- No category
advertisement
HPE Hyper Converged 250
March Update
March 2016
Simple. Scalable. Highly Available.
Refined with HPE Hyper Converged 250
March 2016 Confidential
HPE Hyper Converged 250
Hardware Components
HPE Hyper Converged 250
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
HPE Hyper Converged 250
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)
HPE Hyper Converged 250
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
~ 8.3TB HA storage for 4-node systems
5.6TB Hybrid Storage 4x 1.2TB SAS + 2x 400GB SSDs
2
~ 7.5TB HA storage for 4-node systems
7.2TB Capacity Storage 6x 1.2TB SAS
~ 11.5TB HA storage for 4-node systems
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
( ) 2x Intel E5-2680v3 / 12-core @ 2.5GHz
( ) 1x Intel E5-2660v3 / 10-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.
20 January 2016 Confidential
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
20 January 2016 Confidential
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
HPE Hyper Converged 250
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
HPE Hyper Converged 250
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
HPE Hyper Converged 250
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)
HPE Hyper Converged 250
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
HPE Hyper Converged 250
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
(Automated; 7 minutes)
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)
(Automated; 2 minutes)
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
HPE Hyper Converged 250
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
20 January 2016 Confidential
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
HPE Hyper Converged 250
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
Hyper Converged 250 node
recoverNode
recoverNode
Hyper Converged 250 node
Hyper Converged 250 node
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
HPE Hyper Converged 250
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
Using the Centralized Management Console
Create Management Group and/or Cluster
1. Use the CMC Getting
Started Wizard
2. Create New Management
Group
3. Follow wizard and guidelines
Using the Centralized Management Console
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
Using the Centralized Management Console
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
advertisement
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project