HPE Reference Configuration for

HPE Cloud Bank Storage with

Microsoft Azure
Solution overview, best practices, and recommendations

Reference Architecture
Reference Architecture

Contents
Executive summary ................................................................................................................................................................................................................................................................................................................................ 3
Solution overview ..................................................................................................................................................................................................................................................................................................................................... 3
HPE Cloud Bank Catalyst stores and HPE StoreOnce Catalyst stores comparison ............................................................................................................................................................. 4
Microsoft Azure Blob storage................................................................................................................................................................................................................................................................................................. 4
HPE Cloud Bank stores configured for Microsoft Azure Blob storage .............................................................................................................................................................................................5
HPE Cloud Bank direct restore and optimized restore ................................................................................................................................................................................................................................... 6
HPE StoreOnce housekeeping..............................................................................................................................................................................................................................................................................................6
Micro Focus Data Protector software ............................................................................................................................................................................................................................................................................. 7
Solution components............................................................................................................................................................................................................................................................................................................................7
Micro Focus Data Protector Cell Manager ................................................................................................................................................................................................................................................................. 7
HPE StoreOnce 6600 backup appliance ..................................................................................................................................................................................................................................................................... 7
HPE StoreOnce VSA .......................................................................................................................................................................................................................................................................................................................8
HPE 3PAR StoreServ 8400 storage ................................................................................................................................................................................................................................................................................8
Network configuration ..................................................................................................................................................................................................................................................................................................................8
HPE Common Test Automation Framework ...........................................................................................................................................................................................................................................................8
Environment setup diagram .................................................................................................................................................................................................................................................................................................... 9
Best practices and configuration guidance for HPE Cloud Bank ................................................................................................................................................................................................................. 9
HPE Cloud Bank prerequisites ..............................................................................................................................................................................................................................................................................................9
Configuring the HPE Cloud Bank proxy server ................................................................................................................................................................................................................................................. 10
Creating an HPE StoreOnce Cloud Bank store .................................................................................................................................................................................................................................................. 10
Viewing the HPE Cloud Bank store summary ..................................................................................................................................................................................................................................................... 13
Viewing bandwidth utilization to and from the cloud .................................................................................................................................................................................................................................. 13
Configuring an HPE Cloud Bank device with Micro Focus Data Protector .............................................................................................................................................................................. 14
Catalyst Copy to an HPE Cloud Bank store with Micro Focus Data Protector ..................................................................................................................................................................... 15
HPE Cloud Bank disaster recovery ...................................................................................................................................................................................................................................................................................... 16
HPE Cloud Bank re-attach .................................................................................................................................................................................................................................................................................................... 16
HPE Cloud Bank detach .......................................................................................................................................................................................................................................................................................................... 17
HPE Cloud Bank multi-reader capability ................................................................................................................................................................................................................................................................. 18
Test results and cost comparisons ...................................................................................................................................................................................................................................................................................... 19
Four week backup test: weekly full and daily incremental backups ................................................................................................................................................................................................ 19
Four week backup test: weekly full and daily cumulative backups .................................................................................................................................................................................................. 20
Four week backup test: daily full backups .............................................................................................................................................................................................................................................................. 21
HPE Cloud Bank optimized and direct restore cost comparison....................................................................................................................................................................................................... 22
Testing summary and recommendations ............................................................................................................................................................................................................................................................... 22
Summary ...................................................................................................................................................................................................................................................................................................................................................... 23
Implementing a proof-of-concept......................................................................................................................................................................................................................................................................................... 23
Resources and additional links ................................................................................................................................................................................................................................................................................................ 24
Reference Architecture Page 3

Executive summary
Data is at the core of all organizations and protecting it is more important than ever. The risks of temporarily or permanently losing access to
data are widely documented. Backup remains the key process to mitigate this risk. As data volumes increase, organizations are looking for ways
to efficiently and cost-effectively manage backup footprints within their data centers. The question of where to store enterprise data is essential.
There is a move to use public cloud services to take advantage of generally affordable and flexible data storage, but critical business reasons may
exist to keep data on-site. Since choosing between cloud, on-premises, or a hybrid solution requires an analysis outside the scope of this paper,
we will focus on the cloud storage use case. Refer to the Hewlett Packard Enterprise Decision guide: Public cloud versus on-prem storage for an
informative reference document covering some of the additional choices.

HPE Cloud Bank Storage is a feature of HPE StoreOnce systems designed to lower long-term data retention costs by leveraging cloud-based or
on-premises object storage. It is an extension of HPE StoreOnce Catalyst that uses external object storage to store backup data. It is available to
all backup applications compatible with HPE StoreOnce Catalyst and HPE StoreOnce Catalyst plug-ins. Deploying HPE Cloud Bank Storage
amplifies the low costs of object storage with HPE StoreOnce systems deduplication to enable cost effective, long-term backup data retention.

With support for multiple on-premises and cloud-based object storage targets, such as Microsoft® Azure Blob storage, Scality RING software, and
Amazon Web Services (AWS), HPE Cloud Bank Storage can be deployed without having to change backup processes, reducing the learning
curve associated with a new technology.

