How does Scale-Out Cloud work?

The Cloud Vault is now simply a partner management user interface and API access point that your x360Recover appliances communicate with for coordinating data replication to the Cloud. 

• When an appliance registers a new protected system for Cloud replication, the vault creates a database record for the device and requests a new storage object from the Scale-Out Cloud. 
• When the appliance is ready to perform replication to the Cloud, it starts a transaction with the vault, requesting to begin data transmission. 
• The vault replies with specific URL of the storage server the given protected system resides in and its storage object ID. 
• The appliance then performs the data transfer directly onto the storage server, bypassing the vault entirely. 

Multiple protected systems on a single appliance likely reside on totally different physical storage nodes distributed within our data center.

What are the features and benefits of Scale-Out Cloud?

Scale-Out Cloud benefits:  

1. Performance: 
Compared to our legacy storage system, the raw storage performance of Scale-Out Cloud is many orders of magnitude higher, with the potential for unlimited expansion and growth. 

With legacy vault storage, partners were limited by the constraints of storage, networking, and CPU resources of a single host computer.  This enforced an upper limit to the maximum number of protected systems and total storage capacity that a single vault could efficiently handle.  As a result, many partners had to deploy and manage multiple Cloud vaults to accommodate all their protected systems as their business grew over time.  Overloading of Cloud vaults would occur, leading to replication lag as they struggled to commit data to storage.

2. Scalability: 
Scale-Out Cloud distributes the heavily resource-intensive parts of cloud storage across multiple physical storage nodes that can be expanded to scale indefinitely. 

Partners with many Cloud vaults are now able to consolidate all their protected systems onto a single vault. This simplifies replication configuration management and reduces operating costs and complexity by eliminating multiple vaults that no longer need to be managed and maintained.

Scale-Out Cloud features:

1. Storage and endpoint capacity:

Scale-Out Cloud provides fast, horizontally-scaling storage within the Axcient cloud.

2. Superior performance:

The distributed architecture of Scale-Out Cloud allows for fast storage performance across an ever-expanding storage cluster, eliminating replication lag issues caused by data center storage bottlenecks.

3. Instant virtualization:

Scale-Out Cloud interoperates with our Continuity Cloud architecture to provide for instant virtualization of your protected systems in our Disaster Recovery Cloud.  With the click of a button, your protected systems can be exported to iSCSI targets, providing virtual disks that are instantly available to your Cloud Continuity hypervisor.  Protected systems can be immediately booted as a virtual machine in the Cloud.

4. Reduced total-cost-of-ownership (TCO):

By eliminating the need to maintain multiple Cloud vaults, Scale-Out Cloud reduces total cost of ownership. Partners now need only a single vault, simplifying replication configuration and eliminating the maintenance of many unnecessary devices. 

Recover a protected system walkthrough

1. After you have logged in to the vault, navigate to the Protected Systems page on the lefthand navigation. Click on the name of the System Name of the Protected System you wish to recover.

2. From the Protected Systems Details page, locate the Snapshots section at the bottom of the page and locate the snapshot you wish to recover from.

3. Under the Actions menu for the snapshot that you selected, click on Start iSCSI.

4. When the iSCSI settings screen appears, select Live Mode and enter the encryption password for the protected system (if any) then click Start.

5. The Snapshot actions buttons will be updated. Notice that three of the options for the selected snapshot will be unavailable. The buttons for Start VM, Mount, and Export will be grayed out and the Start iSCSI action button will become Stop iSCSI, as shown below:

6. Next, click the iSCSI option from the lefthand navigation. 

7. Locate the protected system you are recovering and note these iSCSI connection details:

• Portal - iSCSI target address to be connected to
• CHAP User - username required by iSCSI target 
• Password - password required by iSCSI target 
• IQN Names - each Windows Volume (drive) will have an IQN identifier listed
  

8. Once you have the iSCSI target connection details noted, continue configuration from the Continuity Cloud node.

Using the connection credentials you were provided by the Axcient Support team when you requested access to the Continuity Cloud, connect to the node using an RDP client.


9. From within the Windows Control Panel, expand Administrative Tools and open iSCSI Initiator.

10.  Select the Discovery tab and click Discover Portal...

11. Enter the portal address displayed on the vault iSCSI page into the IP address or DNS name field.

12. After you enter the connection address, click Advanced.

13. Check the Enable CHAP log on box and then enter the Name and Target secret password provided on the Vault iSCSI page.

Click OK to close the Advanced Settings window.
Click OK to close the Discover Target Portal window.


