Hyper-Converged Computing
Chris Dunk, Senior Technical Marketing Engineer
• Hyperconverged and Software Defined
• SdS Architecture types and technology
• Introducing Cisco HCI Solution
• Cisco HCI Solution Overview
• HCI Use cases and positioning
Agenda
Cisco Integrated Infrastructure
Integrated Infrastructure
Fast Deployment
Minutes not days
Management and
Operation Simplicity
Intuitive UI
Simple Elasticity
Modular scale
Cost Efficiency
Predictable step growth
Minimise Risk
Validated designs
Any Workload
Mission critical apps
Flexibility
Virtualised and bare metal
Scale in Any Dimension
Building blocks
HyperConverged
SIMPLICITY
SPEED
SCALE
(Increasing Overlap)
Market Trends – Convergence and Automation
Cisco UCS: Platform for All Architectures
Converged InfrastructureCisco UCS
Hardware Abstracted
Policy Driven
Open and Programmable
UCS Mini
Mainstream Computing
Fourth Gen UCS
UCS C3000 Series
UCS M-Series
Modular Servers
Scale Out Computing
Hyperconverged Systems
Software Defined Storage Defined
• The programming that controls the storage is decoupled from the physical hardware
• Consists of the Storage Tier Only
• Emphasises storage services such as deduplication or replication, instead of storage hardware
• A shared pool that runs on commodity hardware
• Utilises Policy Based Management
• Often referred to as Storage Virtualisationand is part of a bigger industry trend of Software Defined Data centre
Definition: What is Software Defined Storage (SdS)
Software Defined Storage
Local Physical Storage
Hosts
Storage Software can run on bare
metal or as a VM on a hypervisor host
• New Converged Infrastructure offering utilisingSoftware Defined Storage
• Tight integration of x86 servers for compute and storage, networking and virtualisation in all-in-one appliance. Packaging
• Integration of hypervisors and physical infrastructure
• Simple Scale one unit/appliance at a time
• Simple deployment measured in hours
• Centralised management, intuitive UI
• Storage presentation via a distributed filesystemor object store
Definition: What is Hyperconvergence?Controller
VMVM VM VM
Local Physical Storage
Hypervisor
Controller
VMVM VM VM
Local Physical Storage
Hypervisor
Controller
VMVM VM VM
Local Physical Storage
Hypervisor
The Promise of Hyperconverged Systems
Efficient
Adaptable
Deploy at Cloud Speed
Instantly Provision, Clone or Snapshot Applications
Flexible DeploymentVariety of Configurations to Handle Diversified Set of Workloads
Scale-As-You-GrowAdd Resources Non-Disruptively and Scale Performance Linearly in Small Increments
Agile
Hyperconvergence
Fast Time to MarketSimplicityFirst Gen HCI
Hyperconvergence
Gaps
Fast Time to MarketSimplicity
New Management SilosInefficient Scaling
and Data OptimisationLimited App Performance
and Workload Support
First Gen HCI
Gaps New Management SilosInefficient Scaling
and Data OptimisationLimited App Performance
and Workload Support
IT Is Looking for a Better Answer
Requirements
Fast Time to MarketSimplicity
Agility
Simplicity+
Integration with Existing Data Centre
Easy Scaling+
Resource Efficiency
Existing Apps+
Next Gen
AdaptabilityEfficiency
First Gen HCI
• Competing for mind-share with traditional Integrated Infrastructures – many customers are now looking more closely after VMware market validation
• Companies like Nutanix, Simplivity, VMware are selling complete, supported, all-in-one appliances with message of simplicity/scale
• New projects like VDI allow customers to try new solutions like Hyperconverged in the name of cost-savings, while keeping segregated
• Software-Defined Storage (SDS) is a good fit for certain customer requirements (i.e. ROBO). Cisco has partnership examples at large customers
