HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware

Confidential | For Training Purposes Only

HPE GreenLake for Block Storage

Built on HPE Alletra Storage MP
Hardware
Appendix A

Course ID 0001207682
Use of this material to deliver training without prior written permission from HPE is prohibited.
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 2

 Copyright 2024 Hewlett Packard Enterprise Development LP

The information contained herein is subject to change without notice. The only warranties for HPE
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.
This is an HPE copyrighted work that may not be reproduced without the written permission of Hewlett
Packard Enterprise. You may not use these materials to deliver training to any person outside of your
organization without the written permission of HPE.
Microsoft® and Windows® are either registered trademarks or trademarks of Microsoft Corporation in
the United States and/or other countries.
VMware® is a registered trademark or trademark of VMware, Inc. in the United States and/or other
jurisdictions.
Oracle® is a registered trademark of Oracle and/or its affiliates.
Linux® is the registered trademark of Linus Torvalds in the U.S. and other countries.
All third-party marks are property of their respective owners.
Printed in U.S.
Appendix A: HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Rev. 24.11
Confidential | For Training Purposes Only

Export Compliance Agreement

Export Requirements. You may not export or re-export products subject to this agreement in violation
of any applicable laws or regulations.
Without limiting the generality of the foregoing, products subject to this agreement may not be
exported, re-exported, otherwise transferred to or within (or to a national or resident of) countries
under U.S. economic embargo and/or sanction including the following countries and regions: Cuba, Iran,
North Korea, Sudan, Syria, and the Crimea region of the Ukraine. This list is subject to change.
In addition, products subject to this agreement may not be exported, re-exported, or otherwise
transferred to persons or entities on applicable Sanctioned Parties Lists, such as listed on the U.S.
Department of Commerce Denied Persons List; U.S. Department of Commerce Entity List (15 CFR 744,
Supplement 4); U.S. Treasury Department Designated/Blocked Nationals exclusion list; U.S. State
Department Debarred Parties List; Office of Foreign Assets Control Specially Designated Nationals
Lists; or to parties directly or indirectly involved in the development or production of nuclear, chemical,
or biological weapons, missiles, rocket systems, or unmanned air vehicles as specified in the U.S. Export
Administration Regulations (15 CFR 744); or to parties directly or indirectly involved in the financing,
commission or support of terrorist activities.
By accepting this agreement you confirm that you are not located in (or a national or resident of) any
country/region under U.S. embargo or sanction; not identified on any U.S. Department of Commerce
Denied Persons List, Entity List, US State Department Debarred Parties List or Treasury Department
Designated Nationals exclusion list; not directly or indirectly involved in the development or production
of nuclear, chemical, biological weapons, missiles, rocket systems, or unmanned air vehicles as specified
in the U.S. Export Administration Regulations (15 CFR 744), and not directly or indirectly involved in
the financing, commission or support of terrorist activities.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 3

Learning objectives
After completing this module, you should be able to:
• Distinguish the updated HPE Storage product positioning
• Explain the architecture and hardware building blocks of HPE GreenLake for Block Storage built on HPE Alletra
Storage MP hardware
• Differentiate storage management and monitoring options for HPE GreenLake for Block Storage
• Plan HPE GreenLake for Block Storage data reduction optimizations, upgrades, and data migration

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 4

Updated HPE Storage product positioning

Before explaining HPE GreenLake for Block Storage Built on HPE Alletra Storage MP hardware, let us start by looking
at the updated HPE Storage portfolio for block storage.

HPE block storage portfolio

HPE hardware offerings for block storage

HPE offers a wide range of storage products with different capabilities suitable for different customers and their
needs. Most of these products are also supported by HPE GreenLake for Block Storage management with more to
come.

Enlargement of left-hand portion of graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 5

Enlargement of right-hand portion of graphic

Storage portfolio high-level attributes

HPE GreenLake for Block
HPE Primera HPE Alletra 9000 HPE Alletra 6000 HPE Alletra 5000 HPE XP8
Storage (Alletra MP)
Centralized cloud-based management

Intent-based provisioning

HPE InfoSight Advanced AI for infrastructure HPE InfoSight

All-flash NVMe all-flash Hybrid Hybrid

Extreme scale
<3.2 PiB (raw) <1.1 PiB (raw) <3.2 PiB (raw) <1004 TiB (raw) <1145 TiB (raw) (>3.2 PiB up to 64 PiB) (raw)
Extreme perf.
<1.8 M IOPS Enterprise-class performance <2.1 M IOPS Best $/Performance Best price (up to 28 M IOPS)
End-to-end NVMe Back-end NVMe End-to-end NVMe

Active/active for latency-sensitive applications (rapid controller failover) Active/standby Active/active

100% availability 6-nines availability Extreme avail. (2N failure)

Stringent DR with 0 RPO/0 RTO Advanced HA/DR set-up

Dark site Dark site

Comparison of HPE Storage products and their features comparison

Use the above table to position HPE Storage products based on their attributes and how they can help your
customers solve their problems.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 6

HPE GreenLake for Block Storage built on HPE Alletra Storage MP hardware
This section introduces HPE Alletra Storage MP hardware as well as HPE GreenLake for Block Storage products
based on HPE Alletra Storage MP.

HPE GreenLake for Block Storage: Built on HPE Alletra Storage MP

HPE GreenLake
cloud control plane

Cloud-delivered software HPE GreenLake

for Block Storage OS

One common
disaggregated
hardware platform
HPE Alletra Storage MP

A high-level view of the components of HPE GreenLake with Alletra MP

HPE Alletra Storage MP is a new modular, disaggregated multi-protocol storage platform with 100% Availability
Guarantee. It consists of standardized composable building blocks—compute (node), capacity (JBOF, or Just a Bunch
of Flash) and switches. This enables the deploying of HPE GreenLake for Block Storage or HPE GreenLake for File
Storage on a common hardware platform through the HPE GreenLake cloud platform (GLCP), thereby collapsing silos
and improving total cost of ownership/return on investment (TCO/ROI).

Note
HPE Alletra Storage MP is a hardware component (compute node/capacity JBOF) of an HPE GreenLake for Block
Storage or HPE GreenLake for File Storage product and cannot be ordered/purchased separately.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 7

HPE GreenLake for Block Storage

HPE Alletra Storage MP chassis

Thanks to the advanced virtualization of HPE GreenLake for Block Storage OS, the hardware is abstracted to provide
great flexibility and easy upgrades, as well as adoption of new items like larger drives, new connectivity options, and
other features.
An expanded HPE GreenLake for Block Storage delivers mission-critical storage at mid-range economics. Built on the
HPE Alletra Storage MP, this modular, disaggregated storage is delivered and managed via the HPE GLCP. This
enhanced cloud data service brings the cloud experience, efficient scale, extreme resiliency, and performance to your
customer’s mission-critical apps—from large databases to modern cloud-native apps and consolidated mixed
workloads—at an affordable midrange price point.
HPE GreenLake for Block Storage leverages the capabilities of HPE Alletra Storage MP, an all-NVMe common HPE
Storage platform. This platform is uniquely designed to abstract data from media, offering a cloud-managed, data-
centric, and self-optimizing architecture. It is engineered for user convenience, featuring customer self-installation -
updating -repair capabilities. This platform includes all-inclusive array software, enhancing its reliability and
functionality.

Features of the HPE GreenLake for Block Storage series arrays

The HPE GreenLake for Block Storage series arrays offer a range of features that emphasize their advanced
technological capabilities. These arrays run on a software defined HPE GreenLake for Block Storage OS, which is
integrated with the common HPE Alletra Storage MP platform.
A standout feature is the 100% Availability Guarantee, ensuring consistent and reliable performance. The arrays are
equipped with an all-NVMe front and back end, including dual-ported NVMe enterprise SSDs, and support 32 Gb FC
and 25 GbE front end, with 100 GbE Remote Direct Memory Access (RDMA) over Converged Ethernet (RoCE) RoCE-
v2 back end. They offer thinly provisioned storage with customer-selectable enhanced data reduction and an
enhanced software architecture that includes single-node virtual volume (VV) ownership and write-through caching.
While the secure cloud management is a preferred option, on-array UI is also available for air-gapped sites. Both
feature replication and Peer Persistence support. The series also boasts comprehensive management integrations
and allows customers to self-install, update, and repair their systems1. Online, data-in-place upgradability is another
notable feature, and the HPE GreenLake subscription includes all array software2, adding to the overall value and
efficiency of these arrays.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 8

Notes
1
Most customers prefer to have the systems maintained by HPE or their trusted HPE partner.
2
HPE Data Encryption is excluded because of legal restrictions in some countries.

HPE Alletra Storage MP—Architecture

2, 4, or 8-node systems
Drives are dedicated to a node-pair
Node interconnect: ASIC to ASIC over PCIe

Node 0 Node 2

Node 1 Node 3

Drive enclosure Drive enclosure

Drive enclosure Drive enclosure

Traditional HPE 3PAR node-pair design

The specific HPE 3PAR architecture is replaced with a set of modular building blocks, based on a commodity
platform:
• Based on a standardized HPE hardware building block, the HPE Alletra Storage MP
• Two-node system interconnect goes via the backplane using a 25 GbE mezzanine adapter
• System managed from Data Services Cloud Console (DSCC)/GLCP and/or from the local UI

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 9

2-node, 8-core system 2-node, 16-core system 2-node, 32-core system

Up to 24 NVMe drives Up to 2 JBOFs directly connected With up to 2 JBOFs directly connected
With up to 72 NVMe drives With up to 2 JBOFs / 72 NVMe drives

IOM 1 IOM IOM 1 IOM IOM IOM

Internal NVMe drives Internal NVMe drives Internal NVMe drives

IOM IOM IOM IOM

NVMe JBOF drives NVMe JBOF drives

IOM IOM IOM IOM

NVMe JBOF drives NVMe JBOF drives

HPE GreenLake for Block Storage MP design

Storage system 1 IOM interconnects over the backplane between 25 GbE mezzanine cards
2 IOM interconnects cabled between 100 GbE OCP cards
2U HPE Alletra Storage MP chassis FC or IP host connections

Legend for previous graphic

Note
IOM = IO Module

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 10

HPE Alletra Storage MP hardware

This section focuses on HPE Alletra Storage MP hardware.

