Change the size of reserved memory on a Linux instance - Elastic Compute Service

kdump is a Linux kernel crash dump mechanism that allows you to capture memory dump information when a kernel crash, such as a kernel panic, occurs in the operating system. This helps you analyze the cause of the crash. kdump requires an amount of reserved memory, which is called crashkernel memory, to ensure that sufficient memory is available to store kernel dump information even in the event of an operating system crash. This topic describes how to view and change the size of reserved memory on a Linux Elastic Compute Service (ECS) instance.

View the size of reserved memory in the operating system

To allow kdump to use reserved memory, you must add the crashkernel parameter to the GRand Unified Bootloader (GRUB) configuration file and specify the size of reserved memory for each memory segment.

Note

By default, the GRUB configuration file is stored in the /etc/default/grub directory. In Alibaba Cloud Linux 3, the GRUB configuration file is stored in the /usr/share/alinux-base-setup/cmdline directory.

When the operating system starts, GRUB passes the crashkernel parameter from the configuration file to the kernel to notify the kernel of the amount of memory that must be reserved for kdump on startup. You can view the /proc/cmdline file at system runtime to check whether the crashkernel parameter functions as expected.

Run the following command to check whether the crashkernel parameter is enabled:
```
cat /proc/cmdline
```
The following command output includes the configuration of the crashkernel parameter, which indicates that the crashkernel parameter is enabled.
crashkernel parameter
This section describes the crashkernel parameter. In this example, the value 0M-2G:0M,2G-8G:192M,8G-:256M is used.
- If the memory size of your ECS instance or the memory size of the operating system on the instance is less than or equal to 2 GB, no memory is reserved for the operating system because the amount of memory available for applications on the instance is insufficient. When the amount of memory in the operating system is less than or equal to 2 GB, you cannot use kdump.
- If the memory size of your ECS instance or the memory size of the operating system on the instance is greater than 2 GB but less than or equal to 8 GB, the operating system reserves 192 MB of memory.
- If the memory size of your ECS instance or the memory size of the operating system on the instance is greater than 8 GB, the operating system reserves 256 MB of memory.
Run the following command to view the size of reserved memory:
```
cat /sys/kernel/kexec_crash_size
```
The following command output is returned, which indicates the size of reserved memory. Unit: bytes. In this example, 201326592 is returned, which indicates that 192 MB of memory is reserved.

Release the reserved memory and disable kdump

If you do not need to enable kdump for your instance and want to return the reserved memory address space to the operating system to prevent resource waste, perform the following operations to release the reserved memory and disable kdump.

Important

If you release the reserved memory and disable kdump, the operating system may be unable to collect kernel crash information when the operating system fails. As a result, you cannot identify and fix kernel-related issues at the earliest opportunity. This affects the service continuity and stability of the instance. Proceed with caution.

Alibaba Cloud Linux and CentOS

Back up the GRUB configuration file.
- Run the following command in Alibaba Cloud Linux 2 or CentOS:
```
sudo cp /etc/default/grub /etc/default/grub.bak
```
- Run the following command in Alibaba Cloud Linux 3:
```
sudo cp /usr/share/alinux-base-setup/cmdline /usr/share/alinux-base-setup/cmdline.bak
```
Delete the configuration of the crashkernel parameter from the GRUB configuration file and save the file.
- Run the following command in Alibaba Cloud Linux 2 or CentOS:
```
sudo vim /etc/default/grub
```
  Find the crashkernel parameter on the GRUB_CMDLINE_LINUX line and delete the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M configuration. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
- Run the following command in Alibaba Cloud Linux 3:
```
sudo vim /usr/share/alinux-base-setup/cmdline
```
  Delete the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M line from the configuration file. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
Run the following command to update the GRUB configurations:
```
sudo grub2-mkconfig -o /boot/grub2/grub.cfg
```
Note
In distributions that are based on Red Hat Enterprise Linux (RHEL) 9, such as CentOS Stream 9, AlmaLinux 9, and Rocky Linux 9, run the sudo grub2-mkconfig -o /boot/grub2/grub.cfg --update-bls-cmdline command to update the GRUB configurations.

