HP Matrix 7.2 KVM Private Cloud Backup and Restore

Table Of Contents

Manual Recovery Actions

Inconsistencies can occur as a result of a restore if the state of the managed resources at the time of backup and

restore are different. In cases where the backup does not contain enough data about the current managed resources,

a manual recovery should be performed. Alerts and audit messages provided by the Matrix KVM Private Cloud can

be used to identify the recovery action that should be performed (Refer to Appendix C).

Cloud Connectivity Subnet

If the cloud connectivity subnet range is added or modified after backup, it needs to be updated again. Ensure that

the same range information and metadata is entered after the restore.

Private Subnet

If the private subnet range is modified after backup, it needs to be updated again. Ensure that the same range

information and metadata is entered after the restore.

Instances

If any new instances (or orphaned instances) were created after the backup, they cannot be brought back onto the

KVM Admin Console. An alert will be presented in the Activity page of the KVM Admin Console for each of these

instances after restore. Refer to “Orphaned Instance Alert” (page 24) in Appendix C. The instance needs to be re-

provisioned by taking a snapshot as described below:

Step 1: Remove public IP association with the instance.

1. Log in to the host for the instance specified in the alert.

2. Use the Private IP provided in the alert and check for floating IP entries using:

iptables -L -t nat | grep <private_ip>

A sample output would look like the following:

DNAT all -- anywhere <floating_ip> to:<private_ip>

SNAT all -- <private_ip> anywhere to:<floating_ip>

3. Use the floating IP obtained above and remove all references of floating IPs from the NAT table.

iptables -t nat -D nova-network-OUTPUT -d <floating_ip>/32 -j DNAT --to-

destination <private_ip>

iptables -t nat -D nova-network-PREROUTING -d <floating_ip>/32 -j DNAT --to-

destination <private_ip>

iptables -t nat -D nova-network-float-snat -s <private_ip>/32 -j SNAT --to-

source <floating_ip>

Step 2: Take a snapshot of the VM and remove the instance from the host.

1. Log in to the host for the instance specified in the alert.

2. Find the disk path of the instance from the domain name obtained in the alert.

virsh dumpxml <domainname>

The disk path can be found in the ‘file’ attribute at /devices/disk/source of the output XML. By default,

the disk path is in the following format:

/var/lib/nova/instances/<domain-name>/disk

Using the disk path specified above, create a snapshot of the image file as follows:

qemu-img snapshot -c <snapshot_name> <disk_path>