O&M
The O&M page supports full-link, multi-layer, and one-stop O&M for resources, applications, and user experience. Specifically, this page displays the following types of cards: infrastructure monitoring, application monitoring, alarm statistics, host monitoring (CPU and memory), component monitoring (CPU and memory), container instance monitoring (CPU and memory), host monitoring (disk), host monitoring (network), cluster monitoring (CPU and memory), and cluster monitoring (disk) cards.
Infrastructure Monitoring Card
This card mainly displays infrastructure metrics. You can select one or all clusters to view information. When you select all clusters, the following information is displayed:
- Host running status, CPU usage, and physical memory usage.
- Trend graph of network traffic in the last 30 minutes. The values of each point in the graph respectively indicate the total receive rate (BPS) and send rate (BPS) of all clusters in one minute. The values above the graph respectively indicate the total receive rate (BPS) and send rate (BPS) of all clusters at the latest time point.
- Trend graph of CPU and memory usage in the last 30 minutes. The values of each point in the graph respectively indicate the average CPU and memory usage of all clusters in one minute. The values above the graph respectively indicate the average CPU and memory usage of all clusters at the latest time point.
Application Monitoring Card
This card mainly displays application metrics:
- Running status of applications, components, containers, and instances.
- When you select an application, the following information is displayed:
- Trend graph of network traffic in the last 30 minutes. The values of each point in the graph respectively indicate the receive rate (BPS) and send rate (BPS) of the selected application in one minute. The values above the graph respectively indicate the receive rate (BPS) and send rate (BPS) of the selected application at the latest time point.
- Trend graph of CPU and memory usage in the last 30 minutes. The values of each point in the graph respectively indicate the CPU and memory usage of the selected application in one minute. The values above the graph respectively indicate the CPU and memory usage of the selected application at the latest time point.
Alarm Statistics Card
This card mainly displays alarms, log usage, threshold rules, and trends of alarms and hosts.
Component Monitoring (CPU and Memory) Card
This card mainly displays:
- The top 5 components with high CPU and memory usage in the last minute.
- Trend graph of the CPU and memory usage of the selected component in the last hour. The values of each point in the graph respectively indicate the CPU and memory usage of the component in one minute.
- CPU and memory usage of the selected component at the latest time point, which is displayed above the trend graph.
- You can select Hide system components in the lower left corner.
Cluster Monitoring (Disk) Card
This card mainly displays:
- The top 5 clusters with high disk usage in the last minute.
- Trend graph of the disk usage of the selected cluster in the last hour. The value of each point in the graph indicates the disk usage of the cluster in one minute.
- Disk usage of the selected cluster at the latest time point, which is displayed above the trend graph.
Container Instance Monitoring (CPU and Memory) Card
This card mainly displays:
- The top 5 container instances with high CPU and memory usage in the last minute.
- Trend graph of the CPU and memory usage of the selected container instance in the last hour. The values of each point in the graph respectively indicate the CPU and memory usage of the container instance in one minute.
- CPU and memory usage of the selected container instance at the latest time point, which is displayed above the trend graph.
- You can select Hide system instances in the lower left corner.
Host Monitoring (Disk) Card
This card mainly displays:
- The top 5 hosts with high disk read/write rate in the last minute.
- Trend graph of the disk read/write rate of the selected host in the last hour. The values of each point in the graph respectively indicate the disk read/write rate of the selected host in one minute.
- Disk read/write rate of the selected host at the latest time point, which is displayed above the trend graph.
Host Monitoring (Network) Card
This card mainly displays:
- The top 5 hosts with high send/receive rate in the last minute.
- Trend graph of the send/receive rate of the selected host in the last hour. The values of each point in the graph respectively indicate the send/receive rate of the selected host in one minute.
- Send/receive rate of the selected host at the latest time point, which is displayed above the trend graph.
Host Monitoring (CPU and Memory) Card
This card mainly displays:
- The top 5 hosts with high CPU and memory usage in the last minute.
- Trend graph of the CPU and memory usage of the selected host in the last hour. The values of each point in the graph respectively indicate the CPU and memory usage of the host in one minute.
- CPU and memory usage of the selected host at the latest time point, which is displayed above the trend graph.
Cluster Monitoring (CPU and Memory) Card
This card mainly displays:
- The top 5 clusters with high CPU and memory usage in the last minute.
- Trend graph of the CPU and memory usage of the selected cluster in the last hour. The values of each point in the graph respectively indicate the CPU and memory usage of the cluster in one minute.
- CPU and memory usage of the selected cluster at the latest time point, which is displayed above the trend graph.
More Operations
Perform the operations listed in Table 1 if needed.
Operation |
Description |
---|---|
Adding a card to favorites |
To hide a card, click in the upper right corner of the card and choose Add to Favorites. After a card is added to favorites, it is hidden from the O&M page. To view the card later, obtain it from favorites. |
Adding a card to dashboard |
Click in the upper right corner of the card and choose Add to Dashboard. |
Zooming in a metric graph |
Click in the upper right corner of the metric graph. |
Drilling down blue texts |
Click the blue texts, such as Host, Application, or Component to drill down to the details page. |
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot