It may happen that an instance refuses to start up or gets unreachable. From this point using a rescue image is the only way to recover the broken instance. This may happen for example in the following cases:
- Username and password data het been lost to gain access to the system
- Network configuration problems
- The system refuses to start properly
- A forced file check of the root partition
This manual describes how to upload an rescue image and boot an instance using the rescue boot image for recovery. This operation can only be performed using the OpenStack CLI toolkit.
Download an rescue image
Within OpenStack we offer 2 rescue images which are invisible within the OpenStack horizon image library.
For recovering Linux and Windows instances please use the SystemRescueCD available on OpenStack 1 and OpenStack 2.
SystemRescueCD voor openstack1 : 90b7c669-a980-4930-814d-ace760214bec SystemRescueCD voor openstack2 : f8b13182-17bc-45ef-b06e-a575189d447d
Windows server 2016 installer and recover disk is available only for openstack 2
Windows server 2016 : 126f413c-e832-4689-915d-d5f7e5f06292
Next, start the concerning instance in rescue mode using the uploaded rescue image and access the instance using the openstack console. After running the openstack command to start the instance rescue you will be provided with a username and password to access the instance rescue image. The user and password accounting data can defer depending on the rescue image, and may provide a own user login.
openstack server rescue --image [UUID_systemrescuecd] [UUID_INSTANCE]
Afterwards to start-up the instance disable the rescue mode. The instance will start-up automatically.
openstack server unrescue [UUID_INSTANCE]