Most data needs to persist longer than infrastructure components. For this reason, access to data stored in HPE Cloud Bank Storage is not
dependent on the HPE StoreOnce system. This also means in the event that an HPE StoreOnce system becomes unavailable or unusable, the
data in the cloud is not lost. An HPE Cloud Bank store can be accessed by reattaching to a different HPE StoreOnce system. Using HPE
StoreOnce Catalyst Copy, data on the HPE Cloud Bank store can be copied to a local Catalyst store, allowing data to be recovered from that
location. The result is that HPE Cloud Bank Storage is a versatile and robust backup data store that can be recovered even in the event of an
HPE StoreOnce system hardware failure.

This Reference Configuration features HPE Cloud Bank Storage with Microsoft Azure Blob storage as the object storage target and Micro Focus
Data Protector as the data mover. This paper demonstrates the benefits of HPE Cloud Bank Storage, along with key use cases. For test results
supporting the cost savings that can be realized using the solution documented here, see the Test results and cost comparisons section, later in
this paper. For more information about the HPE Scality Object Storage solution, refer to HPE Reference Configuration for Enterprise Active
Archive Backup to HPE Scalable Object Storage with Scality RING.
Target audience: Presales consultants, solution architects, storage operators and administrators who design, implement, and maintain common
backup tasks, and others who want to learn more about HPE Cloud Bank Storage with Microsoft Azure.

Solution overview
HPE Cloud Bank Storage (HPE Cloud Bank) is an HPE StoreOnce systems (HPE StoreOnce) feature that allows HPE StoreOnce Catalyst stores
to use external object storage. In use, an HPE Cloud Bank store behaves almost identically to an HPE Catalyst store. This common behavior
enables administrators to utilize HPE StoreOnce deduplication technology when sending, storing, and retrieving public or private cloud storage
data. The key difference is that an HPE Cloud Bank store is supported as a Catalyst copy target, not as a backup target. The key HPE Cloud Bank
use cases are as follows:

• Long-term retention and archive of backups

• Offsite copies of backups for disaster recovery
• Detached HPE Cloud Bank stores for very long-term backup data retention

A key consideration when using public cloud object storage is the cost of uploading and downloading data. Deduplicating the data before writing
to cloud storage significantly reduces the amount of data uploaded. Consequently, retrieving this non-redundant data requires less download
usage. Through deduplication, HPE Cloud Bank reduces the cost of storing the data in the cloud, as well as the cost of writing and accessing the
cloud-stored backup data.
Reference Architecture Page 4

The following are key advantages of HPE Cloud Bank:

• Lower cost
– Optimizes bandwidth use by sending, storing, and retrieving only unique data to and from cloud-based storage for lower Total Cost of
Ownership (TCO).
– Increased capacity in the same management domain: stores up to two times (2x) the maximum configurable capacity of an HPE StoreOnce
system. For example, an HPE StoreOnce 5100 can support up to 216 TB of local usable backup storage and up to 432 TB of HPE Cloud
Bank storage, while HPE Cloud Bank storage capacity can be extended without limit by detaching HPE Cloud Bank stores.
• Flexibility
– Choice of multiple object-storage vendors.
– Broad number of supported backup/recovery Independent Software Vendors (ISVs) and enterprise applications (using HPE StoreOnce
Catalyst plug-ins).
– Orchestrate the optimal balance between performance and cost with automated, intelligent multi-tiered data protection, from on-premises
arrays to the public cloud.
• Simplicity
– Seamlessly move backup data to the cloud of your choice without the need for a separate cloud gateway or virtual appliance.
– Existing backup applications are the single management point for backup, copy, retention, and restore.
– When used in conjunction with HPE StoreOnce Catalyst Plug-ins, integration with native database backup tools gives database
administrators (DBAs) complete visibility and control over data-protection tasks from a familiar user interface.
• Reduced risk
– In case of failure, administrators can reconnect HPE Cloud Bank to an alternate local or cloud-based HPE StoreOnce system for simple,
encrypted, and reliable disaster recovery.
– There is no need for a physical HPE StoreOnce system; backups on HPE Cloud Bank can be restored to a local HPE StoreOnce Virtual
Storage Appliance (HPE StoreOnce VSA).

HPE Cloud Bank Catalyst stores and HPE StoreOnce Catalyst stores comparison
HPE Cloud Bank stores and HPE StoreOnce Catalyst stores are created and behave in similar ways. Any data-protection software that supports
HPE StoreOnce Catalyst Copy functionality can create backup copies on an HPE Cloud Bank store. One fundamental distinction between the two
stores is that HPE Cloud Bank stores write deduplicated data to public or private cloud storage, as opposed to a regular HPE StoreOnce Catalyst
store where data is written to disk in the HPE StoreOnce system. The objects written to the Cloud Service Provider (CSP) contain all of the
information necessary to reconstruct the backup data within the HPE Cloud Bank store.

Microsoft Azure Blob storage

Microsoft Azure Blob storage enables administrators to store all of their unstructured data within the cloud. Data can be scaled up or down as the
need for storing data changes. Microsoft Azure Blob storage account options consist of hot, cool, and archive storage tiers that can be selected
based on how often the data in the cloud is accessed and what type of data is being stored, backed up, or retrieved.
Hot storage tier
The hot storage tier is a price model optimized for storing data that is accessed frequently. For this tier, writing and retrieving data is the least
expensive, but reserving cloud storage capacity is the most expensive relative to other tiers.
Cool storage tier
The cool storage tier is a price model optimized for storing infrequently accessed data. Reserving cloud storage capacity is less expensive, but
writing and retrieving data is more expensive for this tier.
Archive storage tier
The archive storage tier allows an administrator to reserve cloud-storage capacity and data writes to the cloud at the cheapest rate. However,
data retrieval is the most expensive for this tier.
Reference Architecture Page 5

Note
The archive storage tier is only supported with HPE Cloud Bank when the store is detached from the HPE StoreOnce.

