HPE Hyper Converged 250

HPE Hyper Converged 250

March Update

March 2016

Simple. Scalable. Highly Available.

Refined with HPE Hyper Converged 250

March 2016 Confidential


Hardware Components


Chassis

• (2 - 4) Nodes

• (12 - 24) Disk Drives

• (2) 1400W Hot Swapable Power

Supplies

• (8) Redundant Fans

• (1) SUV Connector (Dongle)

• (1) Support Shelf for Racking


XL 170 Compute Node

• (2) CPUs – Selectable Options

• (16) DIMMs – Selectable Options

• Embedded (2) port 1GbE NIC

• iLO 4 Advanced port

• (1) 10-GbE 2-port SFP+ NIC

or

(1) 1 GbE 4-port NIC

• P440 Smart Array controller with Internal mini-

SAS connection

• (1) USB 3.0 Connector and

(1) SUV Connector

Datacenter Processing

HC250 offers processing choice and flexibility

Intel VMCS Shadowing

• Accelerated nested virtualization

Per Core Power States

• Optimize cores by workload

Low I/O Latency

• Bypasses memory

Industry leading power efficiency

• 22nm, 3-D Tri-gate transistors

Powered by Intel

Disk Failure Protection without compromise

HC250 uses Hardware RAID technology

• 4GB Flash Backed Write Cache

• 12Gb Internal mini-SAS port

• Drive Erase

• SSD RAID 0 Write Optimized

• PCI Express Gen3

• Drive Parameter Tracking for Predictive

Analysis

• Dynamic Sector Repairing for automated bad sector remapping

Powered by

HP P440

Smart Array Controller

HPE Integrated Lights-out (iLO)

iLO Advanced v4

• Pre-power up diagnostics

• BIOS Configuration

• Queries and Displays Group Health Status

• Remote Chassis Management and

Configuration

• HPE RESTful API

Networking

HC250 Connectivity Choices

HPE Ethernet 10Gb 2-port 560FLR-SFP+ Adapter

Featuring the next generation of 10Gb Ethernet integration offering a single chip solution on a FlexibleLOM form factor, further reducing power requirements for 2 ports of 10Gb Ethernet.

HPE Ethernet 1Gb 4 -port 331FLR

Adapter

Features the next generation of Ethernet integration, offering a single - chip solution in a Flexible LOM form factor, which offers the option for future upgrades to 10GbE networking speeds. It delivers full line - rate performance across all ports with low power consumption.

Plug and play

– 8x network, 2x power, go!

2x 1400W

Power Supplies

HPE SmartArray P440

(with 4GB cache)

2x 10GbE for primary network connectivty

(SFP+ ports)

4

2

3

1

Serial/USB/Video

(fits HPE 36-pin Dongle)

USB 3.0 connector

( for USB-based Factory Reset)

2x 1GbE

RJ 45 ports

(not used)

* Ports shown available on all nodes. 1GbE ports are not intended for use. Customers are advised against changing the configuration of the pre-configured virtual switches. 1GbE connectivity cannot be used instead of 10GbE.

Dedicated iLO4 port

1GbE for initial installation of first appliance

(RJ 45 port)


Ordering Guide

Our most compact and powerful virtualization platform

Hyper Converged 250 for VMware vSphere

– Compact 2U platform with up to 4 compute/storage nodes based on HPE Apollo 2000 Gen9

– 2, 3 or 4 identically configured nodes in starting configuration with

–

(2) Intel Xeon E5-2640v3 or E5-2680v3 processors, or (1) Intel Xeon E5-2660v3 processor

–

–

128GB, 256GB or 512GB memory

2-port 10GbE SFP+ network adapter, or

4-port 1GbE RJ45 network adapter

–

All Flash, Hybrid and Capacity Storage Options

–

1-year 9x5 NBD on-site warranty, 1-year 9x5 software support

1

–

Every chassis is comes with redundant fans and requires 200-240V power supply

–

Co-existence with ConvergedSystem 200/250-HC StoreVirtual

(ConvergedSystem 240-HC, 242-HC and 250-HC StoreVirtual models)