HPE Alletra Storage MP—The basics

Standard building-blocks SDS personalities Potential storage systems

Compute / Controller
Block or file Stand-alone system

HPE Alletra Storage MP chassis Capacity / JBOF

with versatile hardware configurations NVMe, RoCEv2

Stand-alone system
with JBOFs
Compute + capacity together
Storage array

HPE Alletra Storage MP configuration shown as building blocks

Here are the steps for ordering HPE Alletra MP hardware:

1. Step 1: Choose the storage chassis and controller nodes:
The storage chassis is a 2U 2-node chassis that includes 24 bays for SFF NVMe drives. The storage chassis does
not include any controller nodes or power supplies. One must be ordered for each configuration.
a. HPE GreenLake for Block Storage controller nodes include:
I. HPE Alletra Storage MP 256 GB 8-core Block Controller Node
II. HPE Alletra Storage MP 256 GB 16-core Block Controller Node
III. HPE Alletra Storage MP 512 GB 32-core Block Controller Node
b. Two controller node SKUs must be ordered per chassis. Both controller nodes need to be of the same type.
Different controller node types cannot be mixed in the same chassis. Each controller node contains two Open
Compute Project (OCP) slots for host bus adapters, or HBAs, (slots 3 and 4), and slots 1 and 2 are used for
expansion. Each 32-core Block Controller Node requires 100 GbE 2-port OCP HBA in slot 1.
2. Step 2: Choose power supplies. Various AC and DC models are available.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 11

3. Step 3: Choose adapters:

a. HBAs are rarely used for connection to hosts. They can be ordered standalone to be installed in the field, or
they can be factory integrated into controller nodes. HPE GreenLake for Block Storage does not have any
built-in host ports; therefore, any configuration needs to have at least one HBA per controller node. For
Remote Copy over IP (RCIP), the replication HBAs can be ordered as an option.
b. Each controller node must have at least one FC HBA or one 10/25 GbE HBA for internet Small Computer
Systems Interface (iSCSI) host connectivity. A maximum of two FC HBAs can be ordered per controller node. A
maximum of one 10/25 GbE HBA can be ordered per controller node.
4. Step 4: Choose drives. Drives are orderable at the time the array is purchased or can be added in the future when
additional capacity is required. HPE GreenLake for Block Storage supports SSDs with NVMe interface and triple-
level cell (TLC) NAND technology.
5. Step 5: Choose an optional NVMe expansion shelf (JBOF). The HPE Alletra Storage MP NVMe all-flash expansion
shelf (JBOF) is a 2U chassis to add additional system capacity. The expansion shelf supports eight to 24 NVMe
SSDs, in two-drive increments with up to two expansion shelves per system. The 8-core systems do not support
expansion shelves.
6. Step 6: Choose cables for host connectivity. HPE GreenLake for Block Storage requires cables for host
connectivity and replication. The OM4 fiber cables are used for host connectivity and Peer Motion.
7. Step 7: Choose a Bluetooth connectivity kit option. HPE Alletra Storage MP supports a Chassis Discovery Module
(CDM) that enables the chassis to connect to the DSCC without a controller node present. The chassis also has an
optional Bluetooth setup and configuration.
8. Step 8: Choose racking options. HPE GreenLake for Block Storage is compatible with most industry standard 4-
post Edge Infrastructure Automation (EIA) 19-inch racks with square mounting holes. HPE GreenLake for Block
Storage can be factory-configured and shipped in a rack or shipped without a rack for field integration into an
existing rack. The racks used for factory integration are the HPE G2 Advanced Series Racks or the HPE G2
Enterprise Series Racks.
9. Step 9: Choose software. HPE provides an extensive selection of features in the HPE GreenLake for Block
Storage OS software, including Virtual Copy, Remote Copy, Priority Optimization, Peer Motion, and Peer
Persistence, and so on. Data encryption is offered separately with a License To Use (LTU) at a cost of $1. HPE
GreenLake for Block Storage includes a subscription to HPE GreenLake for Block Storage software and support
software as a service (SaaS) that includes all-inclusive software features for the specified raw capacity and term,
and it enables cloud-based management of the array from the HPE DSCC and access to data services and related
supports.
10. Step 10: Choose support and installation services/options. Self-installation is possible.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 12

Enlargement of left-hand side of graphic

Enlargement of middle and right-hand side of graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 13

HPE Alletra Storage MP chassis

HPE Alletra Storage MP 2U chassis properties

The HPE Alletra Storage MP chassis is a 2U 2-node chassis that includes 24 bays for SFF NVMe drives. The storage
chassis does not include any controller nodes or power supplies. One must be ordered for each configuration.

Note
CDM is mostly reserved for factory and HPE field engineers use.

Note
TPM = Trusted Platform Module
OCP = Open Compute Project
EDSFF = Enterprise and Datacenter Standard Form Factor
U.2 = 2.5-inch form factor and compliant with the PCI Express SFF-8639 Module specification

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 14

Enlargement of left-hand side of graphic

Enlargement of right-hand side of graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 15

HPE GreenLake for Block Storage compute IO module details

Frontend connectivity Frontend connectivity

First FC adapter First 10/25 GbE* or second FC adapter
OCP slot 3 OCP slot 4
1 GBASE-T 1 GBASE-T Serial USB
Admin port ILO Service port ILO

Status LEDs 3-digit LED display

OCP slot 1 UID, Health, Alert UID, Health, Alert OCP slot 2
100 GbE RoCEv2 for backend connectivity 100 GbE RoCEv2 for backend connectivity

HPE GreenLake for Block Storage compute IO module port locations

Note
*Ports 1 and 2 are currently reserved for iSCSI. Ports 3 and 4 are reserved for RCIP.

This module provides connectivity and processing power for the array. Think of it as a storage array node/controller
you may know from other arrays. HPE GreenLake for Block Storage Controller Nodes is available in three variants:
• HPE Alletra Storage MP 256 GB 8-core Block Controller Node
• HPE Alletra Storage MP 256 GB 16-core Block Controller Node
• HPE Alletra Storage MP 512 GB 32-core Block Controller Node
Two controller node SKUs must be ordered per chassis. Both controller nodes need to be of the same type. Different
controller node types cannot be mixed in the same chassis. Each controller node contains two OCP slots for HBAs
(slot 3 and slot 4), and slots 1 and 2 are used for expansion. Each 32-core block controller node requires 100 GbE 2-
port OCP HBA in slot 1.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 16

Extra information on built-in ports:

The admin port is connected to the customer LAN and is configured with an IPv4 or IPv6 address and used for call
home connectivity when Domain Name System (DNS) and proxy are configured for access to the HPE support
infrastructure. The Integrated Lights-Out (iLO) port provides access to the IOM serial console after authentication. A
DHCP server is required to connect and can be connected to the customer network. Access to the iLO port is
managed by the customer; therefore, it may or may not be available for support engineer use. HPE does not support
the connection for the iLO Ethernet to the customer network.
Serial port provides direct access to the IOM/node’s serial console. The pinout is backward-compatible with the HPE
3PAR family serial cable kits. The pre-requisites are:
• The two cables and adapter provided in the serial cable kit are shipped with the system
• A service computer with a USB port is needed to access the console
• A terminal emulator program (such as PuTTY) is required to access the CLI
• A USB iLO port is an alternative access for support engineers when the customer declines access to the standard
iLO port. A USB to Ethernet adapter should be ordered to use this connection

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 17

HPE GreenLake for Block Storage OCP adapters

64 GB FC HBA 10/25 GbE HBA HPE 100 GbE NVMe -oF HBA
Host (SCSI and NVMe-oF), RCFC,
Supported for Host (iSCSI), RCIP JBOF, 32-core node interconnect
Peer Motion, Online Import
Number of ports 4 4 2

SFP+ included No No No
• 10 Gb SFP+
Available SFP 32 Gb SFP28 Yes
• 25 Gb SFP28
DAC cable or AOC supported* No Yes* QSFP28 DAC or AOC

HPE GreenLake for Block Storage OCP adapter options

HPE 100 GbE NVMe-oF

64 GB FC HBA 10/25 GbE HBA
HBA
Host (SCSI and NVMe-oF),
JBOF, 32-core node
Supported for RCFC, Peer Motion, Online Host (iSCSI), RCIP
interconnect
Import

Number of ports 4 4 2

SFP+ included No No No

• 10 Gb SFP+
Available SFP 32 Gb SFP28 Yes
• 25 Gb SFP28

DAC cable or AOC

No Yes* QSFP28 DAC or AOC
supported*

Enlargement of previous table

Note
*See SPOCK for more specific information about cables.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 18

Note
AOC = Active Optical Cable
DAC = Direct Attach Copper
NVMe-oF = NVMe-over-Fabric
RCFC = Remote Copy over FC
RCIP = Remote Copy over IP
SFP = SFF Pluggable

The considerations for Block Storage OCP adapters are:

• FC host connectivity:
̶ Slots 3 and 4 (always start with slot 3)
̶ Ports are numbered from left (port 1) to right (port 4), but there is no indication on the hardware
̶ FC adapters are capable of 64 GB
̶ Currently only offers 32 GB SFPs
̶ Supports switched connectivity to host (direct connection not supported)
̶ Requires NPIV enabled switches
̶ Enables SCSI/FC and NVMe-oF/FC environments
• iSCSI host connectivity:
̶ Requires an Ethernet HBA in slot 4
̶ Can only use ports 1 and 2
̶ Maximum of two iSCSI ports per node
̶ Support for only one VLAN per port
̶ Single iSCSI portal per port
̶ Max 64 initiators per port
̶ Volume replication not supported when using iSCSI
̶ VMware ESXi™ and Linux® environments only

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 19

• RCFC:
̶ RCFC connectivity is supported with FC adapters loaded in OCP slot 3 and/or 4
̶ When RCFC is required, there must be at least one port per node
̶ A port must be dedicated to RCFC (it cannot share a physical port with the target identity)
̶ RCFC requires FC switches; direct connect is not supported
̶ Replication is only available from/to HPE GreenLake for Block Storage, HPE Alletra 9000, and HPE Primera with
1 to 1 topology
• RCIP:
̶ There are no onboard ports for RCIP
̶ This feature requires an Ethernet HBA in OCP slot 4
̶ Only port 3 and 4 of this adapter can be used
̶ A maximum of two RCIP ports can be configured per node
̶ The two RCIP ports must be configured on different subnets. These subnets must differ from the management
subnet
̶ The RCIP connections must be the same on all IOMs

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 20

Slot population rules for different platform models

Node to
Node-to-node
Platform (external) JBOFs Slot 1 Slot 2 Slot 3 Slot 4 Comment on slots 3 and 4
connection
connection
N/A
8-core Mezz card Node internal Empty
• Frontend: Minimum 1 frontend
enclosure only
Empty HBA per node:
FC HBA or
0 to 2 external • FC in slot 3/4 or
• CX-6 Dx: if 1-2 FC 10/25 Gb Ethernet:
16-core Mezz card JBOFs, direct • iSCSI in slot 4
external JBOFs are in HBA • iSCSI: Ports 1 to 2
attached • Optional 10/25 Gb Ethernet in
use • Remote Copy: Ports 3 to 4
slot 4
0 to 2 external CX-6 Dx • Otherwise: Empty
32-core CX-6 Dx JBOFs, direct (node to
attached node)

OCP adapter setup for different platforms

For different platform models, the setup of OCP adapters follows specific rules. Slots 1 and 2 are set aside for back
end connections. Ethernet HBAs can only be used in OCP slot 4, the ports are as follows:
• Port 1 and 2—Dedicated for iSCSI connectivity
• Port 3 and 4—Dedicated for RCIP connectivity

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 21

Platform Node-to-node connection Node to (external) JBOFs connection Slot 1

• N/A
8-core Mezz card
• Node internal enclosure only
Empty
• 0 to 2 external JBOFs
16-core Mezz card
• Direct attached

• 0 to 2 external JBOFs CX-6 Dx (node to

32-core CX-6 Dx
• Direct attached node)

Enlargement of left-hand side of table in the previous graphic

Slot 2 Slot 3 Slot 4 Comment on slots 3 and 4

Empty • Frontend: Minimum 1
FC HBA or 10/25 Gb
Ethernet: frontend HBA per node:
• FC in slot 3/4 or
• CX-6 Dx: if 1 to 2 external FB HBA • iSCSI: Ports 1 to 2
JBOFs are in use • iSCSI in slot 4
• Remote Copy: Ports 3
• Otherwise: Empty to 4 • Optional 10/25 Gb
Ethernet in slot 4

Enlargement of right-hand side of table in the previous graphic

Note
CX-6 Dx: NVIDIA® Mellanox® ConnectX-6 Dx is a highly secure and advanced smart NIC.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 22

Port number: N:S:P

array1 cli% showport

N:S:P Mode State--Node_WWN/IP--- -Port_WWN/HW_Addr- Type Protocol Label
0:3:1 target ready 2FF70002AC07EF88 20310002AC07EF88 host FC -
0:3:2 target loss_sync 2FF70002AC07EF88 20320002AC07EF88 free FC -
0:3:3 target loss_sync 2FF70002AC07EF88 20330002AC07EF88 free FC -
0:3:4 target loss_sync 2FF70002AC07EF88 20340002AC07EF88 free FC -
0:4:3 peer offline - 40A6B76A6CFA free IP -
0:4:4 peer offline - 40A6B76A6CFB free IP -

N:S:P port number locations

Node:Slot:Port identifiers help correlate between user interfaces and physical port locations.
The showport output may refer to slot 5 with ports. This is the internal 25 GbE mezzanine card found in the
8- and 16-core IOMs.