HPE Cloud Bank stores configured for Microsoft Azure Blob storage
HPE Cloud Bank works with a number of object storage vendors, including Microsoft Azure, Scality RING, Amazon Web Services (AWS), and S3
compatible CSPs. For this solution, Microsoft Azure Blob Storage was chosen as the CSP.

An HPE Cloud Bank store configured for Microsoft Azure Blob storage provides offsite protection by enabling data to be backed up to a
Microsoft Azure Blob storage container on the cloud over http or https protocols. Microsoft Azure allows large amounts of data to be stored
within Block Blobs that vary in size depending on the type and the quantity of the data. As data is accessed or overwritten within the container,
the modified data for the associated Block Blob is updated. Figure 1 shows an example Microsoft Azure Blob storage container with contents and
properties, as shown in the Microsoft Azure menu.

Figure 1. Example of a Microsoft Azure Blob storage container

Reference Architecture Page 6

HPE Cloud Bank direct restore and optimized restore

HPE Cloud Bank data can be recovered through either a direct restore or an optimized restore, as shown in Figure 2.
Direct restore
A direct restore is a restore read from the HPE Cloud Bank store. In this case, all of the data to be restored is read directly from Microsoft Azure,
before being written to the restore location, even if some of the data is duplicated. The result is higher costs from the increased reads. In a
scenario where an administrator is only recovering a small amount of data from a backup, it may be more cost effective to read that amount of
data directly from Microsoft Azure.
Optimized restore
An optimized restore is a two-step process. The first step is a Catalyst Copy from the HPE Cloud Bank store to a local Catalyst store. The second
step is a Restore from that local Catalyst store. In this case, only the data chunks that do not exist on the local Catalyst store are read, reducing
the amount of data read from Microsoft Azure, and resulting in lower costs. Additionally, if a restore is interrupted for any reason, data would not
have to be retrieved from Microsoft Azure for the second time.

Figure 2. Direct restore and optimized restore data flow

HPE StoreOnce housekeeping

Housekeeping is a necessary step in any deduplication storage process, required at some time after the host system performs any deletion or
overwrite process. The HPE StoreOnce stores unique “chunks” of data that are referenced with an index count. When these chunks are seen
again in subsequent backups, the index count of that specific chunk is incremented, but the chunk itself remains unaltered. Thus, the HPE
StoreOnce is only storing unique data.

Housekeeping is an ongoing process that works to reclaim disk space by continuously reorganizing these data chunks. When a Catalyst item is
deleted, the index count of the relevant chunks is decremented. If the reference to a particular chunk becomes zero, it is marked for deletion.
Because GET operations can be costly for cloud storage, the housekeeping process has been optimized to reduce capacity used in the cloud
without the need for GET operations.
Reference Architecture Page 7

In the event of a network or cloud service interruption, where connectivity to the cloud provider is lost, any pending housekeeping is unable to
proceed. If the outage is longer than 24 hours, housekeeping may enter a warning state.

Housekeeping is an important process to maximize the deduplication efficiency of the appliance. An administrator must make sure that the
housekeeping process has enough time to complete. Running backup, restore, replication, and HPE StoreOnce Catalyst store operations with no
break (that is, 24 hours a day) can result in housekeeping never being completed.

Micro Focus Data Protector software

HPE Cloud Bank Catalyst stores can be configured with any backup application software that allows HPE StoreOnce Catalyst Copy operations.
For this Reference Configuration, Micro Focus Data Protector was chosen as the backup application. To see a full list of backup applications and
plug-ins that support Catalyst Copy operations, refer to the HPE Data Availability, Protection and Retention Compatibility Matrix.

Micro Focus Data Protector is a backup solution that provides reliable data protection and high accessibility for your organization’s fast-growing
business data. Micro Focus Data Protector offers comprehensive backup, object copy (Catalyst Copy), and restore functionality specifically
tailored to enterprise environments. Figure 3 shows the environment setup for HPE Cloud Bank with Micro Focus Data Protector and Microsoft
Azure Blob storage.

Figure 3. Environment setup for HPE Cloud Bank with Microsoft Azure and Micro Focus Data Protector

Solution components
The environment setup used the components below to effectively utilize HPE Cloud Bank.

Micro Focus Data Protector Cell Manager

An HPE ProLiant DL380 Generation 9 (Gen9) 32-core (2 x Intel® Xeon® CPU E5-2698 v3 @ 2.30 GHz) server was configured with 512 GB of
physical memory.

Micro Focus Data Protector 10.00.0306 was installed on a Windows Server 2016 operating system. This server was designated as the gateway
server for all sections of the test. Micro Focus Data Protector provides an easy-to-use console to perform a variety of data protection tasks,
including HPE Catalyst Copy jobs, and other object operations. HPE Cloud Bank stores are configured in Micro Focus Data Protector as backup-
to-disk devices.