Hyperconverged Market
• Vendors generally saying the same thing – ease of scaling, simple to deploy/size, lower cost vs. SAN
• There are a lot of discrepancies here and Cisco is doing our homework to separate fact from fiction
• Reality is two main tiers exist – those that are trying to replicate enterprise storage and those who are simply SDS software serving a purpose like retail/VDI – Cisco will play in both
• Many limitations exist outside of vendor designed test plans/POC in areas of data integrity, management, scaling and performance
• NOT one-size-fits-all. Does not solve every business need. Cost savings are often inflated
Disruptive Reality
Customer Criteria Converged HyperConverged
Compute, Storage, Network, Server Virtualisation in one Flexible, allows choosing from
Multiple Vendors
Rigid, single vendor offers or
limited choices
Scalability Flexible, can scale up and out
as needed
Has Limitations
Common Management Software Increasingly offered as an
option
Built in by default
Packaged as a Single Unit Possible through SI Offered by default
Pre-Sized and Pre-configured Possible through SI Offered by default
Time to Deployment Days and weeks Hours
Pricing Perceived as high Perceived as low
Performance, Quality, and Uptime High – Designed and validated
for high quality and availability
at various scales and speeds
Low – Tests fail at higher
scales and speeds
Converged vs. HyperConverged
Software Defined Storage Benefits
• Most solutions offer quick deployment and cluster setup
• Simple to scale clusters
• Easy for VM administrators to use – no prerequisite SAN expertise needed
• Dashboards and built-in analytics are comprehensive and easy to understand
• Built in VM storage utilisation visibility and control
• Most offerings have all required advanced storage services• Dedupe, compression, backup, replication, VAAI clone, optimised snapshots
Software Defined Storage Considerations
• Emerging market space does not have the maturity yet, lack of confidence in
deploying enterprise applications
• Hype vs real benefits
• Independent compute and storage scaling, many solutions require purchase
of compute, storage and more software as an appliance
• Mixed workloads with larger active VMs results in dramatic drops in
performance levels –results in requiring more HW nodes
• No Mix and Match – Vendor lock-in
• Traditional SAN arrays offer much higher performance levels and are
changing as well
• As cluster grows the TCO and acquisition cost may surpass most converged
infrastructure offerings
SdS and Hyperconverged Architecture Types
Hyperconverged Architectures
Storage
Controller
VM
VM VM VM
Local Physical Storage
Hypervisor
Storage
Controller
VM
VM VM VM
Local Physical Storage
Storage
Controller
VM
VM VM VM
Local Physical Storage
Filesystem Storage Pool
• Storage VM installed on Hypervisor Host
• Storage Controller Virtual Machines have
direct access to all physical storage
• Pool local storage for distributed capacity
• Hypervisor host presented with an NFS
Datastore
• Examples: Simplivity, Nutanix, Maxta,
Springpath
Hypervisor
Hypervisor
Hyperconverged Architectures – Kernel Based
VM VM VM
Local Physical Storage
VM VM VM
Local Physical Storage
VM VM VM
Local Physical Storage
Filesystem Storage Pool
• Storage SW integrated into Hypervisor
Kernel
• Pool local storage for distributed capacity
• Reduced resource overhead
• Hypervisor host presented with an NFS
Datastore
• VMware VSAN; Gridstore
HypervisorStorage
Controller
SW
Hypervisor
HypervisorStorage
Controller
SW
Storage
Controller
SW
Software Defined Storage Architectures
Storage
Controller
VM
VM VM VM
Local Physical Storage
Hypervisor
VM VM VM
Storage Controller Software