Run the following commands to disable and stop kdump:

sudo systemctl disable kdump
sudo systemctl stop kdump

Run the following command to restart the operating system for the changes to take effect:
```
sudo reboot
```
View the status of kdump and the crashkernel parameter to check whether the changes take effect.
1. Run the following command to view the status of kdump. If kdump is in the inactive (dead) state, kdump is disabled.
```
sudo systemctl status kdump
```
2. Run the following command to obtain the GRUB configuration file of the instance and check whether the configuration of the crashkernel parameter is deleted:
```
cat /proc/cmdline
```

Ubuntu and Debian

Run the following command to back up the GRUB configuration file:
```
sudo cp /etc/default/grub /etc/default/grub.bak
```
Run the following command to delete the configuration of the crashkernel parameter from the GRUB configuration file and save the file:
```
sudo vim /etc/default/grub
```
Find the crashkernel parameter on the GRUB_CMDLINE_LINUX line and delete the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M configuration. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
Run the following command to update the GRUB configurations:
```
sudo update-grub
```

Run the following commands to disable and stop kdump:

sudo systemctl disable kdump-tools
sudo systemctl stop kdump-tools

Run the following command to restart the operating system for the changes to take effect:
```
sudo reboot
```
View the status of kdump and the crashkernel parameter to check whether the changes take effect.
1. Run the following command to view the status of kdump. If kdump is in the inactive (dead) state, kdump is disabled.
```
sudo systemctl status kdump
```
2. Run the following command to obtain the GRUB configuration file of the instance and check whether the configuration of the crashkernel parameter is deleted:
```
cat /proc/cmdline
```

Enable memory reservation and kdump

If you want to use kdump to capture memory dump information of your instance, perform the following operations to add the configuration of the crashkernel parameter and enable kdump.

Alibaba Cloud Linux and CentOS

Back up the GRUB configuration file.
- Run the following command in Alibaba Cloud Linux 3:
```
sudo cp /usr/share/alinux-base-setup/cmdline /usr/share/alinux-base-setup/cmdline.bak
```
- Run the following command in Alibaba Cloud Linux 2 or CentOS:
```
sudo cp /etc/default/grub /etc/default/grub.bak
```
Add the configuration of the crashkernel parameter to the GRUB configuration file and save the file.
- Run the following command in Alibaba Cloud Linux 3:
```
sudo vim /usr/share/alinux-base-setup/cmdline
```
  Add the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M configuration to the configuration file. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
- Run the following command in Alibaba Cloud Linux 2 or CentOS:
```
sudo vim /etc/default/grub
```
  Find the crashkernel parameter on the GRUB_CMDLINE_LINUX line and add the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M configuration to the configuration file. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
Run the following command to update the GRUB configurations:
```
sudo grub2-mkconfig -o /boot/grub2/grub.cfg
```
Note
In distributions that are based on RHEL 9, such as CentOS Stream 9, AlmaLinux 9, and Rocky Linux 9, run the sudo grub2-mkconfig -o /boot/grub2/grub.cfg --update-bls-cmdline command to update the GRUB configurations.
Run the following command to enable kdump:
```
sudo systemctl enable kdump
```
Run the following command to restart the operating system for the changes to take effect:
```
sudo reboot
```
View the status of kdump and the crashkernel parameter to check whether the changes take effect.
1. Run the following command to view the status of kdump. If kdump is in the active (exited) state, kdump is enabled.
2. Run the following command to obtain the GRUB configuration file of the instance and check whether the configuration of the crashkernel parameter is added:
```
cat /proc/cmdline
```

Ubuntu and Debian

Run the following command to back up the GRUB configuration file:
```
sudo cp /etc/default/grub /etc/default/grub.bak
```
Run the following command to add the configuration of the crashkernel parameter to the GRUB configuration file and save the file:
```
sudo vim /etc/default/grub
```
Add the crashkernel=0M-2G:0M,2G-8G:192M,8G-128G:256M,128G-:384M configuration to the GRUB_CMDLINE_LINUX line. Press the Esc key, enter :wq, and then press the Enter key to save and close the configuration file.
Run the following command to update the GRUB configurations:
```
sudo update-grub
```

Run the following commands to enable and start kdump:

sudo systemctl enable kdump-tools
sudo systemctl start kdump-tools

Run the following command to restart the operating system for the changes to take effect:
```
sudo reboot
```
View the status of kdump and the crashkernel parameter to check whether the changes take effect.
1. Run the following command to view the status of kdump. If kdump is in the active (exited) state, kdump is enabled.
2. Run the following command to obtain the GRUB configuration file of the instance and check whether the configuration of the crashkernel parameter is added:
```
cat /proc/cmdline
```