Enlargement of bottom portion of graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 23

HPE Alletra Storage MP chassis—Disks side

Drive Status LEDs
Information pull tab

The components of a key chassis

The disks side of the HPE Alletra Storage MP chassis can be populated with up to 24 drives, arranged from left to
right, and requires a minimum of eight drives. Importantly, all the drives supported in this setup are self-encrypting
SSD drives, ensuring a high level of data security and performance. Available capacities are all Federal Information
Processing Standard [FIPS] or FIPS and Trade Agreements Act [TAA] compliant. Drive capacity includes:
• 1.92 TB
• 3.84 TB
• 7.68 TB
• 15.36 TB

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 24

HPE GreenLake for Block Storage JBOF IO module details

OCP slot 1 OCP slot 2

100 GbE RoCEv2 connectivity 100 GbE RoCEv2 connectivity

OCP slot’s location

The HPE Alletra Storage MP NVMe all-flash expansion shelf (JBOF) is a 2U chassis to add additional system capacity.
The expansion shelf supports eight to 24 NVMe SSDs, in two-drive increments with up to two expansion shelves per
system. The 8-core systems do not support expansion shelves. The connectivity requirements include:
• Inter-IOM connection:
̶ Slot 1: port 1 and port 2 (2x 100 GbE)
̶ Applicable to 32-core IOM variant only as 8-core and 16-core IOM use an internal mezzanine card (two x 25
GbE)
• Drive expansion (JBOF):
̶ Applicable to 16- and 32-core IOM variants only
̶ Direct connect to up to two JBOFs:
▫ Slot 2: port 1 ≥ first JBOF
▫ Slot 2: port 2 ≥ second JBOF
Back end connectivity is based on 100 Gb Ethernet (CX-6 Dx) and is used to connect JBOF IO modules to compute IO
modules, the concept is similar to disk enclosure IO modules that connect to controller nodes on many storage array
models.

Note
This is applicable to 16- and 32-core IOM variants only. The 8-core variant does not support JBOF-based expansions.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 25

Enclosure number—Disk drive number identifier

DSCC UI showing disk drive number identifiers

Identifiers help your customer correlate between user interfaces and physical disk location.
Each disk is identified two separate ways:
• By the disk number—Uses DiskX notation. X is numbers assigned sequentially as the disk area is added
• By the position (CagePos parameter)—Uses C:D notation where C is the drive enclosure number and D is the
disk-slot number within the enclosure

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 26

Enlargement of left-hand side of graphic

Enlargement of right-hand side of graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 27

HPE Alletra MP drive enclosure/cage JBOF connectivity

JBOF #2

JBOF #1

Nodes

Cable connectivity for drive enclosures

The HPE Alletra MP drive enclosure/cage for JBOF connectivity has a few simple rules for its cable connections. It
only supports systems that have either 16-core or 32-core setups. Your customer can connect up to two JBOFs. The
cabling must be done through direct connections. Any errors with the cabling are reported through the alerts or by
using the checkhealth command.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 28

HPE GreenLake for Block Storage releases and features

HPE Alletra release dates

The HPE Storage portfolio has expanded with the introduction of new offering (HPE GreenLake for Block Storage),
which has been released in two phases. The initial release in May 2023 brought the entry-base models, while the
second release in October 2023 introduced scale-up entry and mid-range models. Each phase offers distinct
capabilities and enhancements:

• Entry-base models:
̶ 2-node base enclosure (8- or 16-core)
̶ 1.92, 3.84, 7.68, 15.36 TB NVMe Self-Encrypting Drive SSD
̶ Up to 24 NVMe SSD / 368 TB (raw)
̶ FC host support (SCSI and NVMe-oF)
̶ FC and IP 1:1 replication
̶ Active Peer Persistence
̶ Peer Motion from HPE 3PAR
• Scale-up entry and midrange models:
̶ 2-node base enclosure (8-, 16-, or 32-core)
̶ Up to 2 directly connected add-on JBOFs
̶ Up to 72 NVMe SSD / 1080 TB (raw)
̶ SCSI host support added
̶ Peer Motion from HPE Primera/HPE Alletra 9K added
̶ Remote Copy with HPE Primera/HPE Alletra 9K added
̶ VMware vSphere® Virtual Volumes™ (vVols) support and more planned in the next release

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 29

Software and technologies

This subsection covers the licensing structure of features as well as software and support subscriptions.

HPE GreenLake for Block Storage OS software details

Software components included with HPE GreenLake for Block Storage OS

HPE cloud management is enabled by the subscription-based DSCC. HPE GreenLake for Block Storage ships with all-
inclusive system software licenses as part of the solution. The only exception, due to legal obligations in certain
geographies, is data encryption, which can be ordered together with self-encrypting FIPS drives at a price of
approximately $1.

Note
LDAP = Lightweight Directory Access Protocol
QoS = Quality of Service

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 30

Software and support subscription

HPE Ninja Online showing the configuration of a 16-core system with 184.3 TB raw capacity

The subscription includes:

• All-inclusive HPE GreenLake for Block Storage software features for the selected controller models (8- 16- and 32-
cores)
• Specified raw capacity in TBs grouped into tiers
• Selected term (one to seven years)
• Cloud-based management of the array from the HPE DSCC
• HPE GreenLake supports SaaS

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 31

Available tiers Raw capacity

Tier 1 0 to 16 TB

Tier 2 17 to 25 TB

Tier 3 26 to 50 TB

Tier 4 51 to 100 TB

Tier 5 101 to 200 TB

Tier 6 201 to 400 TB

Tier 7 401 to 800 TB

Tier 8 801+ TB

Raw capacity tiers

SKU Description Qty

HPE GreenLake for Block Storage OS 16-core Tier-5
S1L15A 184
3-Year Software and Support SaaS

SaaS SKU for a 16-core system with 184.3 TB raw capacity and 3-year subscription

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 32

Learning check
1. What is the maximum number of expansion enclosures currently supported by HPE GreenLake for Block Storage?
A. 1
B. 2
C. 3
D. 4
E. 6

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 33

Storage management and monitoring

This section covers management and monitoring options for HPE GreenLake for Block Storage.

HPE GreenLake for Block Storage manageability

• HPE Data Services Cloud Console (within HPE GreenLake cloud platform)
GUI • HPE on-array management UI
• HPE OneView integration

• CLI management client for RHEL and Windows server

• SSH from any supported operating system
• Windows PowerShell toolkit
CLI / API
• DSCC PowerShell toolkit
• WEB API (RESTful)
• SMI-S (SNIA)

• HPE InfoSight
Analytics
• HPE Data Services Cloud Console

• VMware® • Kubernetes • Real-time

Management • Microsoft • Mesosphere performance
monitoring with
integrations • OpenStack • Rancher
Grafana
• Docker • And many more

HPE GreenLake for Block Storage manageability and integration options

The HPE GreenLake for Block Storage operating system helps simplify, automate, and expedite storage management
by handling provisioning, tiering, and change management autonomically and intelligently at a subsystem level with
minimal-to-no administrator intervention.
The system’s user interfaces have been developed to offer autonomic administration, which means that the interfaces
allow an administrator to create and manage physical and logical resources without requiring any overt action.
Provisioning does not require any preplanning, yet the system constructs volumes intelligently based on available
resources, unlike manual provisioning approaches that require planning and the manual addition of capacity to
intermediary pools.
The main interface for array initialization and direct GUI interaction is the DSCC. AI-driven monitoring and predictive
analytics are provided by HPE InfoSight. In addition, we are providing application programming interfaces (APIs),
plug-ins, and more to integrate into management and automation solutions.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 34

HPE Data Services Cloud Console

This subsection covers HPE DSCC, which is part of the HPE GreenLake Edge-to-Cloud Platform.

Data Services Cloud Console overview

Backup and Recovery Block Storage CloudPhysics Data Ops Manager

Automated protection with rapid Simple and easy to provision, protect, Quickly analyzes data center Simplify storage management at scale
restore and cost-efficient, long -term and manage block storage that meets infrastructure to reveal cost and with fleet-wide management and
retention application needs performance optimizations monitoring

J _s l af � � J _s l af � � J _s l af � � J _s l af � �

Storage Fabric Management File Storage Private Cloud Business Disaster Recovery
Accelerates configuring, monitoring, File-based storage accessed through
Edition Continuous data protection for disaster
and managing storage fabrics standard file serving protocols Global lifecycle management of recovery and data mobility
infrastructure and virtualization
resources

J _s l af � � J _s l af � � J _s l af � � J _s l af � �

Data Services Cloud Console dashboard

Increasingly, customers selecting storage products are looking for a platform to adopt, rather than making product-
specific decisions. HPE GreenLake Edge-to-Cloud Platform delivers the largest array of storage services in a single
platform. A comprehensive suite of on-demand hybrid cloud data services spanning block, file, disaggregated
hyperconverged infrastructure (dHCI), storage fabric management, backup and recovery, and disaster recovery—all
delivered with a unified cloud operating experience—enables organizations to modernize data management when
and where they need it, all in just a few clicks.
For example, with HPE GreenLake for Backup and Recovery, your customer can experience effortless cloud backup
from a single pane of glass while meeting every service-level agreement (SLA) across clouds through instant restores,
rapid recovery on-premises, and cost-effective, long-term retention in the cloud. HPE GreenLake for Backup and
Recovery is a SaaS backup service designed to protect on-premises VMware® virtual machines (VMs), and cloud
native workloads like Amazon EBS volumes and EC2 instances across hybrid cloud.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 35

To summarize, DSCC brings the cloud experience to on-premises HPE Storage:

• Common look and feel as other HPE cloud-enabled apps
• Enables customers to manage HPE Storage from anywhere
• Easy access to all storage systems and data management apps
• Redundant, secure, and scalable storage management software
• Enables subscription-based infrastructure
• No storage management software to install or upgrade

HPE GreenLake Edge-to-Cloud Platform services

HPE GreenLake Edge-to-Cloud Platform structure

The HPE GreenLake Edge-to-Cloud Platform is a cloud services platform that is focused on data. At the base is a set
of common services such as login and unified support. Your customer needs to be able to transport it by networking,
compute data streams in real time and protecting their data with storage and backup.
Workloads and applications must be orchestrated across their lifecycle, whether they are bare metal virtual machines
or containerized. In addition, global machine learning and analytics must be applied across your customer’s edge-to-
cloud world to get value from their data and drive outcomes for their business.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 36

The platform does not stop there. To fully deliver end-to-end security and corporate compliance, HPE GreenLake has
advisory services to help your customer tackle complex problems. HPE Managed Services operate everything that is
required in delivering an SLA. HPE Financial Services make it easy to recycle aging equipment and pay for the next
step in your customer’s digital transformation, and all of this is delivered with and through HPE and our partners.