HPE StoreOnce 6600 backup appliance

The HPE StoreOnce 6600 backup appliance used firmware version 3.18.0. The backup appliance used one couplet with two drive enclosures
containing 36 x 4 TB drives per shelf.

The HPE StoreOnce 6600 backup appliance acted as the primary backup device for all portions of this test. Housekeeping, replication, and HPE
StoreOnce Catalyst were licensed and enabled.
Reference Architecture Page 8

HPE StoreOnce VSA

An HPE ProLiant DL380p Generation8 (Gen8) 12-core (2 x Intel® Xeon® CPU E5-2667 @ 2.90GHz) server was configured with 1 TB of HPE
3PAR StoreServ 8400 storage and 64 GB of physical memory.

The HPE StoreOnce VSA software version was 3.18.0. This HPE StoreOnce VSA acted as a disaster recovery device when the primary HPE
StoreOnce device was simulated to have a hardware failure. Housekeeping, replication, and HPE StoreOnce Catalyst were licensed and enabled.

HPE 3PAR StoreServ 8400 storage

The HPE 3PAR StoreServ 8400 storage array, using 24 x 1.2 TB hard disk drives and OS version 3.2.2.612, was connected to a 16 Gb fibre
channel backend fabric.

Network configuration
The local management network was configured to 1 GbE. The local data network was configured to 10 GbE. The primary storage was configured
via 16Gb FC SAN. The internet connection to the CSP was configured to less than 1 GbE.

HPE Common Test Automation Framework

HPE Common Test Automation Framework (HPE CTAF) is the Perl-based testing tool used to generate, modify, and verify the test datasets
used in solution testing for this Reference Configuration.

The HPE StoreOnce was configured to use a separate HPE StoreOnce Catalyst store per use case. The HPE StoreOnce Catalyst store was
recreated before the start of each use case. Housekeeping data was stored and monitored, as well.
Reference Architecture Page 9

Environment setup diagram

Figure 4 shows the environment setup for this solution.

Figure 4. Solution test environment setup

Best practices and configuration guidance for HPE Cloud Bank

HPE Cloud Bank prerequisites
Before HPE StoreOnce can utilize external object storage, an administrator must create an HPE Cloud Bank store, which is configured in the HPE
StoreOnce Management Console. Creating an HPE Cloud Bank store requires an HPE StoreOnce Catalyst license and HPE Cloud Bank licenses
to cover the planned capacity of the HPE Cloud Bank stores. For existing HPE Cloud Bank stores, HPE StoreOnce provides the option to connect
with read-write or read-only access. These options are explained in more detail in the HPE Cloud Bank disaster recovery section.

To utilize HPE Cloud Bank functionality, the following criteria must be met:

• An HPE StoreOnce system based on an HPE ProLiant Gen9 or newer server(s) with the following:
– Version 3.18.0 or newer.
– HPE StoreOnce Catalyst license.
– Sufficient HPE Cloud Bank licenses for the capacity of the connected stores.
– Sufficient local storage for reconnected HPE Cloud Bank store metadata: around 1% of the store capacity.
– A cloud-hosted HPE StoreOnce VSA could also be used if the VSA contains sufficient storage.
Reference Architecture Page 10

For reconnection to an HPE Cloud Bank store, the following criteria must also be met:

• Less than 60 days since the HPE Cloud Bank store was connected to the unavailable HPE StoreOnce system.
• The encryption key and encryption password are required.

Note
Administrators should also consider memory expansion, depending on how close to maximum capacity they plan to utilize local storage.

Configuring the HPE Cloud Bank proxy server

Proxy servers can be enabled to connect an HPE Cloud Bank store to the public cloud service. Follow these steps to configure the proxy server:

1. Log in to the HPE StoreOnce Management Console.

2. Go to StoreOnce  StoreOnce Catalyst.
3. Under StoreOnce Catalyst, select Settings.
4. Under the Proxy Server section, select Edit.

5. Input the host address and the port number of the proxy server, as shown in Figure 5. Optionally, input the username and password for the
proxy server.
6. Click Update.

Figure 5. Configuring proxy server settings for the HPE Cloud Bank server in the HPE StoreOnce Management Console

Creating an HPE StoreOnce Cloud Bank store

An administrator can create an HPE StoreOnce Cloud Bank store by following the steps below. It is necessary to associate the HPE Cloud Bank
store with a cloud storage account (in this case, Microsoft Azure). All of the usual HPE Catalyst Store options are available.

Follow these steps to create an HPE Cloud Bank store with Microsoft Azure Blob storage:

1. Log in to the HPE StoreOnce Management Console.

2. In the navigator pane on the left side of the window, go to StoreOnce  StoreOnce Catalyst  Stores.
3. Click Create at the top-right of the window. The New Store window will appear, as shown in Figure 6.
4. If multiple Service Sets are displayed, select the Service Set to be used for the HPE Cloud Bank store and click OK.
5. Create a Name and Description for the HPE Cloud Bank store and set other configuration settings as necessary.
Reference Architecture Page 11

6. Check the box next to Cloud Bank Storage to configure as an HPE Cloud Bank store.
7. From the Cloud Service Provider dropdown, select Azure.
8. Enter the Account Name of your Microsoft Azure storage account.
9. Enter the Account Key of your Microsoft Azure storage account. (These are 512-bit storage access keys created in Microsoft Azure.)
10. Enter the Container Name of the Microsoft Azure storage account container.
11. Click Create.

