rescue steps
- Force stop the failed ec2 instance
- Detach the root ebs volume
- Launch new ec2 instance need support Nitro System the instance type and attach the previous EBS volume
- Add new user in new ec2 instance and get the
/etc/shadows encrypt password copy to old device same file administrator user overide the line
- Detach the ebs volume from new instance and stop it
- Retach the ebs volume to first ec2 and start it use serial console connect to it and use overide the user credetials to fix some issue
- Or immediately fix some problem in new ec2
useful command or script
1# list the installed kernel
2dpkg --list | grep linux-image
3apt list --installed | egrep '^linux' | grep $(uname -r)
4# view some logs
5grep $(uname -r) /var/log/dpkg.log*
6# in centos
7rpm -qa --last | grep kernel
8
9# try to change grub to old boot
10vi /etc/default/grub
11# to modify some value
12GRUB_DEFAULT=0
13GRUB_TIMEOUT_STYLE=menu
14GRUB_TIMEOUT=10
15GRUB_TERMINAL="console serial"
16GRUB_SERIAL_COMMAND="serial —speed=115200“
17
18# remember to update grub make it work when reboot
19sudo update-grub
20
21# to view grub version may needs different command
22sudo grub2-mkconfig -o /boot/grub2/grub.cfg
23sudo grub-mkconfig -o /boot/grub/grub.cfg
24
25# to view kernel module status
26systemctl status systemd-modules-load.service
27
28# may be finally this command let u surprise
29dpkg --configure -a
The amazon linux 2 upgrade the kernel
1# list the kernel
2amazon-linux-extras list | grep kernel
3# disable the kernel
4amazon-linux-extras disable kernel-5.10 -y
5# install the new kernel
6amazon-linux-extras install kernel-5.15 -y
7# reboot
8reboot
9# confirmation
10rpm -qa | grep kernel
11uname -a
Reference