HPE GreenLake Edge-to-Cloud Platform—Console

HPE GreenLake Edge-to-Cloud Platform console dashboard

After logging in, the HPE GreenLake console dashboard displays. This is your customer’s starting point. Here, they
can access GreenLake services and applications such as DSCC. It also assists with device onboarding that takes place
the first time a new HPE GreenLake for Block Storage solution deploys.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 37

HPE Data Services Cloud Console: A single destination for all new HPE data and infrastructure
services

Backup and Disaster Copy data Security and Data Cloud data
Dev and test Analytics
recovery recovery management compliance migration Services

Fleet Block File Object HCI Storage fabric Policy IaaS Cloud infrastructure
AI-driven manager services services services manager manager manager APIs services
global
intelligence

Cloud-native data
Mission critical General purpose Secondary Big Data/AI Edge Cloud
infrastructure

As a service
HPE GreenLake platform

HPE GreenLake platform for all new HPE data and infrastructure services

Recognizing customer needs for unified data operations, HPE has architected the answer with the HPE Intelligent
Data Platform (IDP). This platform connects applications to infrastructure, innovators to data, and automation to
policies in a seamless, unified cloud experience wherever data lives.
IDP is a cloud-native architecture that includes unified, cloud data services to manage the lifecycle of data across
edge-to-cloud, cloud infrastructure services to deliver cloud operations, cloud-native data infrastructure optimized
for every app and built for the cloud, and artificial intelligence (AI) for autonomous operations—all delivered as an as-
a-service experience.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 38

Data Ops Manager

DSCC Data Ops Manager dashboard

Data Ops Manager is an application within DSCC that enables your customer to manage their organization’s HPE
GreenLake for Block Storage and HPE Alletra cloud native data infrastructure from any location or device.
Your customer can use Data Ops Manager to:
• View system, array inventory and details. Your customer can filter, search, and sort systems, and view detailed
information about issues, capacity, performance, and hardware health. They can view and edit network settings
and system settings; collect support log bundles; and perform locate actions
• View host groups, provisioning, and details. They can also create and delete hosts and host groups; add hosts to
host groups; and view host details, initiator, system, and performance trends
• View volume listing, summaries, and access details
• Define replication partners for systems (your customer can apply protection for volume sets using the Block
Storage UI with HPE GreenLake’s DSCC)
• Service the system to replace hardware components
Data Ops Manager is where the cloud operator or infrastructure engineer manages the underlying infrastructure.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 39

Block storage

With HPE GreenLake for Block Storage service:

• Mission Critical
• Business Critical
• General Purpose

Else:
• HPE Alletra 9000
• HPE Alletra 6000
• HPE Alletra 5000
• HPE GreenLake for Block Storage MP
• HPE Primera
• HPE Nimble

• Windows File Server

• ….
• Microsoft SQL Server
• …
• VMware ESXi
• + Create (your own)

Depending on storage tier and capabilities:

• Daily snapshots | 30 days expiration
• 15-minute snapshot | 1 day expiration
• Sync Replication, zero data loss | Instant recovery
• Sync Replication, zero data loss
• No protection

Block storage volume creation wizard

Block storage is an application in the DSCC that focuses on storage lifecycle management for block provisioning,
native replication, and snapshot-based data protection. The service makes it easy to provision, protect, and manage
volumes and volume sets. To access the service, click the Block Storage tile in the DSCC.
Block storage separates out the application administration features from Data Ops Manager, leaving only the
infrastructure administration in Data Ops Manager.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 40

Your customer can use block storage to:

• Monitor volume and host information, volume issues, and capacity usage in different applications in an integrated
dashboard
• Create, update, or delete volumes
• Add volumes/remove volumes from volume sets
• Provision block storage by storage tier when creating volumes
• Apply predefined or newly created protection policies to volume sets
• Display volume capacity and performance (aggregated at the volume set level)
This is where the cloud consumer or storage administrator would manage and protect the day-to-day activities of
volumes.

Intent-based provisioning

1 Log on to DSCC.
Visibility

2 Start provisioning.
Powered
Prediction

by AI

3 Power the applications. Decision

Steps to achieve intent-based provisioning

With intent-based provisioning, an infrastructure administrator only needs to specify the workload type, capacity, and
host groups that need access to that workload. The intent-based provisioning service in the DSCC optimally
determines where data should be stored with real-time context to resource headroom and application-specific SLAs.
The combination of intent-based provisioning and role-based access control enables self-service data infrastructure
provisioning for application owners and developers—without the need for storage expertise. This delivers a true
cloud provisioning model and accelerates application development cycles—enabling the effortless application
provisioning experience demanded by today’s fast-paced DevOps processes.
Intent-based provisioning can be started either directly from the Intent-based Provisioning tile shortcut or the
Block Storage tile as a part of the provisioning wizard.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 41

HPE GreenLake for Block Storage UI

This subsection covers the HPE GreenLake for Block Storage UI, which is a part of the array operating system.

HPE GreenLake for Block Storage local UI: Features and functions

HPE GreenLake for Block Storage UI dashboard example

The on-node management UI runs as a service in the HPE Alletra OS and provides intuitive insights related to
capacity, performance, and resource allocation for a single array.
To access HPE GreenLake for Block Storage local UI, your customer can browse to the storage system at https://<IP
address or FQDN>.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 42

Compared to the DSCC management apps, this UI provides a limited set of features, including:
• Dashboard screen: Shows system status, performance, and capacity overview
• Storage screen:
̶ Allows your customer to view, create, and edit application sets, volumes, hosts, host sets, replication, and
perform Peer Motion operations
̶ Allows your customer to monitor and configure HPE Remote Copy and HPE Peer Persistence between two
systems in air-gapped sites without connectivity to the internet and HPE DSCC (requires approval from HPE
product management)
• System screen:
̶ Shows system details, hardware schematic, and software versions
̶ Allows servicing of the system including software updates, hardware repairs, and hardware upgrades
• Settings screen: Enables your customer to view and manage key system settings like telemetry, users, LDAP
integration, VMware vCenter® settings, network settings, and certificates

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 43

HPE GreenLake for Block Storage local UI: Storage screen—View and create host and
application sets

Storage screen quick actions users can take

The Storage screen allows your customer to view, create, and edit application sets, volumes, hosts, host sets, and
replication, as well as perform Peer Motion operations. Each tile leads to a given type of object list where actions can
be performed. Most of the tasks are wizard based. Note that some types of actions will only be applicable to certain
types of objects.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 44

HPE GreenLake for Block Storage CLI

This subsection covers HPE GreenLake for Block Storage CLI that is a part of the array operating system accessible
via either a dedicated CLI application or via any Secure Shell (SSH) client.

HPE GreenLake for Block Storage CLI: Overview

HPE GreenLake for Block Storage CLI example

Your customer can use the CLI to monitor, manage, and configure HPE Storage systems. They can access the CLI
from supported host systems using:
• SSH client application:
̶ Most supported host operating systems include an SSH client application that your customer can use to access
the CLI. The exception is Windows, which requires customers to install an SSH client. Using SSH requires no
additional installation or setup of the CLI
̶ SSH eliminates compatibility issues between the client and server because your customer has no CLI client
installed
• HPE GreenLake for Block Storage CLI remote client: HPE provides a CLI remote client that requires installation
on each supported remote host. The CLI is supported on Linux 64-bit and Windows 64-bit operating systems

Note
For more information and a list of supported versions of each operating system, visit the SPOCK website.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 45

The CLI remote client includes commands that, in general, can be divided into four groups:
• create—Command to create object details
• remove—Command to remove/delete objects
• show—Command to display object details
• set—Command to edit objects

HPE GreenLake for Block Storage CLI: Example—Show system information

HPE GreenLake for Block Storage CLI example to show system information

Most CLI commands use the following syntax: cmd [subcmd] [option [arg]]…
[<spec>...][<pattern>...]
The information below provides a description of what each command is used for:
• cmd—Specifies the operation to be executed, such as createcpg or showcage
• subcmd—Specifies an action that a command will execute, such as controlencryption enable (where
enable is a subcommand of the controlencryption command)
• option—Indicates an optional command line element, such as –rw , -f , or -waittask . Options begin with
the dash ( - ) symbol
• arg—Indicates a specific variable (argument) for an option. Often used to identify a specific node, volume, or disk
• spec—Indicates a specifier used with a required command line element such as a command or option

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 46

• | (vertical line)—Requires your customer to choose one option or spec when separated by this character, such as
-y|-n or <VV_name|pattern|VV_Set>
• { } (curly brackets)—Indicates grouped elements. Do not enter the braces; enter only the information inside the
braces
• [ ] (brackets)—Indicates optional elements. Do not enter the brackets; enter only the information inside the
brackets
• < > (angle brackets)—Indicates user-supplied input. See spec, above
• ... (ellipsis)—Indicates that a specifier or an option can be used more than once in a command

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 47

Learning check
2. Which graphical interface should you recommend to your customer if they want to access the widest range of
functionalities?
A. HPE Storage Management Utility
B. HPE GreenLake for Block Storage UI
C. HPE Service Processor Service Console
D. HPE StoreServ Management Console
E. HPE Data Services Cloud Console

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 48

Monitoring
This subsection covers monitoring options provided by user interfaces discussed in the previous section.

Data Ops Manager: Headroom Utilization trend analysis

DSCC configurable settings for enhanced reporting and extended reporting

Enhanced reports
The Performance tab provides enhanced reporting capabilities for HPE GreenLake for Block Storage, HPE Alletra
9000, and HPE Primera systems. Two new charts have been added to show the trend in IO Size and Queue Length.
The following three predefined templates have been added:
• Top volume paths by read latency
• Top volume sets by read IOPS
• Top host ports by write throughput

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 49

Your customer can create customized views by resource type for Volumes, Volume Sets, Host Ports, and Hosts with
“Top 6” or “Select 10” with sorting by a metric.

In addition, your customer can filter the views by duration, by reads or writes, and by object name. They can view
charts in a new browser window, in full screen mode, and download them as CSV or PDF files.

Note
Available resource types and metrics will vary, based on the storage system model, and additional reporting features
will be extended over time.