Figure 6. Creating a new Cloud Bank store with Azure

12. A prompt is displayed to enter an Encryption Key password. Enter a password and click OK.

Important
The encryption password must be stored in a safe place, since it is necessary for reconnection.
Reference Architecture Page 12

13. If Store Encryption is enabled, an encryption key is generated and a password is required for that encryption key. Figure 7 and Figure 8
highlight these steps.

Figure 7. Creating a password for the encryption key for an HPE Cloud Bank store

Figure 8. Generating an encryption key after creating an HPE Cloud Bank store

Note
Direct backups to an HPE Cloud Bank store are not supported for performance reasons. Backups are written to a regular HPE StoreOnce Catalyst
Store and then copied to the HPE Cloud Bank store, using Catalyst Copy operations.
Reference Architecture Page 13

Notice
When selecting the checkbox for HPE Cloud Bank, Store Encryption is enabled by default. Hewlett Packard Enterprise recommends enabling
Store Encryption with HPE Cloud Bank stores to protect the data on the external storage.

Viewing the HPE Cloud Bank store summary

The HPE StoreOnce Catalyst stores section in the HPE StoreOnce Management Console displays the HPE StoreOnce Catalyst and HPE Cloud
Bank stores. HPE Cloud Bank specific information shows the capacity of data stored externally to HPE StoreOnce in the Size in Cloud column in
Microsoft Azure and any other configured and object storage targets. Figure 9 shows the HPE Cloud Bank store summary.

Figure 9. HPE Cloud Bank store summary

Viewing bandwidth utilization to and from the cloud

As of HPE StoreOnce version 3.18, a new Cloud Bank tab has been added to Reports in Reporting Central of the HPE StoreOnce Management
Console. This tab is used to monitor upload and download bandwidth for the HPE Cloud Bank store, as well as the number of requests that are
being sent to and received from the cloud. Individual HPE Cloud Bank stores can be selected for viewing and time periods can be selected, as
required. Figure 10 and Figure 11 display HPE Cloud Bank reports.

Figure 10. HPE Cloud Bank upload bandwidth and requests report summary
Reference Architecture Page 14

Figure 11 displays the HPE Cloud Bank report summary.

Figure 11. HPE Cloud Bank download bandwidth and requests report summary

Note
Throughput will vary based on a number of factors, including internet connection bandwidth, stability, and latency.

Configuring an HPE Cloud Bank device with Micro Focus Data Protector
Regardless of the backup application, a Catalyst Copy job must be configured to write data to HPE Cloud Bank. To perform Catalyst Copy jobs to
an HPE Cloud Bank store in Micro Focus Data Protector, an administrator must first configure HPE Cloud Bank as a backup-to-disk device.
Follow these steps:

1. Log in to the Micro Focus Data Protector Manager.

2. From the drop-down menu, select Devices & Media.
3. Right-click Devices and select Add Device.

4. Provide a name and description for the device.

5. From the drop-down menu, select Backup To Disk for Device Type. Click Next.
6. Provide the StoreOnce System address of the HPE StoreOnce.
7. Provide the Client ID and Password of the HPE StoreOnce.
8. Click Select/Create Store and select the HPE Cloud Bank Catalyst store from the list. Click OK.
9. Click Add to configure a gateway device. Click OK.
10. Click Next to continue.
11. Optionally, configure quotas for the store. Click Next.
12. Click Finish.
Reference Architecture Page 15

Catalyst Copy to an HPE Cloud Bank store with Micro Focus Data Protector
The primary method of transferring backup data into an HPE Cloud Bank store involves HPE StoreOnce Catalyst copy technology that already
exists in many backup applications. HPE StoreOnce Catalyst Copy transfers deduplicated data from one Catalyst store to another. To run an HPE
StoreOnce Catalyst Copy in Micro Focus Data Protector, a Catalyst Copy job must be configured. The data protection application feeds data to
the Catalyst store, then copies the data to the CSP by initiating a Catalyst Copy. Figure 12 shows the data flow for backup and copy operations
with Micro Focus Data Protector, HPE Cloud Bank, and Microsoft Azure.

Figure 12. Backup and copy operations data flow

To perform a Catalyst Copy to an HPE Cloud Bank store with Micro Focus Data Protector, a backup specification must be configured with an
additional post-backup job after the initial full or incremental backup. To add a Catalyst Copy job to a backup specification, complete the
following steps:

1. Log in to the Micro Focus Data Protector Manager.

2. From the drop-down menu, select Object Operations.
3. Go to Copy  Object copy  Automated  Post Backup.
4. Right-click Post Backup and select Add.
5. Select the backup specification and click Next.

