Backup of virtual machines from storage snapshot
View Patent ↗A method for minimizing VMWARE™ snapshot lifetime, by reading data from a storage snapshot created on the VM storage side. The reading is performed by mounting the storage snapshot as a new datastore on an ESX(I) host and accessing the host using a standard VMWARE™ VDDK (vStorage API) method. A pre-configured source ESX(I) host contains a VM that needs to be backed up. A SAN storage connected to the ESX(I) host over iSCSI or FibreChannel or NFS protocols is used. The backup agent, which contains the proprietary code, is installed on a physical or a virtual machine connected to an ESX(I) host. This backup agent maintains the operations with the ESX(I) host and the SAN storage and coordinates the VM virtual disks data read and write operations. The backup agent contains the executable code, which provides compatibility with SAN storage and ESX(I) host.
1. A method for optimizing backup of a Virtual Machine (VM) from a snapshot, the method comprising:
installing a backup agent on a physical machine, wherein the backup agent is configured to access a VM host where a VM to be backed up resides;
creating a native snapshot of the VM by a native snapshot operation;
calling a first storage application programming interface (API) using the backup agent;
creating a hardware snapshot by the first storage API;
removing the native snapshot by a native storage API;
registering the hardware snapshot as a new datastore on the VM host;
reading a virtual disk of the VM into the new datastore using a second storage API;
writing data read from the virtual disk into a final backup archive destination; and
removing the hardware snapshot by a third storage API,
wherein the new datastore is a clone of an original native datastore, containing a state of the original native datastore at creation of the hardware snapshot.
2. The method of claim 1 , wherein the backup agent is installed on a second VM.
3. The method of claim of claim 1 , wherein the VM is a VMware™ VM.
4. The method of claim 1 , wherein the native snapshot operation is a VMware™ API.
5. The method of claim 1 , wherein the VM host is an ESX(i) host and the storage is a Storage Area Network (SAN) storage.
6. The method of claim 5 , further comprising removing SAN hardware snapshot from the SAN storage using SAN API.
7. The method of claim 5 , further comprising reading VM virtual disk data from the new datastore residing on the ESX(i) host using VMware™ Virtual Device Development Kit (VDDK).
8. The method of claim 1 , further comprising installing a backup agent on a VM different from the VM being backed up.
9. The method of claim 1 , wherein the VM host is running the same operating system as the VM.
10. The method of claim 1 , wherein the VM host is running a different operating system than the VM.
11. The method of claim 1 , wherein, a VMWare™ API is used for in the step of creating a snapshot of the VM.
12. A method for optimizing backup of a Virtual Machine (VM) from a snapshot, the method comprising:
installing a backup agent on a physical machine, wherein the backup agent is configured to access an ESX(i) host where a VM to be backed up resides;
creating a native snapshot of the VM using a VMware™ snapshot application programming interface (API);
calling a first Storage Area Network (SAN) storage API from the backup agent;
creating a SAN hardware snapshot by the first SAN storage API;
registering the hardware snapshot as a new datastore on the host;
reading a virtual disk of the VM from the new datastore using VMware™ vStorage API;
writing data read from the virtual disk into a final backup archive destination; and
removing the SAN hardware snapshot from the SAN storage using SAN API, wherein the new datastore is a clone of an original SAN datastore containing a state of the original SAN datastore at creation of the SAN hardware snapshot.
13. A system for optimizing backup of a Virtual Machine (VM) from a snapshot, the system comprising:
an ESX(i) host having VM to be backed running on it;
a physical machine having a backup agent running on it, the backup agent is configured to access the ESX(i) host;
a backup archive connected accessible by the backup agent;
a Storage Area Network (SAN) datastore connected to the ESX(i) host;
wherein the backup agent is configured to:
create a native snapshot of the VM using a VMware™ snapshot application programming interface (API);
call a SAN storage API;
create a SAN hardware snapshot by the SAN storage API;
register the SAN hardware snapshot as a new datastore on the ESX(i) host;
read a virtual disk of the VM from the new datastore using VMware™ vStorage API;
write data read from the virtual disk into the backup archive; and
remove the SAN hardware snapshot from the SAN datastore using SAN API, wherein the new datastore is a clone of the SAN datastore containing a state of the SAN datastore at creation of the SAN hardware snapshot.