Extended reporting
Extended reporting provides capabilities that allow your customer to analyze and troubleshoot performance issues
for an HPE GreenLake for Block Storage system and its volumes. The new system-level graphs are:
• Headroom Utilization with analysis breakdown by Top 6 Volume contribution
• Top Volume Hotspots by Latency provides anomaly detection to identify outliers
The volume-level insights are:
• Latency variation insights, mapped as annotations to the Latency chart
• Workload drift insights, mapped as annotations to the Throughput chart
RC Links and Pools reports
Using the custom reports option, your customer can now view and compare performance metrics across Radio
Control (RC) links for GreenLake for Block Storage, HPE Alletra 9000, and HPE Primera 600.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 50

Enhanced reporting: Available options

Enhanced reporting examples for Data Ops Manager and Block Storage

Enhanced reporting options are available in both Data Ops Manager as well as Block Storage sections and include:
• Three standard templates:
̶ Top volume paths by read latency
̶ Top volume sets by read IOPS
̶ Top host ports by write throughput
• Custom views with:
̶ Top 6 with Sort by Metric
̶ Select 10 objects

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 51

• Resources and metrics:

̶ Volumes, Volume Sets, Host Ports, Hosts, Physical Drives (IOPS, Throughput, Latency, IO Size, Queue Length)
̶ RC links (Round Trip Time, Throughput, Transmitted Data)
• New default trend charts:
̶ IO Size
̶ Queue Length

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 52

Extended reporting: Next generation root cause analysis for system and volume-level
performance

Data Ops Manager UI showing root cause analysis for system and volume-level performance

New features in Data Ops Manager and block storage that consists of:
• Headroom Utilization trend analysis
• Top volume hotspots by latency
• Latency insights
• Workload drift insights

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 53

Overview of headroom
A key performance indicator (KPI) measures the current load of a system and its preferred throughput—the load at
which a system can deliver optimal response times, unique to each system, and based on multiple factors.
This concept was first introduced to DSCC with Intent-based Provisioning. It simulates the workload by using
application-aware workload patterns to derive estimated headroom availability after placement. Then it recommends
the best fit for workload placement after calculating the currently available headroom across all systems and takes
into consideration:
• System configuration—Model, # of Nodes, # of Drives, Drive Type, Port Configuration
• Workload configuration—Request Size, % of Read vs. Write Throughput
• Volume configuration—Dedup, Compression, Thin

Data Ops Manager: Headroom Utilization trend analysis

Example of Data Ops Manager Headroom Utilization trend analysis

The Headroom Utilization graph shows how close the actual throughput is to the preferred or optimal throughput
over time. It charts the top five volume contributors and the remaining contributors during the same period.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 54

The preferred throughput is the load at which the system can deliver reasonable response times. The preferred
throughput is unique to each system and is based on multiple factors, such as:
• Type of workload that is running on the storage system (request size and percentage of read/write throughput)
• Volume type (compression, dedupe, and thin)
• Storage system configuration (model, number of nodes, drive type and count, port type, and so on)
• The utilization levels are characterized as:
̶ Low—The system can handle significant workload increases without impacting performance
̶ Medium—The system can accommodate moderate workload increases without impacting performance
̶ High—The system is currently operating at a level that is close to or reaching its limit. The system may
experience performance degradation if the workload increases
̶ Overutilized—The system is operating at or beyond its limit. Adding any further workload may lead to
performance degradation
The Headroom Utilization graph provides information about the:
• Utilization levels of the storage system during the selected time interval
• Periodic recurrence in the change of headroom utilization levels
• Insights into the top volumes that are contributing to headroom utilization
Your customer can use this graph to answer questions, such as:
• Is the workload on my system as planned?
• Which volumes are causing the system to be overutilized?
• Can I add a workload without seeing performance degradations?
• Can I add any new workload to my system, or is it already overutilized?

Note
Intent Based Provisioning uses headroom utilization when making workload placement recommendations.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 55

Headroom Utilization

Trend of Headroom
Utilization with regions
of overutilization and
periodicity

Identifies the largest

contributors to system
utilization

Examples of Headroom Utilization UI in Data Ops Manager

High headroom utilization over a long period indicates that a system cannot support additional workloads while
maintaining optimal response times for the current workloads.
Think of headroom utilization like the RPM gauge in a car. Your customer can cross into the overutilized region for
short time periods, but it is not recommended to stay beyond the overutilized line for a long duration. When a storage
system remains at a high level of headroom utilization for a long time, it might be an indication that there is no room
for additional workloads.
If the headroom utilization is unreasonably high, the storage system can still deliver more IOPS at this level; however,
latency (or response time) increases.

Note
RPM = Revolutions per minute (tracks how fast the engine is running)

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 56

Data Ops Manager: Top Volume Hotspots

Example of Top Volume Hotspots UI in Data Ops Manager

The graph shows the volumes that are under more stress when compared to other volumes in the storage system. It
uses latency to identify anomalous volumes in terms of latency and ranks them, based on the number of times they
are identified as an anomaly in a time period.
Hotspots indicate bottlenecks in a storage system. For example, hotspots can identify the objects or entities
responsible for saturating specific resources. Hotspots shown as a heat map provide a visual representation of
resource utilization that quickly and efficiently identifies performance bottlenecks.
Your customer can specify whether to analyze read/write latencies and for different time periods (for example, eight
hours or one day). They can correlate the information with other performance metrics for easy troubleshooting.
Your customer can use this graph to address questions like:
• Which volumes are seeing unusually high service times when other volumes are not?
• Is there an issue with the system?
The message, no hotspots detected, indicates that there are currently no volumes under stress. When issues are
reported, your customer can use the hotspots graph to determine if there is a problem with just a few volumes or if
there is a wider problem across the system.
The “Volume Hotspots” report shows up to five volume hotspots. Hotspots are volume outliers consistently reporting
higher latency compared to other volumes. These hotspots are ranked by how frequently they are reported as
outliers during the selected time period.
Hotspots might contribute to bottlenecks. If no volume hotspots are reported during high overall system latency, no
outliers are reported.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 57

Block storage: Latency Insights

Example of Latency Insights that displays after clicking on an insight overlay

By default, the Latency graph reports average IO latency on the device for five-minute intervals. When there are high
variations among IO latencies for the time period, read and write insights appear as overlays on the system-level
latency trends charts for volumes. Your customer can view further details about the variance by clicking the insight
area on the chart. Measurements that are higher or lower than most of the data are known as outliers. Averaging can
mask the impact of outliers on overall system performance.
Suppose that most latency measurements are low, but a few outliers are high. The average value might not reflect the
impact of the outliers and can lead to a false understanding of performance. Your customer might assume that the
system is performing well, based on the average value, while in reality, some requests are experiencing high latencies.
The average latency might not indicate a problem, but the outliers can explain why a user application is experiencing
high latencies.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 58

Averaging can lead to the loss of outliers, as outliers can have a large impact on the overall average, but may not be
represented in the average value. Averaging the latency measurements discards information about the distribution of
the data and may not provide an accurate representation of the overall performance of the storage system. It is
important to use other methods to analyze the performance data and averaging, such as using statistical methods to
analyze the distribution of the data and identify outliers.
For instance, calculating percentiles like 95th or 99th can be used to identify the maximum threshold of latency and
can be used as an indicator of an issue.
The graph serves two key purposes. Firstly, it helps in analyzing whether the latency experienced in an application is
related to storage issues, despite the overall averages appearing acceptable; and secondly, it aids in proactively
scouting for potential problems.
It is important to have a clear understanding of the performance of the storage system, including both the average
value as well as the distribution of the data. That way, your customer can identify and address any issues that might
be impacting the performance of the system, including outliers in latency measurements.
Latency Insights presents latency statistics, including read and write metrics. It overlays additional insights on to
existing trends when there are significant variations in IO latencies within the sampling period. These insights help
identify instances where individual IOs might have higher latencies that are not reflected in the average, potentially
explaining the higher latencies observed in user applications.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 59

Block storage: Workload Drift Insights

Example of Workload Drift Insights in Data Ops Manager

Workload drift is a volume-level insight that indicates whether a workload is running as expected or if there is a
deviation. The drifts are annotated on the throughput trend chart for the volume so your customer can correlate the
insights with the available trend data. Workload drift insights are available on both read and write workloads.
The analysis runs for the selected time duration with five-minute granularity. The time ranges can be predefined or
custom and between eight hours and up to two days. Workload characteristics are a key factor in selecting a storage
array that is well-suited for a particular workload to achieve predictable performance. However, workload
characteristics can change over time so the selected storage array may be unable to handle the data access patterns
and performance requirements of the workload efficiently.
Therefore, it is important to monitor and detect when the predictability of the storage performance is no longer being
met. Your customer can use monitoring and analytics tools to track and measure the performance of the storage
system, including the IO patterns, data access patterns, and performance metrics. Latency is subjective because it
depends on the requirements of the application and the expectations of users. Nevertheless, detecting a change in
the IO pattern is a useful measurement. The insights enable your customer to make informed decisions that prevent
performance issues from affecting business applications.

An example of the summary displayed after selecting a Workload Drift Insight

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 60

Your customer can use the Workload Drift insights graph to answer questions, such as:
• Is the workload running as expected?
• Can a reported issue be attributed to a deviation of the workload for its expected pattern?
A change in the IO pattern can indicate an unusual workload or an issue with the workload that might result in
performance issues in the system and on future workloads. When your customer detects a deviation from the
expected IO patterns:
• They should investigate the cause of the deviation, and then take appropriate action if the change is also
correlated with high latencies in applications using the storage
• Consider applying workload threshold limits, scaling resources, or replacing the storage array with a different one
that better suits the current workload characteristics
When a change in the IO pattern is detected for the duration of the Workload Insight chart, throughput insights are
overlaid on the existing trends. A change in the IO pattern can indicate an unusual workload or an issue with the
workload that might result in performance issues on the system as well as other workloads on the system in the
future, workload drifts are annotated on top of the throughput trend chart for the volume.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 61

HPE GreenLake for Block Storage on-array management UI: Overview, alerts, and tasks

! 1

HPE GreenLake Management UI overview window with tasks and alerts icons in the top right-hand corner

The on-node management GUI runs as a service in the HPE GreenLake for Block Storage OS and provides intuitive
insights related to capacity, performance, and resource allocation for a single array. By default, it displays overall
health of the system, its external monitoring systems connectivity status, as well as brief capacity and performance
metrics.

