Aliyun How to monitor server health

　　Monitoring

You can monitor the health of the cloud server ECS instances from multiple dimensions to ensure that the instance is running smoothly.

You can monitor the operation of an ECS instance from the following two portals:

Instance Details

Cloud Monitoring

Instance Details

Log on to the Cloud Server Management Console.

Click an instance in the left-hand navigation bar. Then select the region at the top of the page.

Locate the instance you want to monitor, and click the instance name.

On the Instance Details page, you can see the monitoring information, including CPU usage and network access and net.

You can set the time period you want to view on the right side of the page and on the top right of the monitor diagram.

Description: Because the aggregation operation is different in the display, for example, the average of 5 minutes and 15 minutes will show different results, so the length of the selection time will affect the accuracy of the display. The smaller the selection time range, the finer the display results.

The following is a resolution of the monitoring information:

CPU: The monitoring data displayed is the percentage of server CPU usage. The higher the percentage, the higher the load on the server CPU.

Windows instances can use Task Manager to view CPU usage on the server, sorted by CPU, to find out what program is consuming the server's CPU resources.

The Linux instance can view CPU usage through the top command. Log on to the server, execute command top on the command line, and then enter Shift+p, which allows top to be sorted by CPU size to see the current CPU-intensive process.

Network: The display of monitoring data is the server's public network traffic, the unit is KBPS,1MB bandwidth =1024 Kbps. Monitoring data can be viewed from the server and into the server in two directions of traffic. 1 MB of bandwidth, if the net traffic reaches 1024 kbps, the bandwidth is basically run full.

Cloud Monitoring

In the admin console, open the product and service > Cloud Monitor, or on the instance details page of the instance, click Set Alarm rule.

Click the cloud server ECS in the left-hand navigation bar, and then select the instance name to monitor for the item.

Click on the install, you can monitor the instance's operating system. Click the monitor icon to view the various base parameters, and click the alert rule to set the alert rule.

For more information about cloud monitoring, please refer to the product documentation for cloud monitoring.

Background knowledge: Conversion of bandwidth units

First, explain the difference between KB and KB:

The information in the computer is binary to indicate that each 0 or 1 is called a bit (bit), expressed in lowercase b;

8 bit is 1 byte (byte), in uppercase B, that is, 1B = 8b;

Uppercase K or lowercase k is thousand, that is, thousands of bits (KB) or thousands of bytes (KB).

In the network monitoring information that is seen in ECS console, PS refers to/s, that is, per second. Kbps refers to the network speed, which is the number of thousands of messages transmitted per second. Generally describe the bandwidth of the BPS is often omitted, for example: 4M bandwidth, complete the writing should be 4Mbps.

Myth: How much is the bandwidth, download speed is how much.

The correct conversion relationship, taking 1M bandwidth as an example:

1KB=8KB (1 byte=8 bits)

1mbps=125kb/s

The download rate for 1M bandwidth is theoretically 125kb/s, and some applications (including remote connections) in the system consume a small amount of bandwidth, so the actual rate of 100--110kb/s is more common.