14. Select the Targets tab on the iSCSI Initiator Properties window and locate the IQN name(s) matching the disk(s) belonging to the protected system you are recovering. 

For each IQN, click Connect...


...and then click Advanced to configure connection properties.



15. Check the Enable CHAP log on box and then enter the Name and Target secret password provided on the Vault iSCSI page.

Click OK to close the Advanced Settings window.
Click OK to close the Connect to Target window. 

Repeat for any additional IQN’s being configured. 

16. The iSCSI disk(s) exported by the x360Recover vault are now connected as locally attached devices on the Windows Continuity Cloud hypervisor node and are ready to be used to create a virtual machine.

• To determine exactly which local disk volume a given IQN name is associated with, select the IQN from the targets page within iSCSI Initiator Properties and click Properties.
  

17. On the Devices page, note the disk's Name from the Name column. 

This corresponds to the disk name within Disk Management on the Windows host.

Create a virtual machine walkthrough 

1. On the Continuity Cloud node, open Hyper-V Manager from within Control Panel -> Administrative Tools.

a. Right-click the host from the left menu pane, navigate to new, and click Virtual Machine to open the New Virtual Machine Wizard. 

b. Click Before You Begin on the left-hand navigation panel of the New Virtual Machine Wizard screen, go to the bottom of the page and click Next to begin.

2.  Enter the host Name of the protected system being recovered and click Next.

3. If the system you are recovering is a legacy / MBR BIOS-style machine, select Generation 1. If the system you are recovering is an EFI BIOS-style machine, select Generation 2.

Click Next to continue. 



4. Enter the amount of RAM you would like to assign to the virtual machine and click Next to continue. 

Note: To determine how much total system memory is available to the Continuity Cloud node, open the System tool from Control Panel. It is recommended to leave at least 4GB of RAM for the host operating system.



5. Select the network profile to be connected with the virtual machine. Typically, you will select Internal-LAN, which will place the virtual machine behind the pfSense virtual firewall inside your Continuity Cloud environment.

Click Next to continue. 

6. Select the option to Attach a virtual hard disk later and click Next to continue.

7. Review the summary information and click Finish to create the virtual machine.

Configure the virtual machine walkthrough

1. Locate your new Virtual Machine within the Virtual Machines pane of Hyper-V Manager.

2. Right-click the virtual machine and select Settings

• If this is a Generation 1 virtual machine, select IDE Controller 0
• If this is a generation 2 virtual machine, select SCSI Controller
  (IDE Controller will not be present)

Select Hard Drive and click Add.

3. Under Hard Drive,  select Physical Hard Disk and then select the physical disk to be connected. (This disk was identified earlier when attaching the iSCSI targets.)  Click OK to continue.

Generation 1 Example: IDE Controller 0

Generation 2 Example: SCSI Controller

What is a 4K Advanced Format disk?

Until relatively recently, all computer disk drives were manufactured with 512 bytes of physical block size for data storage.  

In 2010, manufacturers shipped the first prototype 4K Advanced Format drives to hardware vendors for experimental use. 4K Advanced Format drives are able to store 4096 bytes (4K) in each block. Since 2010, these 4K Advanced Format drives have slowly entered the market in production. 

The physical space saved by combining (eight) 512-byte blocks into a single 4096-byte block is about 7-11% of total platter space. This space savings is due to combining the redundant block headers into a single set.

Why does this matter?

Block size determines the smallest amount of data storage that a file can consume. 

• A 1-byte file consumes an entire block on the disk.
  • On a traditional drive, this 1-byte file would consume 512 bytes.
  • On a 4K Advanced Format disk, this 1-byte file requires 4096-bytes. 

Disk structures, partitions, and volume sizes are also calculated as:

(Number of blocks)  x  (Block size)

So, it is important to understand this size difference when manipulating disk data. 

Most importantly, partitions must begin and end on a block boundary, and the underlying storage controller driver must be 4K-Advanced-Format-block-aware to process these drives correctly.

How has 4K Advanced Format affected virtualization and recovery? 

The legacy compatibility storage controller emulation in KVM does not understand or support 4K Advanced Format block devices.

This means that it was not possible to construct a 4K Advanced Format virtual disk to be compatible with the default IDE controller emulation used for virtualizing a protected system in x360Recover. 

4K Advanced Format drives are only supported using VirtIO storage controller modes within KVM.  

Windows does not include native driver support for VirtIO. This has historically meant that VirtIO drivers must be installed into our virtual machine images via a complicated manual process  - before such machines could be booted as a virtual machine.