Example of a task running in the HPE GreenLake for Block Storage OS

Example of an alert in the HPE GreenLake for Block Storage OS

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 62

HPE GreenLake for Block Storage on-array management UI: Performance

Example of the performance UI for HPE GreenLake for Block Storage

The Performance dashboard provides a summary of the storage system IO performance with connected hosts. Your
customer can easily check the performance behavior of their arrays, and find the host consuming the most IOs.
The performance dashboard screen displays:
• IOPS properties:
̶ The number of IOPS for exported volumes during the last 24 hours
̶ The number of read/write IOPS for exported volumes during the last 24 hours
• Latency properties:
̶ The storage system average aggregate IO latency in a one-hour period. The average is taken for the last 24
hours for all volumes exported to all hosts
̶ The storage system read/write IO latency in a one-hour period. The average is taken for the last 24 hours for all
volumes exported to all hosts

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 63

• Bandwidth properties:
̶ Bandwidth—The storage system average aggregate IO bandwidth during the last 24 hours for all volumes
exported to all hosts
̶ Bandwidth (graph)—The storage system read and write bandwidth during the last 24 hours for all volumes
exported to all hosts
• Top hosts properties IOPS (block chart)—The hosts with the most IO operations with the storage system during
the last 24 hours. Each block in the chart shows the host name and number of IOPS. Up to 10 hosts can be shown

HPE GreenLake for Block Storage on-array management UI: Capacity

Example of the capacity UI for HPE GreenLake for Block Storage

Your customer can easily check their arrays’ current capacity consumption, savings, trends, and more by using the
capacity dashboard screen, this displays:
• Overprovisioning properties:
̶ Overprovisioned (ratio)—The overprovisioning ratio for all volumes on the storage system
̶ Exported capacity—The total capacity of volumes that is exported to hosts
̶ Usable capacity—The total capacity that can be stored on the storage system, after RAID and system overhead.
This does not include savings from data reduction (deduplication and compression) and thin provisioning
̶ Used capacity—The total capacity that has been consumed

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 64

• Savings properties:
̶ Total savings (ratio)—The compaction ratio for all volumes on the storage system. The ratio indicates the
virtual size of all volumes compared to the space used on the physical drives. For example, a ratio of 2.5:1
indicates that the virtual size is 2.5 times the amount of space used on the physical drives
̶ Provisioned capacity—The capacity of the storage system that appears to be available to hosts
̶ Host written capacity—The amount of space on the storage system that would have been used for host writes
if data reduction (deduplication and compression) had not been applied
̶ Used capacity—The total capacity of the storage system that has been consumed (the remainder is free). This
does not include savings for data reduction or thin provisioning (data reduction includes deduplication and
compression)
̶ Thin savings—The amount of capacity on the storage system that has been saved due to thin provisioning of
volumes
̶ Data reduction savings—The amount of capacity on the storage system that has been saved due to data
reduction of volumes (data reduction includes deduplication and compression)
• Free space and breakdown properties:
̶ Free space—The total amount of free capacity on the storage system
̶ Used capacity in volumes—The percentage of storage system capacity used in volumes. The percentage
includes base and clone volumes
̶ Used capacity in snapshots—The percentage of storage system capacity used in snapshot volumes
̶ Used capacity in system volumes—The percentage of storage system capacity used in system volumes. System
volumes are not accessible to users
̶ Free capacity—The percentage of storage system capacity that is free for creating volumes
̶ Failed capacity—The percentage of storage system capacity on the physical drives that is marked as being
failed
• Capacity trend properties:
̶ Provisioned capacity—The provisioned capacity trend during the last 12 months
̶ Used capacity—The used capacity trend during the last month

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 65

Performance metrics available using API and CLI

Service
Queue Average
Bandwidth time IOPS IO size
length busy
(latency)

Physical Cache
CPU Ports

Histograms
disks memory
Historical

Logical Virtual Remote

VLUNs
drives volumes Copy

Real-time

Reporting capabilities visualized for System Reporter

The on-node performance collector, called System Reporter, captures and reports the whole subsystem—from
physical disks at the back end, through the CPU and cache, into the logical disk and volume management layer, and
finally into the presentation, logical, and physical connectivity layer.
Some of these statistics can be collected in real time for immediate, detailed, and fine grain monitoring, analysis, and
diagnostics. Statistics are also collected on a regular basis and stored on node for recall through the tools.
Your customer can also set thresholds against performance metrics that can trigger alerts, which are useful to
highlight and act upon potential performance issues or bottlenecks before they become critical.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 66

CLI—Stat commands
The CLI stat commands can be used to look at real-time performance data at various levels. The stat output
loops continuously to show mean values over a sample period. There are various options available for the stat
commands, common options include:
• -iter iteration (number of loops)
• -d duration or time interval (default is two seconds)
• -ni non-idle objects will not display
To display histogram performance data, a series of hist commands is available as well:
Stat command Usage
statcache Cache performance

statcmp Cache management page stats

statcpu Cache CPU statistics

statpd Physical disks stats

statld Logical disks stats

statch Chunklet stats

statvv Virtual volume stats

statvlum VLUN stats

statport Port stats

HPE GreenLake for Block Storage CLI Stat commands

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 67

Using stat commands for performance

Front end

statcache, statcmp, statcpu

Exported volumes
statport -disk
Host ports statvlun, statvv

Controller node cache

Controller node CPU statport –host
Enclosure ports
statpd
Physical drives

Back end

Layers available for monitoring using HPE GreenLake for Block Storage CLI stat commands

The stat commands allow your customer to view performance counters at the layers involved in servicing IO:
• statvlun—Measures the roundtrip time of an IO as seen by the system. IOPS, bandwidth, and service times can
all be viewed
• statport -host—Displays port statistics for host-facing ports
• statcmp—Displays cache statistics. This command allows your customer to see whether IO flow is evenly
distributed across the nodes. The statcmp command can also be used to see the number of delayed
acknowledgements
• statcpu—Displays the node CPU usage
• statvv—Displays the internal system response to the IO. Comparing statvlun to statvv, your customer
can determine whether the array is having a performance issue
• statport -disk—Displays port statistics for disk-facing ports
• statpd—Displays the performance of the physical drives

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 68

CLI—Hist commands
By default, each command displays captured data from the last 24 hours, collected in five-minute increments. The
hist* CLI commands, in general, can be broken down into areas for investigation:
Hist
command Usage
histld Displays a histogram of service times for logical disks (LDs) in a timed loop

histpd Displays a histogram of service times for physical disks

histport Displays a histogram of service times for ports within the system

histvlun Displays virtual logical unit number (VLUN) service time histograms

HPE GreenLake for Block Storage CLI hist commands

CLI—Srstat commands
The CLI sr* commands draw from the capacity and performance metrics from the. srdata database volume. The sr*
commands display historical performance and capacity data, with an option to display metrics over a specific period
of time. By default, each command displays captured data from the last 24 hours, collected in five-minute increments.
The sr* CLI commands, in general, can be broken down into areas for investigation. The general areas are as
follows:
• Space—Displays data space consumed or free in different categories: srcpgspace, srpdspace,
srvvspace, srldspace, srsysspace
• History—Displays historical histograms for various performance collections: srhistld, srhistpd,
srhistport, srhistvlun
• Remote Copy—Displays performance statistics for Remote Copy over a measured period: srstatrcvv,
srstatrcopy
• Stat—Displays historical performance statistics over a measured period: srstatcmp, srstatcpu,
srstatld, srstatlink, srstatpd, srstatport, srstatqos, srstatvv,
srstatvlun

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 69

• stat commands:

̶ srstatcmp—Performance reports for cache memory

̶ srstatcpu—Performance reports for CPUs
̶ srstatld—Performance reports for logical disks
̶ srstatpd—Performance reports for physical disks
̶ srstatport—Performance reports for ports
̶ srstatqos—Performance reports for QoS rules
̶ srstatvv—Performance reports for VVs
̶ srstatvlun—Performance reports for VLUNs
̶ srstatlink—Performance reports for links
̶ srstatiscsi—Performance on iSCSI ports
̶ srstatiscsisession—Performance on iSCSI sessions
• Space commands:
̶ srcpgspace—Space reports for CPGs
̶ srpdspace—Space reports for physical disks
̶ srvvspace—Space reports for VVs
̶ srldspace—Space reports for logical disks
̶ srsysspace—Space reports on overall system capacity
• Histogram commands:
̶ srhistld—Histogram performance reports for logical disks
̶ srhistpd—Histogram performance reports for physical disks
̶ srhistport—Histogram performance reports for ports
̶ srhistvlun—Histogram performance reports for VLUN exports

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 70

Using srstat commands for performance

Front end

srstatcache, srstatcpu

Exported volumes
srstatport –port_type disk
Host ports srstatvlun, srstatvv

Controller node cache

Controller node CPU srstatport –port_type host
srstatiscsi,srstatiscsisession
Enclosure ports
srstatpd
Physical drives

Layers available for monitoring using HPE GreenLake for Block Storage CLI srstat commands

The srstat commands allow your customer to view historical performance counters at the layers involved in
servicing IO.

Performance alerts
Using the CLI, the administrator can create customizable alerts to identify performance issues that the array might be
experiencing. There are no predefined values or alerts within the tool, so your customer must identify their own
standards to which they will alert. Your customer should set values on observances that were made during the
monitoring of the array using performance reporting. Performance alerts highlights are:
• Performance alerts allow users to define criteria such that when a sample is taken and if all criteria are met, an alert
is invoked
• A maximum of 64 alerts can be defined
• Removal of alert criteria also automatically clears all corresponding alerts
• Alerts are configured and managed using the CLI
• If possible, enforce standard naming conventions for objects to facilitate the use of glob-style (wildcard) pattern
matching in alert criteria
• High-resolution sampling is the finest granularity that is possible (take the average rates into consideration vs.
transients when defining thresholds)

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 71

Performance alerts—Create
createsralertcrit
<type><options><condition>[,<condition>…]<comment><critname>
• Where <type> can be: port, vlun, ld, pd, cmp, cpu, link, qos, rcopy, rcvv,
vvspace, cpgspace, pdspace, sysspace, ldspace
• Where <options> can be:

̶ -daily/hourly/hires—Evaluated daily when a sample is taken

̶ -count <number>—The number of matching objects that must meet the criteria for the alert to be
generated
̶ -recur <recurrences>/<samples>—The alert will only be generated if the other conditions of the
criterion recur repeatedly. <samples> represent the number of previous samples in which the
<recurrences> will be examined
̶ -btsecs <relative_time>—A negative number indicating the number of seconds before the data
sample time used to evaluate conditions that compare against an average
̶ -major/minor/info—The alert should require urgent action/not require immediate action/is
informational only (default)
• Where <condition> is of the format: <field><operator><value>, where field is a specific field for the
statistic, operator is: =, !=, <, >, <=, or >=, and value is the numeric value for the field
• Where <comment> is comment or additional information for the criterion
• Where <name> is a user-friendly name for the criterion

Performance alerts—Examples
Some examples of performance alerts are listed as follows:
• To generate an alert for each port that has more than 50 write IOPS in a high-resolution sample:
̶ cli% createsralertcrit port write_iops>50 write_port_check
• To generate an alert for each port that has more than 10,000 write IOPS in a high-resolution sample if the sum of
all write IOPS for all ports exceeds 50,000:
̶ cli% createsralertcrit port write_iops>10000,sum_write_iops>50000
write_port_check
• To generate an alert if an hourly sample discovers that the sum of all the write IOPS performed on node 1 exceed
10,000:
̶ cli% createsralertcrit port –hourly –port 1:*:* sum_write_iops>10000
node_1_writes

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 72

Learning check
3. What is Headroom Utilization?
A. A comparison of the amount of storage space left to the historical consumption growth
B. A comparison of the current IOs to a theoretical maximum IOs that a similar device can deliver
C. A comparison of the current host connections to a maximum that an HBA can handle
D. A comparison of the system’s licensed capacity to the currently used capacity

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 73

Data Reduction Optimizations, Upgrades, and Data Migration

This section covers topics associated with product optimization and maintenance tasks such as data reduction
optimizations, upgrades, and data migration.

HPE GreenLake for Block Storage data reduction overview

HPE GreenLake for Block Storage data optimization flow

HPE GreenLake for Block Storage data reduction technologies, based on data reduction (compression, deduplication,
and packing), as well as thin conversion and thin copy reclamation, achieve advanced data reduction through
leveraging built-in hardware and software capabilities.
The savings depend on the type of data. The savings are seen in the range of 1.5:1 to 6:1.
For data that cannot be reduced, like host encrypted and zipped data, your customer can choose thin provisioned
only. This saves CPU cycles, and features include:
• Built-in
• In-band
• Reservation-less
• Integrated

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 74

Data reduction technologies overview

HPE GreenLake for Block Storage OS is designed for thin provisioning and advanced data reduction.
Sequences of incoming data blocks are analyzed inline in 16 KB chunks. Chunks full of zeroes are not written to the
back end, while the remaining chunks are deduplicated, and finally, compressed.
Thin provisioning draws fine-grained increments from a single free space reservoir without any preallocation. Other
vendors’ implementations might require a separate, preallocated pool for each data service level.
Application Data compaction ratio*
Virtual desktop infrastructure (VDI) 3.75 to 9.0X

Virtual server environments 2.25 to 3.75X

Databases 3.0 to 3.75X

Usual data savings using deduplication and compression

Note
*Based on an internal HPE study, the average data compaction savings per workload is derived from HPE Storage
telemetry data at the time of publication. HPE Store More Guarantee might be available for other workloads with a
storage assessment. Your customer should contact their HPE Sales or channel partner representative for more
information.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 75

Configuration best practices

This subsection covers best practice rules when configuring HPE GreenLake for Block Storage.

HPE GreenLake for Block Storage high availability

Traditional enclosure-dependent array set up

The concept of “RAIDlets” allows your customer to selectively create highly available (HA) volumes. Data availability
is guaranteed, even when your customer loses a complete drive enclosure.
Traditional storage arrays use enclosure-dependent RAID: Enclosure failure might mean no access to data as
illustrated in the graphic above.
HPE GreenLake for Block Storage uses enclosure-independent RAID:
• RAIDlet groups are built across enclosures
• Data access is preserved with an HA enclosure
• RAIDlets are automatically configured by the system if the number of enclosures and drives allows it

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 76

Encl.
?/ ?2 ?0 ?3 ?1 ?4
RAID 60 volume
Wide striping
Encl.

@/ @2 @0 @3 @1 @4 R6 R6 R6
A1 A2 A3 A4 A5 A6
B1 B2 B3 B4 B5 B6
Encl.

C1 C2 C3 C4 C5 C6
A/ A2 A0 A3 A1 A4
D1 D2 D3 D4 D5 D6
Encl.

B/ B2 B0 B3 B1 B4

HPE GreenLake for Block Storage enclosure-independent array setup

Drive configuration best practices: Single enclosure

Visualization of single-enclosure drive configuration

In the case of a single enclosure configuration the smallest supported number of drives is eight SSDs of the same size.
Drives can be added in increments of two SSDs of the same capacity as already installed.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 77

Drive configuration best practices: Multi-enclosure

16 drive minimum 24 drive minimum

Optimal 4 drive upgrade Optimal 6 drive upgrade
48 drive maximum 72 drive maximum

2-enclosure configuration 3-enclosure configuration

Visualization of multi-enclosure drive configurations

In the case of a multi-enclosure configuration, rules are more complex:

• The smallest supported initial number of drives per enclosure is eight SSDs of the same size
• The minimal supported drive upgrade increment is two SSDs of the same capacity as already installed
• The optimal minimal upgrade is two SSDs per enclosure
• Distribute the drives symmetrically across the enclosures whenever possible
• A system with a total of three enclosures will autoconfigure an High Availability (HA) cage (enclosure) and RAID 6
4+2
If your customer would like an HA disk (SSD) and a higher set size like 10+2, your customer will need to manually
change the default provisioning

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 78

HA cage vs. HA disk with a three-enclosure system

HA cage HA disk
• Default configuration with 3 enclosures • Manually configurable
• CPG layout: RAID 6 set size 4+2 HA • Layout: From RAID 6 set size 4+2 to 10+2
• Logical disks are created on SSD across enclosures • Logical disks are created across any available SSD
• Resulting net capacity with 72 x 15.26 TB SSD • Resulting net capacity with 72 x 15.26 TB SSD 10+2
610.1 TiB 762.6 TiB
Layout and availability differences for a three-enclosure system

There are two levels of availability for a three-enclosure system:

• Enclosure/cage (Default): No two members of the same RAID set can be in the same drive enclosure. This helps
ensure that data is still available if access to an entire drive enclosure is lost
• Drive (lower): No two members of the same RAID set are in the same drive. This allows a wider stripe with fewer
drive chassis and increases space efficiency, but at the same time decreases resiliency
A system with a total of three enclosures will autoconfigure the HA cage (enclosure) and RAID 6 4+2:
• An HA cage protects against the loss of a complete enclosure, which is a very unlikely scenario
• If your customer wants to configure an HA disk with a higher set size like 10+2, they will need to manually change
the default provisioning

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 79

Enhanced software architecture: Accelerated write-through architecture and single-node

virtual volume ownership

Paths from both controllers being reported with the same preference

The traditional storage array layout is based on mirrored DRAM write cache and must be protected by batteries from
data loss in case of a power failure. Volumes are accessed through more than one controller.
The new architecture eliminates the need for write caches and backup batteries. This results in a simpler architecture
with fewer dependencies and failure points.

Note
AO = Active optimized paths
ANO = Active nonoptimized paths

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 80

Write operation sends data to one controller that then mirrors it to the second controller

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 81

Data gets destaged from cache to disks

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 82

Only paths leading to the given LUN/volume owning controller are reported as optimized

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 83

ALUA-aware server writing data only via optimized path

Note
ALUA = Asymmetric Logical Unit Access

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 84

Data destaged from staging area to the actual volume

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 85

Traditional mirrored write cache architecture data flow

Traditional mirrored write cache architecture with no operations taking place

The traditional storage array layout details are as follows:

• Write cache is essential with an HDD for IO acceleration
• The volatile DRAM cache must be mirrored and protected from data loss by batteries
• Data write flow:
̶ Host write goes to the local controller node write cache
̶ Write gets mirrored to the second controller write cache
̶ Writes are acknowledged when written to the second controller write cache
̶ Data gets compressed and deduplicated when destaged to the back end
• Data read flow: Data is read from cache or the back end, if there is no cache hit

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 86

Data written to the first controller’s cache gets mirrored to the second controllers’ cache

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 87

Data gets compressed and deduplicated when destaged to the back end

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 88

Data is read from the cache or back end, if there is no cache hit

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 89

Enhanced software architecture: Accelerated write-through architecture

Architecture layout with no operations taking place

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 90

Data written by the host is written to both read cache and VVSTAGE areas

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 91

Data deduplicated, compressed, and destaged to the volume’s logical disks

The HPE Alletra Storage MP architecture provides extremely low latencies, which:
• Eliminates the need for write cache mirroring and write cache battery backup
• Makes node failover simpler and faster
• Offers performance like on mirrored write caching architectures thanks to the low-latency NVMe back end and the
reduction of overhead
• Uses new architectural components, including triple RAID 1 SA spaces without deduplication and compression:
̶ VVSTAGE, one for each VV (very fast host write data staging space)
̶ VVLOG, one for each VV (very fast metadata and pointer space for recovery purposes)

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 92

• Data write flow:

̶ Writes go simultaneously to the read cache and the VVSTAGE space
̶ Writes are acknowledged after they are written to VVSTAGE
̶ Data in VVSTAGE gets asynchronously destaged, compressed, and deduplicated in full stripes to the SD RAID 6
LDs disks of the VV
• Data read flow: Data is read from cache or the back end, if there is no cache hit

Note
SA = Snapshot Admin
SD = Snapshot Data

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 93

Data is read from cache or the backend, if there is no cache hit

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 94

Enhanced software architecture: Single-node virtual volume ownership and failover

A VV and all logical disks of a VV are owned by one node

How single-node virtual volume ownership and failover works:

• Each VV is assigned to one single node. This means that all metadata and data processing for a volume is done by
the owning node
̶ The vLUN path connecting to the owning node is tagged with ALUA active/optimized
̶ The vLUN path to the non-owning nodes is tagged with ALUA active/nonoptimized
• The initial volume-to-node assignment cannot be controlled, and it cannot be altered later either. It follows a
round-robin-like pattern
• If a node fails, its volume ownerships transfer to the other nodes in the cluster (a simple ID pattern-based algorithm
creates an even distribution of the volume count between all the surviving nodes)
• The host operating system must be configured to only use the active/optimized path
• A nonoptimized path is still active, and using it will cause internode traffic, which could have negative impacts on
the storage performance

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 95

VVs are automatically distributed across all available nodes

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 96

Host multipathing only has Active Optimized paths to the VV owning node

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 97

If a node fails, the VV and all logical disks are failed over and distributed to the remaining node

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 98

Maintenance
This subsection covers maintenance tasks such as software updates, servicing, and health check.

Software updates and recommendations

HPE GreenLake for Block Storage update process UI

These features allow your customer to easily identify available software updates across the fleet, and prioritize the
critical ones first. Software updates can be performed from HPE GreenLake Data Ops Manager, a local UI, or the CLI.
HPE GreenLake for Block Storage automatically downloads software updates when connected to the internet and
runs readiness checks on the array when updates are prestaged:
• If the readiness check fails, click the Failed link for details
• Readiness checks can be run again after the issue has been remediated

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 99

Details of a failed readiness check

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 100

Data Ops Manager: Running readiness checks and installing updates

Running readiness checks and installing updates in HPE GreenLake for Block Storage UI

The UI allows your customer to run readiness checks and then install updates. Updates also provide information on
the time to update and the update path. It will also inform on requirement for node reboots or restarts, if applicable.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 101

User notified of update requirements

Notification of an update with added warning of a system restart in HPE GreenLake for Block Storage UI

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 102

On-array management/local UI

Each array only receives a verified set of required updates

Telemetry is collected from the array and connected SAN components and sent back to HPE. That version
information is filtered, allow-list and deny-list rules are applied, and a set of recommended updates is generated. That
set of recommendations is then displayed in the management tools, so users know which updates apply specifically to
their array. This gives the customer full control of updating their arrays.

Validate the Send telemetry

environment data to HPE.
then install array
updates.

Analyze telemetry
Send
recommendations. data, then create
allow and deny lists.

Telemetry data analysis and update flow

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 103

HPE GreenLake for Block Storage on-array management UI

Click here to start a guided

maintenance activity.

Click here to see, verify, and install

recommended updates.

HPE GreenLake for Block Storage systems screen showing how to run a maintenance check

Now we will cover some of the actions you can perform from the Quick actions screen:
• Service storage system—From here, users with an appropriate role can initiate service actions. The Service dialog
guides your customer through the selection of serviceable components and provides helpful information to
consider before they start the upgrade or repair operation
• Create maintenance window—Normally, maintenance mode is set automatically when the user services the
system to repair or upgrade the array. There might be other reasons, such as bringing a node or system down to
test an RC failover mode. In this case, the customer does not want to receive any calls from HPE and just wants to
perform a test, which is why this option was added. A free-form reason field lets your customer describe the
maintenance and a list of durations and types of maintenance from which to choose. When in maintenance mode,
the banner is shown, and the user can come to the system detail page to see how long the window is open and why
it is open. This can also be edited to add more time. If there was not enough time, or they need to update the
reason, it can also be edited

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 104

• Check health—If your customer notices any unusual behavior with their HPE GreenLake for Block Storage
systems, they should consider conducting a health check. Before they run the check, they will typically encounter a
screen that is mostly blank, showing a hyphen (-) in the last run time. They should click Run to start the check.
After completion, your customer can click Check health again to see the results. The results will display a list of
issues, sorted by severity, along with a summary, detailed information about each issue, and brief guidance on
corrective actions for Check health

Check health

Quick action menu to perform Check health

The Check health procedure checks and displays the status of storage system hardware and software components, it
can check for unresolved system alerts, display issues with hardware components, or display information about
unoptimized VVs.

HPE GreenLake for Block Storage Onboard UI for Check health

By default, the CLI checkhealth command checks most storage system components, but your customer can also
check the status of specific components.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 105

The procedure can be started from:

• DSCC (Data Ops Manager)
• Onboard UI
• CLI (checkhealth):

̶ cli% checkhealth
̶ Checking alert
̶ Checking cabling
̶ Checking cage
̶ Checking cdm
̶ Checking cert
̶ Checking dar
̶ Checking date
̶ Checking file
̶ Checking fs
̶ Checking host
̶ Checking hostkeys
̶ Checking ilo

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 106

Check health UI in Data Ops Manager

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 107

Checkhealth CLI command

The syntax of the commands are:
checkhealth [<options> | <component>...]
Options:
• -lite—Performs a minimal health check
• -svc—Performs a thorough health check. This is the default option
• -full—Performs the maximum health check. This option cannot be used with the -lite option
• -list—Lists all components that checkhealth can analyze
• -quiet—Suppresses the display of the item currently being checked
• -detail—Displays detailed information regarding the status of the system

Note:
The <component> is the command specifier, which indicates the component to check. Use the -list option to
get the list of components.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 108

Using the checkhealth command

Use the checkhealth command without any specifiers to check the health of all the components.

cli% checkhealth
Checking alert
Checking cabling
Checking cage
Checking cert
Checking dar
Checking date
Checking file
Checking fs
Checking host
Checking ld
Checking license
Checking network
Checking node
Checking pd
Checking pdch
Checking por
Checking qos
Checking rc
Checking security
Checking snmp
Checking task
Checking vlun
Checking vv
Checking sp
Component -----Summary Description ----- Qty
Alert New alerts 7
Cage Cages not on current firmware 8
LD LDs with reduced availability 3
-------------------------------------------
3 total 19

Summary and detailed information for the hardware and software components

If there are no faults or exception conditions, the checkhealth command indicates that the system is healthy.

Example of a healthy system report using checkhealth

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 109

HPE GreenLake for Block Storage customer self-repair option

Data Ops Manager repair wizard example

Replacing and upgrading parts is safe, extremely easy, and there are many instructional videos provided for your
customer. The HPE GreenLake for Block Storage is designed for customer self-update and repair. The array
components that can be serviced by the customer are:
• IO modules:
̶ Memory DIMMs
̶ Boot drive
̶ IO adapter (OCP)
̶ FC and GbE SFP
̶ Fans
• NVMe drives
• Power supplies
• CDM
• External cables

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 110

Repair actions are guided from DSCC Data Ops Manager or the on-node management GUI.

Note
Watch the Learning about Data Services Cloud Console Customer Self Update and Repair video to see how to
perform software updates as well as hardware upgrades and repairs from either DSCC or the on-node UI.

Part replacement

Data Ops Manager maintenance UI during a system service

The maintenance window action is used to prevent unnecessary events and alerts from being sent. These events and
alerts are logged, but will not generate alerts or service calls.
Your customer should be aware of the following, when initiating maintenance:
• When driving the repair from the CLI, create a maintenance window
• The onboard UI automatically creates the maintenance window when service is initiated
• Always initiate the repair from the UI (Data Ops Manager, onboard UI, or CLI)
• Only remove the faulty device when instructed by the automation

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 111

HPE Call Home

HPE GreenLake for Block Storage Call Home settings UI

HPE highly recommends enabling the Call Home functions to get the most out of your systems. Call Home enables
HPE to provide predictive remote support and cloud analytics, Call Home is configured in the HPE on-array UI in the
Telemetry panel under Settings. Your customer’s firewall must also be configured to allow port 443:
• Enable Call Home options are enabled and disabled
• Enable remote support access options are:
̶ Allow non-root access
̶ Allow access, including root, only after asking me
̶ Do not allow access

Note
All telemetry data sent to HPE is encrypted and subject to the HPE Data Privacy Policy. See the Hewlett Packard
Enterprise Privacy Statement website for more information.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 112

HPE GreenLake for Block Storage CLI

Every maintenance activity started from the HPE GreenLake for Block Storage UI also has a corresponding CLI
command. The table below shows some examples:
Command Summary

servicenode Informs the system that a certain component will be replaced, and
causes the system to indicate the physical location of that component

servicecage Prepares a drive cage for service

servicemag Prepares a disk drive for service

createmaint Creates a maintenance window record

showmaint Displays maintenance window records

setmaint Modifies the maintenance window record

HPE GreenLake for Block Storage CLI Service and maintenance commands

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 113

Learning check
4. What is the optimal minimal upgrade number of SSDs in multi-enclosure configuration?
A. 2
B. 2 per enclosure
C. 4
D. 4 per enclosure
E. 24

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 114

Data migration
This subsection presents various options for data migration to HPE GreenLake for Block Storage with focus on Peer
Motion and Online Import.

Storage migration methodologies

• Host software solutions such as HPE Zerto, VMware
vSphere® Storage vMotion®, Robocopy, Double-Take Done by the customer,
Host-based

• DB tools such as Table Copy, Standby DB, DB HPE Services, or an HPE

Export/Import, ASM partner
Software • Backup/restore

Done by the customer,

Volume manager such as LVM, VxVM HPE Services, or an HPE
partner

Usually done by
Data migration SAN appliance
SAN appliance such as CirrusData
HPE Services
as a service
SAN-based

Done by the customer,

HPE Peer Motion and HPE Online Import for selected
HPE Services, or an HPE
source arrays (see HPE SPOCK)
HPE Storage partner
array capability Primary Data Migration Method for virtually any third-
Done by HPE Services
party block storage array to HPE GreenLake for Block
as a service
Storage, HPE Primera, or HPE Alletra 9000

HPE GreenLake migration options

HPE has a large series of tools and partners we can choose from to choose the right migration method for the right
workloads.
Which migration method is used depends on different factors such as online/offline, type of server operating system,
requested speed, agentless or not, and so on.
This is not an exhaustive list, but shows the most commonly used methods today for HPE and HPE partner projects
for their customers’ projects.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 115

Enlargement of right-hand side of previous graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 116

HPE Peer Motion

Peer Motion topology information in HPE GreenLake for Block Storage UI

With HPE Peer Motion, your customer can move data from HPE 3PAR, HPE Primera, and HPE Alletra 9000 Storage
arrays to HPE GreenLake for Block Storage built on HPE Alletra Storage MP—online, nondisruptively, and without
complex planning or dependency on extra tools. HPE Peer Motion is built into the onboard HPE GreenLake for Block
Storage UI.
Available options to use it are:
• Option 1: Migration with the HPE GreenLake for Block Storage UI
• Option 2: Migration with the Peer Motion Utility
Supported source array models are:
• HPE 3PAR Storage 7000, 7000c, 8000, 9000, 10000, 20000
• HPE Primera C630, A630, C650, A650, C670, A670
• HPE Alletra 9060, 9080

Note
For further information, visit the HPE SPOCK website for configuration details and the HPE Support Center for
documentation and manuals.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 117

Peer Motion and third-party Online Import

Source arrays* HPE 3PAR, Host servers Target arrays*

HPE Primera
NetApp Dell EMC Hitachi IBM HPE GreenLake for Block Storage,
HPE Alletra 9000
Fibre HPE Alletra 9000, HPE Primera
Channel
SAN

Peer Motion Utility/

Online Import Utility server

Source plug-ins Target plug-ins

HPE 3PAR HPE
Alletra
IBM
HPE Primera
HDS
HPE GreenLake
EMC for Block Storage
Hitachi Vantara
HiCommand
SMI-S Provider Business logic
Dell EMC
SMI-S Provider

Peer Motion and third-party migration architecture

Legend

Peer Motion enables your customer to migrate volumes and host configuration information from HPE 3PAR, HPE
Primera, or HPE Alletra 9000 arrays to a destination HPE GreenLake for Block Storage system without changing host
configurations or interrupting data access.
Online Import enables your customer to migrate volumes and host configuration information from selected third-
party arrays to a destination HPE GreenLake for Block Storage, HPE 3PAR, HPE Primera, or HPE Alletra 9000
storage system without changing host configurations or interrupting data access, in most cases.
Peer Motion Utility and Online Import Utility coordinate the movement of data from the source to the destination
while servicing IO requests from hosts.
During data migration, host IO service to the source storage system takes place through the destination storage
system. The host and volume presentation implemented on the source storage system is maintained on the
destination storage system.

Note
*See SPOCK for the list of supported source and target arrays.

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 118

Enlargement of left-hand side of Peer Motion and third-party migration architecture graphic

Enlargement of left-hand side of Peer Motion and third-party migration architecture graphic

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 119

Learning check
5. Which UI can be used for Peer Motion? (Select two.)
A. Storage Management Utility (SMU)
B. Remote Web Console
C. Local array UI
D. Data Ops Manager
E. StoreServ Management Console (SSMC)
F. Peer Motion Utility

6. Write FYI a summary of the key concepts presented in this module.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 120

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 121

Summary
• HPE GreenLake for Block Storage delivers mission-critical storage at mid-range economics with the industry’s first
disaggregated, scale-out block storage with 100% data availability. Built on HPE Alletra Storage MP, modular,
disaggregated hardware, and delivered and managed via the HPE GreenLake cloud platform
• HPE GreenLake for Block Storage can be managed and monitored via multiple GUI and CLI options, but the main
one is HPE GreenLake Data Service Cloud Console. It offers a full range of provisioning, replication and
maintenance tasks
• HPE GreenLake for Block Storage offers options similar to HPE Alletra 9000 when it comes to data reduction,
maintenance, upgrades, and migration. The difference is scalability and internal architecture that sometimes may
require a different planning approach

Course ID 0001207682
Rev. 24.11
HPE GreenLake for Block Storage Built on HPE Alletra Storage MP Hardware
Confidential | For Training Purposes Only Page 122

This page intentionally left blank.

Course ID 0001207682
Rev. 24.11