk, WeCom, Lark, and webhook. On-call scheduling is supported. For more information, see Alert Contact.
	Alert notification policies	The Flink development console supports alert denoising policies. The ARMS console supports silence and escalation policies. For more information, see Escalation Policy.	Supports dynamic threshold alerting, merging alert notifications, and blacklist policies.
	Alert callback	Supported	Supported
	One-click alerting	Not supported	Supported
OpenAPI	Monitoring and alerting OpenAPI	API overview Obtain Remote Read, Remote Write, and HTTP API addresses Note For details about the metrics, see Flink metrics.	Cloud service monitoring API Alerting service API

Switch monitoring and alerting services

The platform lets you switch between the two monitoring services to meet different business needs.

In the Realtime Compute for Apache Flink console, find the target workspace. In the Actions column, click More to switch to the other monitoring service.

Note

Read the notes on switching service types carefully. You can proceed with the configuration only after you select the confirmation checkbox.

For more configuration details, see Configure monitoring and alerting.