Rescue mode
Rescue mode is a mechanism that allows you to boot an instance in rescue mode in order to diagnose and repair issues with the operating system of the virtual machine.
This tool is useful when an instance has problems starting up or accessing remotely.
Procedure for Systems Launched from an Image¶
This applies when the following options were selected in the properties during virtual machine creation:
- In the OpenStack panel, select
Project --> Compute --> Instances --> "Rescue Instance" option
.
- In the next step, select the "systemrescue 10.02" image (other images/tools can be loaded and used). Leave the password field empty (this option is inactive).
- Access the instance via the VNC console:
- To exit rescue mode: in the OpenStack panel, select
Project --> Compute --> Instances --> "unrescue Instance" option
.
Procedure for Systems Launched from a Volume:¶
BST Region¶
- Remember or note down the name of the volume attached to the instance:
- The "root" disk cannot be detached from the instance, so you need to delete the virtual machine.
- The remaining volume should be connected to another virtual machine to perform repair procedures, depending on the problem.
- Similarly, detach the volume. Create a new virtual machine based on the repaired volume.
DCW Region¶
- CLI access required: CLI Client Configuration for OpenStack - For Linux (Ubuntu).
- Use the "systemrescue 10.02" ISO image ID f86f3b96-c5e6-4ca6-982b-c0cd694e3a54 for system repair.
- Access the instance via the VNC console.
Commands:
openstack image list | grep system - need to verify if the "systemrescue 10.02" image is available on the list
openstack server list - look for the ID of the virtual machine that requires repair
openstack --os-compute-api-version 2.90 server rescue --image IMAGE_ID VM_ID
Example:
openstack --os-compute-api-version 2.90 server rescue --image f86f3b96-c5e6-4ca6-982b-c0cd694e3a54 d01f669e-dd78-4f43-958d-944f63c35c07
After completed repair works:
openstack server unrescue SERVER_ID