This document describes how to troubleshoot and solve the problem of not being able to log in to Windows and Linux CVM instances due to the instances’ overly high CPU or memory usage.
Troubleshooting Approaches
1. Log in to the instance and identify the process that is causing high CPU or memory usage.
2. Analyze the process.
If the process has an exception, the exception may be caused by a virus or a trojan. In this case, terminate the process or use an antivirus application to scan your system.
If the process is a service process, check whether the high CPU or memory usage is caused by an access volume change and whether it can be optimized.
If the process is a Tencent Cloud component process, submit a ticket and we will help you locate and troubleshoot the problem.
2. On the instance management page, locate the target CVM instance and click Log In, as shown in the following figure:
3. On the Log in to Windows instance page that appears, click Log In Now under Alternative login methods (VNC) to log in to the CVM instance.
4. On the login page that appears, select Send CtrlAltDel in the upper-left corner and click Ctrl-Alt-Delete to access the system login page, as shown in the following figure:
Viewing the resource usage of processes
1. In the CVM instance, right-click the taskbar and select Task Manager, as shown in the following figure:
2. In "Task Manager", you can view the resource usage, as shown in the following figure:
Note:
You can sort the processes in ascending or descending order by clicking CPU or Memory.
Analyzing the processes
Analyze the processes on the Task Manager page to troubleshoot and solve the problem.
The problem is caused by a system process
If a system process is occupying too many CPU or memory resources, complete the following steps:
1. Verify the name of the process.
Several viruses use names that are very similar to system processes, such as svch0st.exe, explore.exe, and iexplorer.exe.
2. Locate the executable file that corresponds to the process.
System processes are usually located in C:\\Windows\\System32 and have valid digital signatures and descriptions. To locate the corresponding executable file such as svchost.exe, select the process on the Task Manager page, right-click the process, and choose Open file location, as shown in the following figure:
If the executable file is not in C:\\Windows\\System32, it is likely that your CVM instance has a virus. Kill the virus manually or by using an antivirus application.
If the executable file is in C:\\Windows\\System32, restart your CVM instance or end safe but unnecessary processes.
The following describes typical system processes:
System Idle Process: a process that displays the percentage of time that the processor is idle for
system: a memory management process
explorer: the desktop and file management process
iexplore: the Microsoft Internet Explorer process
csrss: the Microsoft client/server runtime subsystem
svchost: a system process that is used to execute DLLs
Taskmgr: the task manager process
Isass: the local security permission service
The problem is caused by processes with exceptions
If you find that high CPU and memory usage is caused by processes with strange names such as xmr64.exe (a cryptocurrency mining malware), your CVM instance may be infected with viruses or trojans. In this case, use a search engine to verify whether the processes are in fact viruses or trojans.
Use an antivirus application to remove the virus or trojan. Then, back up the data and reinstall the operating system when necessary.
If the process is not a virus or a trojan, restart your CVM instance or end safe but unnecessary processes.
The problem is caused by a service process
If you find that the problem is caused by a service process such as IIS, HTTPD, PHP, or Java, further analyze the problem.
For example, check whether your business volume is high.
If yes, we recommend that you upgrade your CVM instance. If you do not upgrade your CVM instance, optimize your service processes.
If no, use service error logs to further analyze the problem. For example, check whether the resources are wasted due to incorrect parameter settings.
The problem is caused by a Tencent Cloud component process
If the problem is caused by a Tencent Cloud component process, submit a ticket, and we will help you locate and troubleshoot the problem.
Linux CVM instances
Logging in to the CVM instance
Select a CVM login method based on your actual needs.
Log in to the Linux CVM remotely via third party software.
Note:
If the Linux CVM has a high CPU load, you may fail to log in to the CVM.
If the Linux CVM has a high CPU load, you can log in normally via the Console.
Viewing the resource usage of processes
Run the following command to view the system load. View the %CPU and %MEM columns and identify which processes consume more resources.
top
Analyzing processes
Analyze the processes on the Task Manager page to troubleshoot and solve the problem.
If the problem is caused by a service process, analyze whether the service process can be optimized and accordingly optimize the process or upgrade the CVM configuration.
If the problem is caused by a process with an exception, the instance may have a virus. In this case, you can terminate the process or use an antivirus application to kill the virus. When necessary, back up the data and reinstall the operating system.
If the problem is caused by a Tencent Cloud component process, submit a ticket and we will help you locate and troubleshoot the problem.
Common Tencent Cloud components include:
sap00x: a security component
Barad_agent: a monitoring component
secu-tcs-agent: a security component
Terminating processes
1. Compare the resource consumption of different processes and record the PID of the process that needs to be terminated.
2. Enter k.
3. Enter the PID of the process that needs to be terminated and press the Enter key to terminate it, as shown in the following figure:
Suppose you need to terminate a process whose PID is 23.
Note:
If kill PID 23 with signal [15]: appears after you press Enter, press Enter again to keep the default settings.
4. If the operation is successful, the message Send PID 23 signal [15/sigterm] will appear. Press Enter to confirm the termination.
Other related problems
CPU usage is low but load average is high
Problem
The load average is an indicator of CPU load. The higher the load average, the longer the queue of pending processes is.
After the top command is executed, information similar to the following is returned, indicating that the CPU usage is low but the load average is very high.
Solution
Run the following command to view the process states and check whether any process is in the D state, as shown in the following figure:
ps -axjf
Note:
The D state refers to the uninterrupted sleep state. A process in this state cannot be terminated nor can it be exited by itself.
If there are many processes in the D state, restore the resources on which the processes depend or restart the operating system.
CPU usage of the kswapd0 process is high
Problem
Linux manages memory by using the pagination mechanism and sets aside a portion of the disk as virtual memory. kswapd0 is the process responsible for page replacement in the virtual memory management of the Linux system. When system memory becomes insufficient, kswapd0 will frequently replace pages, which will result in high CPU usage.
Solution
1. Run the following command and find the kswapd0 process.
top
2. Check the state of the kswapd0 process.
If the process is not in the D state and has been running for a long time and consuming too many CPU resources, perform step 3 to check the memory usage.
3.
Run commands such as vmstat, free, and ps to check how much memory is being consumed by processes in the system.
Based on the memory usage, restart the system or terminate safe but unnecessary processes. If the si and so values are also high, pages are frequently replaced in the system. If the physical memory of the current system can no longer meet your requirements, please consider upgrading your system memory.
Yes
No
Was this page helpful?