Storage Option Drives

9.6TB All-Flash Storage 6x 1.6TB SSDs

2

~ 15.5TB HA storage for 4-node systems

6.4TB Hybrid Storage 4x 1.2TB SAS + 2x 800GB SSDs

2


5.6TB Hybrid Storage 4x 1.2TB SAS + 2x 400GB SSDs

2


7.2TB Capacity Storage 6x 1.2TB SAS


1

Software support covers HPE software included with product only; it does not cover hypervisor support. 3- and 5-year HPE Care Pack Services available.

2

SSD capacity as used by the HC 250 after additional capacity reservation to meet performance and wear life targets.

•

Apollo r2600 Chasis 24 SFF

•

HPE ProLiant XL170r Gen9 Server

Nodes

20 January 2016 Confidential

Appliance simplicity with server flexibility

1-2-easy

– Ordering the next-generation hyper-converged solution

– Discrete options for simple yet customizable configuration

– Choices apply to all four nodes in the system for optimally balanced configurations

– ConvergedSystem 200/250-HC StoreVirtual equivalent configurations for compatibility

HPE Hyper Converged 250 for VMware vSphere

Base system

Quote licenses:

[ ] VMware vSphere Enterprise

[ ] VMware vSphere Enterprise Plus

[ ] VMware vSphere for Desktop

[ ] VMware Horizon* (View, Adv, Ent)

Nodes in the System:

( ) 2 nodes (New!) ( ) 3 nodes ( ) 4 nodes

Node options

Hypervisor pre-integrated:

( ) VMware vSphere 5.5

( ) VMware vSphere 6.0

CPUs per node:

( ) 2x Intel E5-2640v3 / 8-core @ 2.6GHz



(New!)

Primary network connectivity per node:

( ) 2P 10GbE SFP+

( ) 4P 1GbE RJ45

(New!)

Amount of memory per node:

( ) 128GB

( ) 256GB

( ) 512GB

Drive configuration per node:

( ) 6x 1.2TB SAS

( ) 4x 1.2TB SAS + 2x 400GB SSD

( ) 4x 1.2TB SAS + 2x 800GB SSD

(New!)

( ) 6x 1.6TB SSD

(New!)

Selection applies to all nodes in the system

* When including VMware Horizon licenses on the same quote, Horizon-specific components are not pre-installed on the system.


HPE Hyper Converged 250 for VMware vSphere

List of SKUs launching on February 16, 2016

Group SKU

Base

Software

Description

M0T03B HPE Hyper Converged 250 System for VMware vSphere

M0T04B HPE Hyper Converged 250 Node

P9B50A HPE Hyper Converged 250 Software LTU for VMware vSphere 5.5

P9B51A HPE Hyper Converged 250 Software LTU for VMware vSphere 6.0

All CPU/memory/networking/storage options are ISS options, all components sold as part of the HC 250 count towards PL FE quota.

Included with System/Nodes:

•

StoreVirtual VSA, Microsoft Windows VM with OneView InstantOn,

•

•

•

OneView for vCenter, Software Update Manager, StoreVirtual CMC

VMware vSphere pre-integrated; Enterprise/Enterprise Plus licenses must be purchased for HC 250

(or use licenses out of Enterprise License Agreement/vCloud Air Agreement, or un-used Enterprise Purchasing Program tokens)

1-year NBD parts/labor/onsite warranty; 1-year 9x5 SW support

Base system is quoted with 200-240V AC power supplies


HPE HC250 (Tiered) - Hybrid Storage

• Up to four server nodes in 2U, scale out to up to 16 nodes in one cluster

• Hybrid Tiered Storage Options

• 1.2TB SAS and 400GB SSD

• 1.2TB SAS and 800GB SSD

New!

Tier 0

Hybrid Storage

Tier 1

StoreVirtual vSphere / StoreVirtual (Boot device)

StoreVirtual

Nodes 1.2TB SAS & 400GB

SSD (NR10)

2

3

4

New!