Local Physical Storage
BlockStorage Pool
• Software Defined Storage 2 ways for
deployment
• VM controller on hypervisor host
• Controller bare metal installation
• Pool local storage for distributed capacity
• Storage is presented via iSCSI LUN or
NFS
• ScaleIO, Datacore, Stormagic, Atlantis
Operating System
Hypervisor
Storage
Controller
VM
HCI Architectures
Hypervisor
VM
Hypervisor Datastore
Storage Software
• Min of 2-3 nodes in a cluster providing storage
• Hardware to be similar with equivalent drive
technology
• Typical cluster scale 8 to 64 nodes
• Disks: Locally attached disks
o Hybrid: spinning disks and flash devices
o All-Flash: Flash devices only
• Network
o 1GB Ethernet OR
o 10GB Ethernet (preferred)
• “Witness” node acts as tie-breaker during availability
decisions
Storage Network
Data Tiering
• All data ingested goes through SSD caching tier
• SSD caching for read and write
• Allocation varies by vendor
• Writes are committed replicated and
acknowledged at the SSD cache tier
• Data is moved from SSD cache to spinning drives
• Time spent in cache is vendor specific
• Performance is impacted when SSD tier is
flooded with heavy IO
• Cache misses
• Some vendors also use Memory Caching
• Moving data closer to the processor
• More performance
VM
Write
Typical best practice is to have a 10:1 ratio of HDD to SSD capacity
• Data is protected by Replication or Erasure Coding
• Most SdS/HCI stacks use replication but more are switching
• Replication factor min of 2
• Cost is 50% of disk capacity
• Every block is written to min of 2 different nodes in the cluster
• Can survive a single node/device failure
• Higher device protection requires more replication
• RF3 reduces RAW disk capacity to 33%
• Most platforms requires min 5 nodes
HCI Data High Availability
VM
Replicate
Write IO
• Replication factor
• Fast rebuild time
• Simple technology
• Consumes large amounts of capacity
• Erasure Coding
• RAID like protection (5 or 6)
• More usable capacity
• Overhead is high for low number of nodes, need to have larger clusters to get full benefits of capacity efficiency
• Many vendors building proprietary versions
Data Replication vs Erasure Coding
Network Considerations
• Many solutions require Multicast and IPv6 for node discovery and
scale out
• Multicast used for exchanging metadata and storage network
heartbeats
• Redundant 10Gb networking required for storage data
• Management/VM networking separate from storage network
• Density solutions will require a large number of 10Gb uplink ports
• Direct access to switches for low latency storage communication
Cisco HyperFlex SystemsHX-Series
Cisco HyperFlex Systems
1 Complete Hyperconvergence
Unified Compute and
Network Infrastructure Wide Array of Rack and Blade Form Factors
Integrated, High
Performance
Network Fabric
Automated
Management of all
Hardware
Virtualisation
Aware
Cisco UCS: The Ideal System for Hyperconvergence
Cisco HyperFlex Systems
1 Complete Hyperconvergence
Unified Compute and
Network InfrastructureCisco HyperFlex
Cisco HX Data Platform
Data Services and Storage Optimisation
2 Next Gen Data Platform
Designed for
Distributed Storage
Cisco HyperFlex Systems
1
3
Complete Hyperconvergence
Unified Compute and
Network Infrastructure
Part of a Complete Data Centre StrategyElastic and Secure at Enterprise Scale
Cisco HyperFlex
Cisco HX Data Platform
Data Services and Storage Optimisation
2 Next Gen Data Platform
Designed for
Distributed Storage
Cisco One
Enterprise Cloud SuiteCisco ACI
HyperFlex Portfolio
HX220c Nodes
Smallest footprint
3-8 Node Cluster
(VDI, ROBO)
HX240c Nodes
Capacity-heavy
3-8 Node Cluster
(Server Virtualisation; IT/Biz
Apps, Test/Dev)
HX240c + B200 Hybrid
Compute-heavy
