What changed in the Synthetic Monitoring upgrade
Synthetic Monitoring was upgraded on June 27, 2023. The upgrade introduces new monitoring point types, flexible task management, richer analysis dimensions, and built-in integrations with Managed Service for Prometheus, Managed Service for Grafana, and Simple Log Service.
Starting from June 1, 2024, ARMS Synthetic Monitoring no longer supports activating pay-as-you-go billing or purchasing subscription resources. For more information, see Product Change: Discontinuation of Synthetic Monitoring Activation.
Monitoring point types
The upgrade adds two monitoring point types:
| Monitoring point type | Description |
|---|---|
| ECS instances | Run synthetic checks directly from your Elastic Compute Service (ECS) instances. |
| Mobile clients | Test over 3G, 4G, and 5G networks across more than 100 cities in China. |
Scheduled tasks
Scheduled tasks gain several capabilities:
| Feature | Description |
|---|---|
| Availability assertions | Specify assertions for availability for each synthetic monitoring task. |
| Custom monitoring point groups | Save frequently used monitoring points into reusable groups. |
| Batch operations | Start or stop multiple tasks at once. |
| Task tags | Organize tasks with tags for easier filtering and management. |
| Alert status on the task list | View alert status inline and sort tasks by availability. |
| Quick real-time checks | Launch a real-time synthetic monitoring run directly from a scheduled task's configuration. |
Real-time synthetic monitoring
Real-time synthetic monitoring lets you run ad-hoc checks without setting up a scheduled task:
Launch checks instantly with custom monitoring points.
Compare results across multiple tasks side by side.
Monitor network quality and webpage performance.
Run file transfer tasks.
View results on a redesigned data analysis page.
Task analysis
The upgraded analysis tools provide deeper visibility into task performance:
| Feature | Description |
|---|---|
| Built-in Grafana dashboard | A pre-configured dashboard provides a task overview out of the box. |
| Custom analysis dimensions | Filter and group results by custom conditions and dimensions. |
| Comparative analysis | Compare performance across multiple tasks, regions, and conditions in a single view. |
| Percentile latency metrics | Track P50, P95, and P99 latencies for precise performance measurement. |
| Automatic error capture | Error response headers and bodies are recorded automatically for faster debugging. |
| Redesigned task details page | A new layout makes it easier to navigate task details. |
Alert management
Alerting supports finer-grained control over rules, dimensions, and notifications:
| Feature | Description |
|---|---|
| Multi-dimensional alerts | Set alerts by task, region, ISP, or other dimensions. |
| Multiple rules per task | Attach more than one alert rule to a single synthetic monitoring task. |
| Customizable alert properties | Configure the check interval, severity level, and alert content independently. |
| Tags and notes | Add tags and notes to alerts for better organization and context. |
| Notification policies | Define how and where alert notifications are delivered. |
Open-source integrations
Synthetic Monitoring integrates with the following services by default:
| Integration | Description | Learn more |
|---|---|---|
| Managed Service for Prometheus | Writes metric data to free Prometheus instances automatically. | Metrics written from Synthetic Monitoring to Prometheus |
| Managed Service for Grafana | Enables custom task dashboards through Grafana. | -- |
| Alibaba Cloud Simple Log Service | Writes detailed monitoring data to Logstores of Simple Log Service. | Fields written to Simple Log Service |
Access Synthetic Monitoring data programmatically through the remote read addresses and HTTP API endpoints of Prometheus, or through the Simple Log Service API.
Costs
Cloud host monitoring: CNY 0.001 per probe (pay-as-you-go).
Mobile monitoring points: CNY 0.08 per probe (pay-as-you-go).
Prices for other monitoring points remain unchanged.
Supports task-level usage statistics, providing a clear view of usage for different monitoring point types.