3.8TB Useable

5.7TB Useable

7.5TB Useable

1.2TB SAS & 800GB

SSD (NR10)

4.3TB Useable

6.4TB Useable

8.3TB Useable

HPE HC250 (non-Tiered)

• Up to four server nodes in 2U, scale out to up to 16 nodes in one cluster

• Capacity storage options

• 1.2TB all SAS

• 1.6TB all SSD

New!

Capacity Storage vSphere / StoreVirtual (Boot device) / Management VM

StoreVirtual

Nodes

2

New!

1.2TB SAS (NR10)

5.8TB Useable

New!

1.6TB SSD (NR10)

7.7TB Useable

3 8.6TB Useable 11.5TB Useable

4 11.5TB Useable 15.5TB Useable


Network RAID and Availability

Disk RAID, Network RAID

RAID: Redundant Array of Inexpensive Disks

•

Drives are the weak link in reliability

– we EXPECT them to fail

•

All StoreVirtual products include hardware RAID to preserve volumes through drive failure and rebuild

Network RAID provides an additional protection above hardware RAID

•

HW RAID + NR 10 

“five nines” of availability (AKA “the gold standard”)

•

Best protection at the expense of storage efficiency (2TB storage for each 1TB volume)

•

Default for SV4000, CS250-HC appliances, and VSA

Five Nines of WHAT?

Availability

is the percentage of time in a year that a volume provides storage services to the host. It’s the “uptime” of storage.

HW RAID + NW RAID 10, 5 or 6 provides “five nines” of availability.

Availability %

90% ("one nine")

95%

98%

99% ("two nines")

99.9% ("three nines")

99.99% ("four nines")

Downtime per year

36.5 days

Downtime per month

72 hours

18.25 days

7.30 days

3.65 days

36 hours

14.4 hours

7.20 hours

8.76 hours 43.8 minutes

52.56 minutes 4.38 minutes

99.999% ("five nines") 5.26 minutes

99.9999% ("six nines") 31.5 seconds

99.99999% ("seven nines") 3.15 seconds

25.9 seconds

2.59 seconds

Downtime per week

16.8 hours

8.4 hours

3.36 hours

1.68 hours

10.1 minutes

1.01 minutes

6.05 seconds

604.8 milliseconds

262.97 milliseconds 60.48 milliseconds

99.999999% ("eight nines") 315.569 milliseconds 26.297 milliseconds 6.048 milliseconds

99.9999999% ("nine nines") 31.5569 milliseconds 2.6297 milliseconds 0.6048 milliseconds

We live here.

StoreVirtual in 5 Easy Bites: Network RAID

Provides options for protection vs. space efficiency

Network RAID 0: stripe the data over the nodes

•

Hardware RAID protects against (multiple) drive failures

•

Loss of a node causes loss of access to the volume

Network RAID 10: stripe and mirror the data across nodes

•

Protects against node failure

•

Costs twice as much (1TB volume requires 2TB)

Network RAID 5 & 6: stripe with parity encoding across nodes

•

Covered later in the course

The choice is per-volume

•

Essential volumes  NR 10

•

Non-critical volumes  NR 0

B

C

D

A

Cluster

A

D

A B

C B

C D

Network RAID 10 with Single 4-node Appliance

Network RAID-10

2

B/A1

1

A/D1

D/C1

4

C/B1

3

Network RAID 10 in Cluster with Four Systems

A B C D E F G H I J K L M N O P

Network RAID 10

5

D’

E

1

A

P’

7

F’

G

3

B’

C

L’

M

13

H’

I

9

N’

O

16

J’

K

11

6

E’

F

2

A’

B

8

G’

H

4

C’

D

M’

N

14

I’

J

10

O’

P

16

K’

L

12

Network RAID 10 in Cluster with Four Systems

A B C D E F G H I J K L M N O P

Network RAID 10

Node 1

Node 2

Node 3

Node 4

Node 5

Node 6

Node 7

Node 8

Node 9

Node 10

Node 11

Node 12

Node 13

Node 14

Node 15

Node 16

A B C D E F G H I J K L M N O P

A

P’

A’ B

B’ C

C’ D

D’ E

E’ F

F’ G

G’ H

H’ I

I’

J

J’ K

K’ L

L’ M

M’ N

N’ O

O’ P

23


StoreVirtual Provisioning

StoreVirtual in 5 Easy Bites: Provisioning

All these choices are available for each volume:

Protection (Network RAID level)

Thin Provisioning

•

Create a volume that eventually will be 10TB

•

Only consume storage in 50GB increments

•

Auto-grow it to 10TB as your application expands

Adaptive Optimization

•

Mix SSD (flash) and traditional HDD storage

•

Frequently accessed data automatically moves to the faster SSD drives

Snapshots

•

Instantaneously capture the state of the volume

•

No data copied until needed

Unmap

•

Allow storage to be reclaimed when no longer needed

Thin Provisioning

No Reservation Required

Cost Savings

•

Purchase only what you need today

•

Allocate pages as needed

•

Growth is automatic (set by policy)

•

No reserve necessary

•

Integrated with SmartClones, Snapshots and

Remote Copy

Delay Future Storage Purchases

•

Grow capacity as needed

Simple to Manage

•

Enable/disable per volume

Full Provisioned

Allocated 200 GB

Available 0 GB

Used 100 GB

Wasted 100 GB

50 GB written

50 GB written

200 GB Volume

Thin Provisioned

Allocated 100GB

Available 100GB

Used 100GB

Wasted 0GB

50 GB written

50 GB written

200 GB Volume

Waste

Available


Adaptive Optimization

Network RAID 10 & Adaptive Optimization

CS 200HC-SV with Hybrid Storage Pack

Maximizing the effectiveness of Flash real estate

Hypervisor Hypervisor

– Network RAID is configured at the volume-level and replicates across storage nodes

A

D’

Tier 0

A

1

2

Hypervisor

Cluster with NR10 & AO

A’

Tier 0

B

B’

Tier 0

C

– More I/O performance due to optimization on each storage node

D’

AO

Tier 1

4

A

3

A’

AO

Tier 1

B

B’

AO

Tier 1

C

H’

E

E’

F

F’

G

Node 1 Node 2 Node 3

Hypervisor

C’

Tier 0

D

C’

AO

Tier 1

D

G’

H

Node 4

Workload acceleration

Adaptive Optimization

– Hybrid Model ONLY

– Adaptive Optimization is a StoreVirtual feature

– Dynamic movement between the SSD and SAS storage tiers for optimized utilization and performance

– 16x 1.2TB SAS drives + 8x 400GB SATA SSD

– 16x 1.2TB SAS drives + 8x 800GB SATA SSD

– Granular and efficient movement of data at 256KB

Tier 0 Tier 1 vSphere / StoreVirtual (Boot device)

StoreVirtual

StoreVirtual

Data block “heat map”

Tier 0

Access Frequency

923

740

523

160

Tier 1

Access Frequency

142

113

42

12

Adaptive Optimization

Simple Application Acceleration

StoreVirtual Cluster

 Tier 0 is always ~100% utilized

 Granularity is a single LeftHand OS page (256 kB)

 Each storage system continuously maintains a heat map about the pages and moves data accordingly



Disabling “Adaptive Optimization” on a volume will cause pages to be gradually moved to the lower tier

System

AO permitted

AO permitted

AO not permitted

System

Tier 0

(2 SSD)

Tier 1

(4 HDD)

Tier 0

(2 SSD)

Tier 1

(4 HDD)


StoreVirtual Snapshots and Smart Clone

Snapshots

A Snapshot is the state of a volume at a point in time

• a collection of pages whose contents were “frozen” in time

• no data are copied when the snap is taken

• Snapshots are incremental from previous snap

“Snap @ 11 includes pages 7 and 8”

“Snap at midnight is Snap @ 11 plus page 9”

• Data copied when needed (e.g. “copy Snap @ 11 to tape”)

• Snapshots are widely used by applications

• scheduled from Windows or VMware

(“AASM”)

• snapshots consumed in multisite SANs (“RIPC”)

• snapshots used as VDI copies (“Smart Clone”)

• snapshots controlled by upper level site failover (VMware SRA)

... metadata

7

10:23

8

10:24

7

8

9

9

22:11

Cluster

7

8

...