(CPU bound apps; VDI)
Ω Cisco UCS Fabric
Interconnects
Easy to Buy,
Deploy, Scale
• 3 base cluster
options and full
configure to order
• Annual
subscription
software model
• Factory integration
of HW/SW
including VMware
vSphere options
• Single call, WW
Cisco support
HyperFlex Key Customer Benefits
Efficient
AdaptableScale-As-You-Grow
AdaptiveScaling
Non-Stop Infrastructure
Built on the UCS Platform
FlexibleDeployment
Always-on Data Optimisation
CompleteHyperconvergence
IntuitiveManagement
CloudSpeed
All 3 Layers of Network, Compute
and Storage Intelligently Integrated
Into a Single Solution for <30 Minute
Deployment
Eliminate Infrastructure Silos and Use
Existing Management Tools to Manage
and Automate Operations
Instantly Provision, Clone or Snapshot Applications
Unified Management Using Existing
Skillsets in the Organisation
Variety of Configurations to Handle
Diversified Set of Workloads
Inline Dedupe and Compression Ensuring Hyper-Efficient Resource Utilisation
Add Resources Non-disruptively
and Scale Performance Linearly
in Small Increments
Scale Compute or Capacity to Match
Application Needs
Self-Healing Fabric Based Hyperconvergence with Cloud Monitoring
Agile
HyperFlex Differentiation
Break Down Silos
Simplify Management
Networking Integration
AdaptableEfficientAgile
Deploy the Entire System Faster, Not Just Computing and Storage
Standardise Your Configurations and Apply Across All Your HCI Platforms
Manage in the Same Way You Manage Your UCS Blade Servers, Rack Servers and Converged Infrastructure Solutions
Use Cases
Primary HyperFlex Use Cases
• Agile provisioning
• Frequent iterations
• Instant cloning and snapshots
Test and Development
• Low upfront costs
• Consistent performance
• Predictable scaling
Virtual Desktop Infrastructure
• Reduce operational complexity
• Adaptive scaling
• Always-on resiliency
Server Virtualisation
• Simple deployment
• Centralixedmanagement
• No “fly-and-fix” missions
Large RemoteBranch Office
Cisco HyperFlex Systems Deployment
Fast and Easy Installation
STEP 0
STEP 1
60 MINUTES
OR LESS
VMware and HX Data
Performance Software
Installed on Servers at
Factory/Reseller Before
Shipping to Customer
Rack Up Servers, Power
on, and Add to vCenter
Drag and Drop
Configuration (JSON) File
Create, Cluster,
and Datastore
Start Provisioning VMs
HyperFlex Deployment WorkflowCustomer Onsite
Download HX Data Platform/UCSM bundle from Cisco.com
HyperFlex Installer: UCS Configuration• Auto-config. of UCSM policies and core templates• Instantiate and assign Service Profiles
Configure ESX pre-requisites (manual)• Set static IP Address for ESXi hosts• Add hosts/datacentre/cluster to vCenter
HyperFlex Installer: HX Data Platform Cluster Creation
UCS Integrated HyperFlex Networking
Shipped From Factory With Integrated Networking
High Performance, Flat and Predictably Latent Fabric
Simplified Network Deployments• No specialised protocol
requirements such as IPv6, Multicast
Plug and Play Networking with UCS Unified Fabric
Reduced Complexity, Simplified Decision Making and Deployment
CVM DataPort GroupData vSwitch
CVM Mgmt Port GroupManagement vSwitch
Predefined DC Network Policies
CONTROLLER VM CONTROLLER VM CONTROLLER VM CONTROLLER VM
UCS: Embedded AutomationRapid Deployment of HyperFlex with Service Profiles
Uplink port configuration,
VLAN, VSAN, QoS, and
EtherChannels
Server port configuration
including LAN and SAN
settings
Network interface card (NIC)
configuration: MAC address,
VLAN, and QoS settings;
host bus adapter HBA configuration:
worldwide names (WWNs), VSANs,
and bandwidth constraints;
and firmware revisions
Unique user ID (UUID),
firmware revisions,
and RAID controller settings
Service profile assigned to server,