6. As necessary, complete the following optional steps:

a. Select to apply the copy specification to an existing copy specification. Click Next.
b. Select to apply the copy specification to an existing consolidation specification. Click Next.
c. Apply object filters to the selection. Click Next.
d. Apply a library filter to the selection (default is all libraries). Click Next.
e. Select the source device(s) to apply the copy job (default is “automatic device selection”). Click Next.
7. Select the destination device, which in this case is the HPE Cloud Bank store. Highlight the gateway device and click Properties. In the
device properties, select the media pool for the HPE Cloud Bank store from the drop-down menu and click OK. Click Next.
8. Under Options, make sure that Use replication is checked. Optionally, change other copy specification options. Click Next.
9. Click Save as and name the copy specification. Click OK.
Reference Architecture Page 16

HPE Cloud Bank disaster recovery

An important feature of HPE Cloud Bank is that it is not dependent on the availability of the host HPE StoreOnce System. In the event of an HPE
StoreOnce becoming unavailable or unusable, HPE Cloud Bank can be recovered from a separate HPE StoreOnce system and still be part of the
overall disaster recovery process. Unlike other cloud-connector systems with deduplication technology, data in HPE Cloud Bank is still accessible
if the host HPE StoreOnce is unavailable. The data-protection software can connect to a new HPE StoreOnce system to manage the recovery.

HPE Cloud Bank re-attach

In the event of disaster recovery, an existing HPE Cloud Bank store can be reconnected to another HPE StoreOnce device and backed data
recovered using that device. This is possible because the local metadata on the HPE StoreOnce system, which is primarily used for matching, is
also duplicated across various objects in the cloud storage. In the HPE StoreOnce Management Console, an administrator must select Connect in
the HPE StoreOnce Catalyst menu to initiate the re-attach process. After supplying the cloud-store credentials, the window shown in Figure 13 is
displayed.

Note
If the HPE StoreOnce Catalyst Store has been connected to an HPE StoreOnce system within the last 24 hours, the HPE StoreOnce disables
Connect (Read-Write).

An HPE Cloud Bank store can be reattached using the Connect (Read-Only) option or the Connect (Read-Write) option, as shown at the bottom
of Figure 13.

Figure 13. Connecting an existing HPE Cloud Bank store to an HPE StoreOnce system
Reference Architecture Page 17

Connect (Read-Only)
The Connect (Read-Only) process is similar to the Connect (Read-Write) process, with some key differences. An administrator can use the
Connect (Read-Only) process to read data that has been written to another HPE Cloud Bank store. This process does not require 24-hours
before the data can be utilized. There are two use cases for the Connect (Read/Only) functionality.

1. Read data from a store that has been detached.

2. Read data from a store attached to another HPE StoreOnce system.

In the HPE StoreOnce Management Console, an administrator clicks Connect (Read-Only) to initiate the connection process. Once the process is
complete, the store will be in a read-only state and does not require HPE Cloud Bank licenses to access and use.
Connect (Read-Write)
The Connect (Read-Write) process can be executed up to 60 days from the date that a store was disconnected, and is temporarily disabled if the
Catalyst store was connected to an HPE StoreOnce system within the last 24 hours. After 60 days from disconnection, reconnection is possible
only with intervention by HPE Pointnext, an HPE IT services organization. 1 The Connect (Read-Write) process requires that the host system has
HPE Cloud Bank Read/Write licenses available that are equal or greater than the capacity of the connecting data. Once the process of connecting
is complete, the store will be in an online state.

HPE Cloud Bank detach

If long-term storage is required, an administrator can choose to detach an HPE Cloud Bank store. Detaching an HPE Cloud Bank store removes it
from the HPE StoreOnce Management Console and returns any HPE Cloud Bank Capacity licenses to the HPE StoreOnce system for later use. A
detached store can be reattached within 60 days after disconnection or moved to the archive tier of Microsoft Azure storage. Archive tier data is
optimized to be the least expensive access tier for long-term data storage, see Microsoft Azure Blob storage. Once in an archived state, a cost
based on the number of read operations is present when moving a Blob from archive tier access to a cool or hot tier access.

Note
HPE Cloud Bank detach is a licensed feature that requires HPE Cloud Bank capacity licenses. Detaching an HPE Cloud Bank store recycles HPE
Cloud Bank licenses for later use.

As shown in Table 1, three options are presented when removing an HPE Cloud Bank store: disconnect, detach, and delete.
Table 1. Disconnect, detach, and delete options
Disconnect Detach Delete

Cloud data state after executing Read-write Read-only Unavailable

Removes objects from the cloud storage No No No
Removes store from the HPE StoreOnce Yes Yes Yes
system
Options after executing Connect (read-write) after 24 hours and Connect (read-only) within 60 days No longer accessible
up to 60 days after disconnect after disconnect
Connect (read-only) within 60 days after
disconnect

1 To find out more about HPE Pointnext services, go to https://www.hpe.com/us/en/services.html.

Reference Architecture Page 18

HPE Cloud Bank multi-reader capability

HPE Cloud Bank multi-reader functionality grants the ability to read from a single Microsoft Azure Blob storage container, or another object
storage target concurrently, using multiple backup applications. Only one backup application at a time can write to the object storage target,
while multiple backup applications from other locations can simultaneously access the storage data in read-only mode, as shown in Figure 14.

Figure 14. HPE Cloud Bank multi-reader functionality

HPE Cloud Bank multi-reader functionality addresses several potential use cases:

• Streamlined data archiving while minimizing impact to a production environment

• Mining backup data for analysis without impacting production backup and restore processes
• Restore or restore testing outside of a production environment
• Disaster recovery testing outside of a production environment
• Seeding a new site with an entirely separate SAN environment
When more than one HPE StoreOnce system is connected to the object store, the first HPE StoreOnce can be set for read/write access. Access
for additional systems is limited to read-only. The HPE Cloud Bank licensed capacity for each HPE StoreOnce must be equal to or greater than
the physical capacity of the data held in the HPE Cloud Bank storage. If the capacity of the HPE Cloud Bank store exceeds the licensed capacity,
Reference Architecture Page 19

a warning is issued. An HPE StoreOnce grants a seven-day grace period to allow an administrator to apply the required HPE Cloud Bank license.
After that time, the multi-reader feature is disabled until the appropriate HPE Cloud Bank license is applied.

