Alert Notifications

Tingyun Network provides an alert notification function. Users can flexibly develop alert strategies based on the needs of existing tasks, and supports three alert methods: performance alerts, error alerts, and smart alerts. The system can be triggered based on performance, errors, and availability respectively, and the alert system will send emails and SMS notifications. Intelligent alerts are based on machine learning principles, eliminating the need for complicated threshold configuration work. Through a long-term learning process, they gradually adapt to the user's understanding of alerts based on user feedback.

Performance Alert

Set multiple performance indicators and availability alerts for monitoring tasks. When the performance values or availability of these performance indicator items of the monitoring task exceed the set thresholds, the system will send emails and text messages to alert you, and generate alert logs.

The performance alert function provides static alerts and dynamic alerts. System administrators can flexibly configure alerts according to the conditions of their own websites.

Static alert directly sets a static performance value and a static availability value, sets the conditions for alert triggering, and triggers when the trigger is met or one of the conditions is met.

Since the content of some websites or other changes at any time, using static alerts will not be flexible enough. The biggest difference between dynamic alerts is that they can change dynamically based on an alert. The system administrator can specify a calculation period, during which the alert can change flexibly.

Configure Performance Alert Basic Info:

Configure performance alert conditions:

Configure performance alert rules:

Configure Performance Alert Notifications:

Error Alert

System administrators can set alerts for monitoring tasks when errors occur, and can select one or more errors as alert triggering conditions. When these errors occur in a monitoring task, an alert is triggered and an alert log is generated.

Configure Error Alert Basic Info:

Configure error alert conditions:

Configure error alert rules:

Configure Error Alert Notifications:

Intelligent Alert

Intelligent alerts use anomaly detection and intelligent alert algorithms based on time series prediction.

Anomaly detection algorithm based on time series prediction
Statistical methods based on data models are the most widespread anomaly detection technology. Its basic principle is to perform statistical modeling on the training data set (usually normal samples). If a data sample does not conform to the random model obtained by training, it will be identified as an abnormal sample. The fact that the data sample does not conform to the random model means that the sample is unlikely to be generated by the model. It is generally determined through statistical testing, based on the following assumption: normal data samples occur in the high probability area of the statistical random model, while abnormal data samples occur in the low probability area of the model. That is to say, we can get the probability that an unknown data sample is generated by a known model based on the test statistic. If the probability is less than a predetermined standard, the sample is considered an anomaly.

Intelligent alert algorithm based on machine learning
Through the anomaly detection algorithm based on time series prediction, we can allow the alert system to complete fault detection and alert work automatically and relatively accurately, thereby greatly reducing the arduous manual setting of alert thresholds for various performance indicators and manual management tasks, and achieving the purpose of automatic and intelligent fault detection.

Alert Log

You can view specific information when alerts occur through alert logs, including performance alert logs, error alert logs, and smart alert logs.

In the alert details, you can view the alert overview, alert timeline, performance scatter plot, and Probe access records.

Alert overview: including alert name, alert type, task name, Task Type, alert status, alert duration, and grouping method.

Alert timeline: represents the process of alert triggering, displayed in order from left to right, including normal alerts (orange), severe alerts (red), and alert cancellations (green). After selecting a point in the timeline, the scatter chart data below will be refreshed to display the monitoring data during that time period.

Alert detailed data: Displays the monitoring data in the current alert period with scattered point data.

Alert push

There are two types of alert push for Network products: HTTP push and UDP push. More and more customers are beginning to use a unified information access platform to connect information pushed by third-party service providers, such as alert information. Network products implement a templated alert push method, which can splice standard alert fields into the required format according to customer needs to meet customers' diverse needs for alert information.

Performance Alert#

Error Alert#

Intelligent Alert#

Alert Log#

Alert push#

Performance Alert

Error Alert

Intelligent Alert

Alert Log

Alert push