chassis slot, or pool
configuration: MAC
address,
VLAN, and QoS settings;
host bus adapter HBA
configuration:
worldwide names (WWNs),
VSANs,
configuration: MAC
address,
VLAN, and QoS settings;
host bus adapter HBA
configuration:
worldwide names (WWNs),
VSANs,
configuration: MAC
address
VLAN, and QoS settings;
host bus adapter HBA
configuration:
worldwide names (WWNs),
VSANs,
Application Profiles
Virtualisation Policy
Network Policy
Storage Policy
Server Policy
Storage SMENetwork SMEServer SME
1Subject Matter Expert
Define Policies 2Policies Used to Create
Service Profile Templates 3Service Profile Templates
Create Service Profiles 4Associating Service Profiles
with Hardware Configures
Servers Automatically
Service
Profile Templates
Pre-Defined at the
Factory for Quick
Deployment
Hyperflex Deployment Factory Automation
HX Servers Will Be Pre-Installed at Manufacturing With the Following Firmware and Software:
UCS 2.2(6f)
ESXi 6.0 U1
HX Data Platform
HX Server CIMC will be Configured in UCSM Mode
UCS 2.2(6f) Is the Minimum Required Version to Support Hyperflex
Newer UCS and HX Data Software Versions Will be Installed When they Are Qualified
Powered by UCS Platform
Hyperconvergence Meets Unified ComputingHyperFlex HX-Series
HCI Built on Cisco UCS
SystemLinkIntegrated
Compute
Centralised
Management
Optimised for
Virtualisation
Scale Without
Complexity
Low Latency
Physical & Virtual
Data & Management
Proven Performance
Enterprise Reliability
Blades & Rack Mount
Complete HW Abstraction
Policy & Profile Driven
Optimised for Scale
Comprehensive Hypervisor
Integration
Native I/O Virtualisation
Hypervisor Switching in HW
Maintains Single Point of
Management
Self-Integrating
Fewer Components
Unified Management Across UCS Portfolio
UCS Mini
Edge-Scale
ComputingCloud-Scale
Computing
UCS M-Series Composable Infrastructure
UCS B-Series Blade Servers
UCS C-Series Rack Servers
UCS HX Series Nodes
Core Data Centre Workloads
UCS Management
UCS Central Policy-based Management Architecture
• Simplified, centralised management
• Based on UCS Manager, which remains the execution engine
• Global policies are read-only in UCS Manager and are only stored on UCS Manager instances when they are in use
• Local policies are read-only in UCS Central with visibility to all policies in all UCS Manager instances
• Global policies, local policies, or a combination of both may be used at the same time
User
Access
User
Access
Full Two Way
Configuration Changes
With Traceability
Element
Manager
Element
Manager
Element
Manager
UCS
Manager
UCS Central
Cisco UCS HyperFlex Security
Hyper Flex
Security
Integrated
Protection
Dynamic Control
Unmatched Agility
Through a pervasive, policy-driven approach
Resource
lifecycle
Provisioning Spin up Security at the Same
Time as Other Resources
MonitorEnforce Policy Automatically,
Even When Workloads Move
DecommissionRemove Associated Rules and
Policies Automatically
Cisco HyperFlex with Cisco ACI
HXCONTROLLER
HYPERVISOR/CONTAINER
Web App DB AppWeb
HXCONTROLLER
HYPERVISOR/CONTAINER
Web App DB AppApp
HXCONTROLLER
HYPERVISOR/CONTAINER
App App DB WebWeb
EPG EPG EPG EPGEPG EPG
EPG EPG EPG
EPG EPG EPGEPGEPG EPG EPGEPG EPG
Web Web Web AppWeb
EPG EPG EPG EPGEPG
Diverse Workload Deployments at Enterprise Scale
HX DATA PLATFORM
Simplified Capacity Expansion
HXCONTROLLER
HYPERVISOR/CONTAINER
EPG
HX DATA PLATFORM
Web App App DBDB
EPGEPG EPGEPGEPG
EPG
HXCONTROLLER
HYPERVISOR/CONTAINER
Cisco HyperFlex Systems
Automated,
Application Centric
Policy Model
Flexible
Application
Service Policies
Elastic Infrastructure
CONTROLLERHYPERVISOR CONTROLLERHYPERVISOR
Hyperconverged Data Platform
Cisco ONE Enterprise Cloud Suite
PublicCloud
Cisco HyperFlex