Snap at 23:00

Snap at 24:00

9

Application Aware Snapshots

Cluster

Snapshot Orchestration

Windows

VSS

AASM

Agent

The application must participate in snapshot creation

•

The application has data in flight

• Only the application “knows” when a snap will be meaningful

VSphere

Client

AASM

Agent

•

Example: a database needs a snap of the data and index volumes at the same moment.

• Example: a virtual machine needs a snap of both its “home” volume and its data volumes at the same moment.

•

Example: a virtual machine needs to stop and flush its memory before taking a backup.

100

GB

Idx

Vol

1 TB

Data

Vol

How it works:

•

StoreVirtual AASM agent installed on the host.

• CMC sets the snap schedule (or asks for one “now”)

•

AASM agent communicates with the appropriate host service, asking for a signal when the snap can be safely taken.

•

Sometimes the snap is copied to a backup service such as HP Data Protector or Veeam.

Smart Clone

Efficient volume copies

Realize cost savings through cloning

•

Pro-active, not reactive de-duplication.

•

No duplicate data, only changes are stored.

•

Greatly improved read performance.

•

Write performance becomes COFW.

• Up to 100s of volumes from a single “base”.

•

Managed the same as non-cloned volumes.

Use cases

•

Virtual desktop images.

•

Boot drives for like blades, rack servers, virtual machines.

•

Read-only data volumes for multiple servers and/or applications.

•

Volume copies images for test and development.

0GB

SAN/iQ Cluster

100

GB

0GB

SmartClone Volumes

0GB 0GB


Software and Deployment

Software ready-togo for true ‘plug-and-play’

All software preinstalled, “no assembly required”

Hyper-converged systems from HPE provides virtualized compute and HA storage with rich data services.

Pre-integrated software components

• VMware vSphere 5.5u2 or 6.0u1

*

(upgradable, mix is supported)

• HPE StoreVirtual VSA with LeftHand OS 12.5 for clustered, highly available storage

• HPE OneView InstantOn 1.3

• HPE OneView for vCenter 7.8

* vSphere and vCenter licenses must be purchased for the HC 250. Valid software license required for preinstalled software. Customer with ELA and vCloud Air Partner can reuse their licenses on this platform. When including VMware Horizon licenses on the same quote, Horizon components are not pre-installed on the system.

Mgmt VM

App App App App

StoreVirtual

VMware vSphere

HPE HC 250

Prepare the installation

Gather information about network setup and prepare workstation

Choose a network deployment model

 Flat/single network with 23 contiguous IPv4 addresses

 Multiple networks for ESXi, vMotion, storage

•

VLAN w

: 5 IPv4 contiguous addresses for ESX hosts

(3 per hosts + management)

•

1 per vSphere hosts and 1 for Management VM/vCenter

•

VLAN x

: 4 IPv4 contiguous addresses for vMotion

•

1 per node

•

VLAN y

: 14 IPv4 contiguous addresses for storage

•

5 IPv4 for StoreVirtual VSA (4+1 VIP), 8 iSCSI Initiators (2 per vSphere node), 1 Management VM

IPv6 enabled?

