PHD Virtual Backup for VMware vSphere

Storing Backups in the Cloud with CloudHook

PHD Virtual Backup's CloudHook technology provides multiple options for storing your backed up data in the cloud, including some of the most popular cloud storage providers like Amazon S3, Google Cloud Storage, RackSpace CloudFiles, and OpenStack.

In addition to writing backups to the cloud directly you can also use CloudHook to store archived copies of your backups on cloud storage by enabling Backup Archive. For details, see Configuring Backup Archive.

Note:    Using a cloud backup storage option requires version 6.2 or later and a CloudHook license. To obtain a CloudHook license, contact PHD Virtual (a CloudHook license is not required when running the trial version of the product).

PHD Virtual Backup enables cloud storage backups by first caching data locally on an attached virtual disk known as the Write Space. This is the same storage used for Instant VM Recovery. The Write Space disk must be large enough to handle the daily changed data you will be backing up, refer to the information in the configuration sections below for specific guidelines.

In some situations, cloud backups can provide an alternative to tape, allowing for long term, offsite storage in addition to regular backups to local or network storage. Some considerations should be made when deciding how cloud backups can best suit your needs, for instance, backing up to the cloud is highly dependant on the network speed and bandwidth available and in some cases may take additional time to complete when compared to backing up to other storage locations.

Configuring Cloud Backup Storage

The following steps describe how to enable and configure cloud backup storage with PHD Virtual Backup. These instructions assume you have already deployed and initially configured a PHD Virtual Backup Appliance.

Step 1: Apply the CloudHook license update

1. The license update is applied like any other PHD Virtual Backup Appliance update. Open the PHD Console to the Configuration area and select the PHD VBA you want to apply the update to in the dropdown menu at the top of the page. Note that the license update is not required while running the trial version of the product.
2. Click the Support tab.
3. In the Version Information area, click Upload Appliance Patch.
4. Select the CloudHook license update file (file type is .phd) then click Open.
5. When the update is applied, you will see the following text beneath your license details:

Step 2: Configure Write Space

1. Use the hypervisor client to attach a new virtual disk to the PHD VBA virtual machine by editing the virtual machine settings. This disk will be used as the Write Space disk to cache data being sent to cloud backup storage.

   A guideline for selecting the correct size Write Space disk is to calculate at least 50% of your daily changed data, then attach a disk that can accommodate that size. Typically, changed data is about 5% of the total of all virtual disks in your environment that are being backed up. So for example, if you were backing up ten VMs every day that totaled 1 TB, you should attach a virtual disk that is at least 25 GB (Changed data is approximately 5% of 1 TB, which is 50 GB; of which 50% is 25 GB). Write Space can be expanded at any time by adding additional attached virtual disks.

2. In the PHD Console, go to the Write Space tab and in the Storage Type dropdown menu, select Attached Disk.

3. Click Configure.

   

4. Use the Write Space configuration dialog to select the disk you added and move it from the Available Disks list to the Write Space Configuration list.
5. Click OK.

Step 3: Configure Cloud Backup Storage

1. Go to the Backup Storage tab and select the Cloud storage you will be using from the list of available options.

   

2. Enter your account credentials and the bucket name or container for your cloud storage. This is location where you will be storing your backups with your cloud storage provider. To obtain this information, you may need to log in to your cloud storage account and review the details. Additional folders can be created by including them after the bucket name or container. Note that Rackspace does not support creating additional sub-folders.

3. Creating sub-folders can be useful when writing to the same bucket or container with multiple PHD VBAs (if using Rackspace, you can use different File Prefixes to handle multiple PHD VBAs writing to the same bucket).

4. The File Prefix is a string that you supply that is added before the name of all files written to the cloud storage location. This is a required field and cannot contain any slashes. For example, when using Amazon S3, if the Storage Path and File Prefix were entered as, phdvb/mybackups and phd respectively, all backup files would be stored within the mybackups folder and prefixed with the value phd.

5. Finally, the Encryption Passphrase is a string you provide to secure your encrypted cloud backup data. In the case you need to attach a new PHD VBA to your cloud backup data store, you will need to include this passphrase along with the path and credentials to successfully mount the data store. Once set, the passphrase cannot be changed.

Using Cloud Storage

When determining if cloud storage is right for you, one main consideration should be the estimated backup window. Backing up to the cloud will be slower in most cases when compared to other backup storage types, so determining the length of time an average backup will take is crucial.

You can estimate this time by first determining the average changed data in a typical backup then dividing that by your transfer rate. If you are already running backups, the job logs will contain the number of bytes transferred. Transfer rates can be estimated using online speed test tools in most cases or you can check with your ISP. The result of this calculation should give you an estimate of the time required to backup to the cloud.

Note that the initial backup will take much longer when compared to subsequent backups that include changes only. For example, transferring 250 GB of actual used data with an upload speed of 30 Mbps would take about 9 hours to complete the first backup. Each backup thereafter would take a fraction of the time, a few minutes to an hour, as only changes from the initial backup are included.

Other Considerations

• For redundancy, consider using two PHD VBAs to perform backups - configuring one PHD VBA to back up locally and the other configured to back up to the cloud.for offsite backup redundancy. In this scenarios, backup jobs must be staggered, for instance, begin the local backup job at 8 PM then the cloud backup job at 10 PM.
• When configuring your backup plan using cloud, consider backing up critical VMs with fast RTO requirements locally and store your less critical VM backups in the cloud.

Cloud Storage Notes

• Cloud Storage backup data stores are not available as replication sources.
• When using cloud storage, Write Space must be an attached virtual disk. The option to use the VBA's backup data store is not available in the Write Space Storage Type dropdown menu when a cloud option is selected as the backup data store.
• To maximize the benefits of using cloud storage you should ensure you have a stable Internet connection and a reliably fast connection (Tier 1 or better is recommended).
• If the network connection is interrupted, the PHD VBA must re-mount the cloud storage. When re-mounting, a file consistency check is required to ensure the integrity of your backup data. The length of time required for this check depends on multiple factors including the size of the cache, the available network speed and bandwidth, and your individual cloud storage provider. In general, file consistency checks will take longer when using cloud storage when compared to other storage types. The stability of your network connection can directly impact how often file consistency checks are required.
• If you stop using a PHD VBA that was writing to the cloud, to avoid additional charges from your cloud provider, ensure no backup data is left behind. You can remove backup data using the Backup Catalog via the PHD Console if the PHD VBA is still available. If the PHD VBA is no longer available you will need to manually remove the backup data by logging in to your cloud storage account and using their interface.
• As with all other backups, you should periodically check your PHD VBA's health using the Tools area of the PHD Console to ensure they are performing optimally.