Test results and cost comparisons

The cloud cost savings achieved by HPE StoreOnce deduplication technologies are apparent after running the first backup. To give tangible
evidence of the improvement that deduplication has on the cost of reading from or writing to Microsoft Azure Blob storage, a test was conducted
as follows:

• A four-week backup schedule was simulated by running 28 backups in total, with every seventh backup being a full backup, and everything
else being an incremental backup.
• The size of the dataset being backed up was slightly larger than 100 GB with an initial compression ratio of 2:1.
• One percent of the data was changed between each backup.
• The four-week backup schedule test was repeated two more times with cumulative backups and exclusively full backups.
• The dataset existed on an HPE 3PAR LUN.
• A Microsoft Azure account was created with a cool access tier subscription.
• The charts in this section compare the costs to write the deduplicated dataset to cloud storage with the costs to write that same dataset to
cloud storage if it was not deduplicated.
• The costs were calculated using the Microsoft Azure Pricing Calculator tool at https://azure.microsoft.com/en-us/pricing/calculator/ and the
Microsoft Azure Block Blob pricing information at https://azure.microsoft.com/en-us/pricing/details/storage/blobs/. At the time this document
was published, the Microsoft Azure costs, using Locally Redundant Storage (LRS) to the West US region, were as shown in Table 2.
Table 2. Microsoft Azure costs at the time of publication
Hot access tier Cool access tier

First 50 TB (per GB, per month) $0.0208 $0.0152

Next 450 TB (per GB, per month) $0.02 $0.0152
Over 500 TB (per GB, per month) $0.0192 $0.0152
Data Retrieval (per GB) Free $0.01
Data Write (per GB) Free $0.003

Four week backup test: weekly full and daily incremental backups
Figure 11, earlier in this document, shows the immediate cost savings expected from using HPE Cloud Bank. After 28 backups using Microsoft
Azure with HPE Cloud Bank, the data write cost was $0.25. By comparison, without HPE Cloud Bank, the data write cost was $1.37. In this
instance, the cost was reduced by 81%.
Reference Architecture Page 20

The cost savings are proportional to the approximate size of the dataset. For example, this test uses a 100 GB dataset, resulting in a data write
cost of $1.37 without HPE Cloud Bank. If a 5 TB dataset is used instead, the data write cost is projected to be $12.50 with HPE Cloud Bank and
$68.50 without HPE Cloud Bank. Figure 15 shows physical data written to the cloud, which was measured between each backup. In this figure,
every seventh backup is a full backup, and everything else is an incremental backup.

Figure 15. Physical data written to Microsoft Azure Blob storage with weekly full and daily incremental backups

Four week backup test: weekly full and daily cumulative backups
An administrator with a similar dataset and environment could expect to pay around $0.26 in data write costs after 28 backups with HPE
StoreOnce deduplication technology, compared to a $1.54 without HPE Cloud Bank. The result is a solution that is 83% less expensive after 28
backups, as shown in Figure 16, where physical data written to the cloud was measured between each backup. In this figure, every seventh
backup is a full backup, and everything else is a cumulative backup.

Figure 16. Physical data written to Microsoft Azure Blob storage with weekly full and daily cumulative backups.
Reference Architecture Page 21

Four week backup test: daily full backups

HPE Cloud Bank is highly recommended when conducting daily full backups to Microsoft Azure. As shown in Figure 17, after 28 full backups in a
similar environment, an administrator using HPE Cloud Bank would spend $0.27 in data write costs, compared to $9.00 without HPE StoreOnce
deduplication technology. In this case, the cost has been reduced by 97%. Again, the cost savings are proportional to the size of the dataset.
Figure 17 shows physical data written to the cloud, which was measured between each backup.

Figure 17. Physical data written to Microsoft Azure Blob storage with daily full backups.
Reference Architecture Page 22

HPE Cloud Bank optimized and direct restore cost comparison

Cost savings are also present when restoring data from the cloud. In this test, a 100 GB dataset was restored from Microsoft Azure, first directly,
and then by using a Catalyst Copy job from HPE Cloud Bank to a regular HPE StoreOnce Catalyst store. The resulting download utilization was
about 32 GB less, resulting in a restore that cost 29% less than a direct restore, as shown in Figure 18.