•

VLAN z: Virtual machines

HPE recommends 5700

 Network switches have to be IPv6 enabled or 5900 series switches

 Network services: DNS, mail server, (NTP)

SFP+ or

DAC

VLANs

w,x,y,z

Hewlett Packard Enterprise Hyper-Converged 250

HPE IP under the hood

00:00

Hardware and software component discovery and validation

(Automated; 2 minutes)

04:00 06:00 08:00

Deployment of

StoreVirtual clusters and enabling discovery by

OneView for vCenter


15:00

Connect and login to management software, configure network addresses, host names and credentials

(Customer, user dependent;

4 minutes of keyboard time)

Deploy configuration to components

(vSphere, StoreVirtual, management software)


Connect and login to vCenter

(Customer; 1 minute)

OVIO Flow

Discovery

Collect Config

Info

Network Mgmt

VM

Network Hosts

Start VSAs Build SAN

Build ESXi

Cluster

Migrate Mgmt

VM to SAN

WWAS16 | Confidential

HP IP Under the Hood:

A closer look at the software architecture

Automating for hyper-converged in less than 15 minutes

OneView InstantOn in Management VM configures the ESX servers and StoreVirtual VSAs

Key components on the CS 200-HC StoreVirtual:

OneView InstantOn & Discovery Agent

– Discovery Agent on nodes & mgmt VM uses IPv6 link-local addresses and makes the node available for deployment

– OneView InstantOn shows health data that is collected from systems

– OneView InstantOn automates the vSphere and StoreVirtual cluster creation

– iSCSI Rescan Agent automates storage availability upon system reboot.

VSA DA

VSA DA

HP OneView

InstantOn

IR

ESX

ESX

IR

VSA

VSA

DA

DA

IR

ESX

ESX

IR

WWAS16 | Confidential

Connect to the Management VM - Node 1 Port 2

Windows 2012 Start Screen

• To access the HPE HC 250, you need a laptop/workstation with a

1 GbE port that is capable of running Microsoft Windows

Remote Desktop Services

• On the WindowsStart screen, select Desktop. If HPE OneView

InstantOn does not display, click the HPE OneView InstantOn shortcut on the desktop.

OneView InstantOn

Customization of the appliance

Customer-installable and great out-ofbox experience

1.

Connect to the management VM from your workstation/laptop

2.

OneView InstantOn starts up

3.

Accept VMware/HPE EULAs and go through the wizard for

– IP address assignments

– StoreVirtual credentials

– Network services

4.

Deploy configuration out to the system/appliances

OneView InstantOn (OVIO)

Deploy…..

– IP the Management VM (2 IPs)

– IP the ESX Hosts (4 IPs per Host)

– Configure the StoreVirtual VSAs (IP, Name,

...)

– Add the ESX Hosts into vCenter Cluster (4 hosts per appliance)

– Add the StoreVirtual VSAs into SAN group/cluster (4 VSAs per appliance)

– Set the affinity rules for Management VM

(Nodes 1,2)

– Attach the SAN to the vCenter Cluster

– Create a SAN Datastore for Management

VM

– Migrate the Management VM to the SAN

Datastore


2-Node Quorum

Always-on compute and storage for small sites

Availability in two-node clusters without putting data integrity at risk

– Automatic and transparent failover in two-nodes installations, like remote and branch offices

– Low-resource alternative for “Failover Manager”, as no separate equipment in site required

Quorum Witness

– Requires a secured NFSv3 share in central site on in central data center secured NFSv3

– Multiple remote sites can use the same file share share

Networking requirements

– Supports <300ms latency

– Supports >= 1Mbps connectivity


45

Achieve high availability storage with two nodes

StoreVirtual two node quorum- no FOM required

• Low-resource alternative for Failover Manager (FOM) in two-node management groups

• Automatic, transparent failover

• No additional equipment in remote sites

• Implemented as file on NFSv3 file share in e.g. central data center

• Good fit for small, centrally managed remote sites

Quorum Witness as file on NFSv3 share

Link requirements for Quorum Witness

• Supports <300ms latency

• Supports >= 1Mbps connectivity

2-node StoreVirtual VSA provides

HA storage for Enterprise ROBO, e.g. retail environments


HC2XX Recovery

Quick Reset for HC 250 for VMware vSphere

Reset of individual notes or entire appliance

– Typical use for Quick Reset are

– Quickly recover from failed installations

– Reset the unit after a test deployment

– Reset a demo unit

– Two modes allow users to reset an individual nodes or the entire appliance (multiple nodes)

