Description
You can set traffic-related alerts such as alerts for transactions, queries, and timeout events in a cluster or tenant based on the business characteristics. Then, the system will send you notifications in the case of abrupt business traffic increases due to abnormal factors.
Note that the business characteristics and request types vary in different clusters. We recommend that you set the metrics based on the stress testing results of the application system.
Alert rule
Alert item | Metric type | Metric | Default threshold | Duration (consecutive cycles for triggering an alert) | Detection cycle | Alert level |
|---|---|---|---|---|---|---|
Tenant write requests | Single metric | tps | Set the threshold to 80% of the maximum value of the stress testing result | 3 | 1 minute | Warning |
Tenant read requests | Single metric | qps | 3 | 1 minute | Warning | |
Response time for tenant write requests | Single metric | tps_rt | Set the threshold within the acceptable range of the application system | 3 | 1 minute | Warning |
Response time for tenant read requests | Single metric | qps_rt | 3 | 1 minute | Warning |
Impact on the system
This issue degrades the service performance of the application system.
Possible causes
You have performed operations such as releasing new applications.
The existing resources cannot cope with your rapid business growth.
Solutions
Increase the tenant or cluster resources.