Figure 18. Cost savings with HPE Cloud Bank restore

Testing summary and recommendations

• For all backup schedules, whether they are full backups, cumulative backups, or incremental backups, an administrator can achieve cost
savings using HPE Cloud Bank to back up to Microsoft Azure or other cloud-storage services.
• The following results have been observed with the test dataset and scenarios described above:
– After 28 daily incremental backups and weekly full backups, HPE Cloud Bank was 81% less expensive compared to not using deduplication
technology. (See “Four week backup test: weekly full and daily incremental backups”.)
– After 28 daily cumulative backups and weekly full backups, HPE Cloud Bank was 83% less expensive compared to not using deduplication
technology. (See “Four week backup test: weekly full and daily cumulative backups”.)
– After 28 full backups, HPE Cloud Bank was 97% less expensive compared to not using deduplication technology. (See “Four week backup
test: daily full backups”.)
– Using a Catalyst Copy job was 29% less expensive. When restoring from Microsoft Azure or another CSP, if the ISV allows, use a Catalyst
Copy job rather than a direct restore. (See “HPE Cloud Bank disaster recovery”.)
Reference Architecture Page 23

Summary
As cloud services become increasingly integrated into data center infrastructure, the need to efficiently utilize cloud technologies has become
critical. HPE Cloud Bank offers an end-to-end solution that significantly reduces the costs associated with backing up to and restoring from the
cloud. HPE StoreOnce provides streamlined deduplication technology, requiring only unique, deduplicated data to be sent over the network and
stored in the external storage.

With the combination of HPE StoreOnce deduplication and object storage, the HPE Cloud Bank Storage solution enables high-performance
backup and recovery, as well as long-term backup retention. In addition to significant cost-savings, HPE Cloud Bank provides a secure offsite
location for disaster recovery, with adaptable options. In the event of an HPE StoreOnce becoming unavailable or unusable, the data in the cloud
is not lost. HPE Cloud Bank can re-attach to a separate HPE StoreOnce or HPE StoreOnce VSA to provide reliable and flexible data recovery.
(For more information, see “HPE Cloud Bank disaster recovery”.)

By leveraging HPE StoreOnce Catalyst capabilities already integrated into a broad collection of backup applications and database backup tools,
HPE Cloud Bank Storage provides an easy-to-implement hybrid data protection solution that amplifies the cost benefits of object storage with
HPE StoreOnce deduplication. The result is an economical, efficient, flexible, and secure backup and recovery infrastructure for short-term and
long-term backup data.

Implementing a proof-of-concept
As a matter of best practice for all deployments, HPE recommends implementing a proof-of-concept using a test environment that matches as
closely as possible the planned production environment. In this way, appropriate performance and scalability characterizations can be obtained.
For help with a proof-of-concept, contact an HPE Services representative (hpe.com/us/en/services/consulting.html) or your HPE partner.
Reference Architecture Page 24

Resources and additional links

HPE StoreOnce systems and HPE Cloud Bank Storage
HPE Cloud Bank solution brief, hpe.com/us/en/pdfViewer.html?resource=/content/hpe/country/us/en/resources/storage/Solution-brief/cloud-
bank

HPE Cloud Bank Storage: A Data-Protection Solution You Can Bank On, community.hpe.com/t5/Around-the-Storage-Block/HPE-Cloud-Bank-
Storage-A-Data-Protection-Solution-You-Can-Bank/ba-p/6965903

HPE StoreOnce Data Protection Backup Appliances, hpe.com/us/en/storage/storeonce.html

HPE Cloud Bank Storage Video, youtube.com/watch?v=Uz04jZPONeE

HPE Cloud Bank Storage Chalk Talk, youtube.com/watch?v=tmUM_5NA2xk

HPE Cloud Bank Storage Blogs

Multi-tiered data protection with HPE Cloud Bank Storage for HPE 3PAR, community.hpe.com/t5/Around-the-Storage-Block/Multi-tiered-data-
protection-with-HPE-Cloud-Bank-Storage-for/ba-p/6991279

From All-Flash to the Public Cloud – HPE Cloud Bank Storage for HPE 3PAR, news.hpe.com/from-all-flash-to-the-public-cloud-hpe-cloud-bank-
storage-for-hpe-3par/

Cloud backup made simple with HPE Cloud Bank Storage, community.hpe.com/t5/Around-the-Storage-Block/Cloud-backup-made-simple-with-
HPE-Cloud-Bank-Storage/ba-p/6989075

Microsoft Azure
Microsoft Azure, https://azure.microsoft.com

Micro Focus Data Protector

Micro Focus Data Protector, software.microfocus.com/en-us/products/data-protector-backup-recovery-software/overview

Additional references
Public Cloud Versus On-Prem Storage, hpe.com/us/en/insights/articles/decision-guide-public-cloud-versus-on-prem-storage-1701.html

HPE Pointnext, hpe.com/us/en/services.html

HPE Data Availability, Protection and Retention Compatibility Matrix, support.hpe.com/hpsc/doc/public/display?docId=c04616269

HPE Reference Configuration for Enterprise Active Archive Backup to HPE Scalable Object Storage with Scality RING,
h20195.www2.hpe.com/V2/GetDocument.aspx?docname=a00038287enw

To help us improve our documents, please provide feedback at hpe.com/contact/feedback.

Learn more at hpe.com/storage

© Copyright 2018 Hewlett Packard Enterprise Development LP. The information contained herein is subject to change without notice.
The only warranties for Hewlett Packard Enterprise products and services are set forth in the express warranty statements accompanying
such products and services. Nothing herein should be construed as constituting an additional warranty. Hewlett Packard Enterprise shall
not be liable for technical or editorial errors or omissions contained herein.

Microsoft, Windows, and Windows Server are either registered trademarks or trademarks of Microsoft Corporation in the United States
and/or other countries. Intel and Intel Xeon are trademarks of Intel Corporation in the U.S. and other countries.

a00043317enw, March 2018