– Components reset to factory : ESXi, StoreVirtual

VSA and Management VM (incl. vCenter and

OneView InstantOn)

– Does not a compliance-safe/sanitization reset

– Available on HC 250 for VMware vSphere, and

Converged System 200/250-HC StoreVirtual that have been reimaged

20 January 2016

Two modes: Reset of individual nodes or entire appliance

Confidential

Quick Reset

Hyper Converged 250 node

–

Node 1

Zip files containing:

• StoreVirtual VSA

• Windows Management

recoverNode

(and additional scripts for each component)

(rebuilds node)

quickReset

(invokes recoverNode on nodes 2, 3, 4)

Hyper Converged 250 node

–

Node 2, Node 3, Node 4

Zip files containing:

• StoreVirtual VSA

• Windows Management

recoverNode

(and additional scripts for each component)

(rebuilds node)

Hyper Converged 250 nodes

recoverNode script will :

1.

Stop all running VMs

2.

Delete all VMs

3.

Delete all Virtual Disks

4.

Delete all non-ESX data stores

5.

Run rebuild scripts to recreate ESX VMs

6.

Reset ESX IP address back to factory DHCP setting

7.

Restart ESX

quickReset script will:

1.

Start recoverNode script on each ESX host node

(as background task to run simultaneously)

2.

Provide results of ESX four-node recovery

Recovery Folder:

/vmfs/datastore1/recovery/

Quick Reset

VM VM

VM

VM

VM VM

recoverNode

Hyper Converged 250 node


recoverNode

recoverNode



recoverNode quickReset

VM VM VM

VM VM

Mgmt

VM

READY FOR OVIO

Resetting an HC 250 for VMware vSphere

QUICK RESET

Recovery from failed installation

Reset a demo unit

Reset Test Deployment

Built into the appliance

Reset one at time or all at once

Not suitable under complete RAID failure

Available on HC Platform Stack 1.3 or re-imaged systems

USB RESET

Recovery from failed installation

Rebuild after RAID failure

Requires image download and USB key build out

Single node at a time or concurrently

Upgrade entire HC Platform Stack on demo units

Recovering Hyper-converged Appliances

Download and Thumb Drive Method

– Engineering notice went out to registered users

•

Include customers and resellers with “entitlement” (registered/licensed)

•

Did NOT include HPE Internal unless signed up to receive the notices

•

Software available on the HPE Software Depot (links in GUI and Install

Guide)

– HPE internal can access the same software via another

URL

•

HPE Presales and Support personnel

•

No registration/licensing required

•

Same Software Depot but via another URL the bypasses the login

• https://hpsc-pro-athp.austin.hpe.com/portal/site/hpsc/public/

•

HPE Passport Required to login


Adding Storage

Additional Storage for Hyper Converged Environments

Supported Storage

Add StoreVirtual storage-only resources that can also be provisioned from HP

OneView for vCenter (OV4VC)

– StoreVirtual Appliances (i.e. 4335,

4730, etc.)

– StoreVirtual VSA on x86 server

* New management group recommended; HP Peer Motion on StoreVirtual not available across management group boundaries.

Using the Centralized Management Console

Install and Discover

1. Install via standalone instructions

2. Do NOT create a Management

Group during the initial install.

3. Use CMC on HC2XX to Find

System


Create Management Group and/or Cluster

1. Use the CMC Getting

Started Wizard

2. Create New Management

Group

3. Follow wizard and guidelines


Skip Volume Creation

•

Skip Volume creation during the MG creation as a best practice.

• You can provision volumes later using same wizard or OV4VC.

57


iSCSI Initiator Association

58

Using vCenter

iSCSI Initiator Association (continued)

59

Storage Administrator Portal

– OV4VC

Adding the Storage

60

Storage Administrator Portal

– OV4VC

Storage Verification

61

Спасибо!

Ждём ваших вопросов:

+7 (495) 925-5519 [email protected]

62

No results