SeriesAgile and Efficient
Infrastructure
Capacity
Augmentation in
AWS and Azure Leverage Public Cloud to
Manage Peaks
Back Up and Async DR
VM VMVM
VM VMVM
VM VMVM
VM VMVM
CONTROLLERHYPERVISOR
VM VMVM
VM VMVM
Hyperconverged Data Platform
Cisco HyperFlex SystemsAdaptable Infrastructure
Adaptive Scaling
on PremAdaptive Scaling
to Public Cloud
PublicCloud
“Out of the Box” Hybrid CloudCisco HyperFlex with Cisco Enterprise Cloud Suite and Cisco ACI
IaaSNative Secure Cloud
SegmentationHybrid Cloud
1. Factory configured Cloud ready Infrastructure and Management SW
2. Complete infrastructure automation including HX
Create/edit/delete data store
Add capacity—nodes / discs to existing node
3. Rich API set for PaaS integration
1. Default micro segmentation with best in class virtual services
2. HX controller created in secure zone w/ Distributed Firewall
3. Micro segmentation w/ ACI infrastructure
1. Capacity Augmentation
with consistent policy to
manage peaks
Deploying enterprise workloadin cloud
2. Back Up and Async DR
Cisco HX Data Platform Overview
Integrated Management and
Data Services
Dynamic Data Distribution
Continuous Data Optimisation
Independent Scaling of Compute and Capacity
Real Innovation
Hyperconverged Scale Out and Distributed File System
CONTROLLERVMHYPERVISOR
VM VM VM
HYPERCONVERGED DATA PLATFORMHYPERCONVERGED DATA PLATFORMHYPERCONVERGED DATA PLATFORM
Start With as Few as Three Nodes
Hyperconverged Data Platform
Installs in Minutes
Add Servers, One or More at a Time
Linearly Scale Compute, Storage
Performance, and Capacity
Distribute and Rebalance Data Across Servers Automatically
Retire Older Servers
HYPERCONVERGED DATA PLATFORM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
Software Modules Inside a Server
CONTROLLERVMHYPERVISOR
VM VM VM
HYPERCONVERGED DATA PLATFORMHYPERCONVERGED DATA PLATFORMHYPERCONVERGED DATA PLATFORMHYPERCONVERGED DATA PLATFORM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
Controller VM Has Direct Access to Drives
VAAI Plugin Offloads Snapshots and Clone Operations
IO Visor Module Presents NFS to ESX and Stripes IO
DATASTORE/VOLUME
CONTROLLER
VMHYPERVISOR
VMVMVMVMVM
HDD
HDD
SDD
SDD
IO Visor
VAAI
Building on the Right FoundationCisco HX Data Platform
Built From the Ground Up for
Hyperconvergence
Distributed Log-Structured File
System Designed for Scale-out,
Distributed Storage
Advanced Data Services (Snapshots,
Clones) and Data Optimisation
(Inline Dedupe, Compression) Without
Trade-offs
Better Flash Endurance and
Disk Performance
Computing, Storage,
Networking, and Hypervisor Integration
Eliminates Management
Silos
Distributed File system
Local FS Local FS Local FS Local FS
Unique
Architecture
Independent Scaling of Compute and Capacity
Scale Compute
HX Data Platform
Add NodesScale Capacity Within Nodes
HX Data Platform
CONTROLLERHYPERVISORCONTROLLERHYPERVISOR CONTROLLERHYPERVISOR CONTROLLERHYPERVISOR
IOVisorIOVisor
IOVisorIOVisor
IOVisorIOVisor
IOVisorIOVisor
VM VM VM VM
Non-HyperFlex Hosts Can Connect to
Storage with IOVisor
VM VMVM VM VMVM VM VMVM VM VMVM
Independent Scaling of Compute (Blades) and Storage Capacity (Rack Servers)
HX
DA
TA
PL
AT
FO
RM
B200 M4 Blades
Connect Existing Blades for Compute
Add More Blades for Scaling Compute
Add HYPERCONVERGED
Rack Servers
Start with HYPERCONVERGED Rack
Servers (Compute + Storage Capacity)
Dynamic Data Distribution
• HX Data Platform stripes data across all nodes simultaneously, leveraging cache across all SSDs for fast writes
• Balanced space utilisation: no data migration required following a VM migration
Systems built on conventional file
systems write locally, then replicate,
creating performance hotspots
CONTROLLERHYPERVISORHYPERVISOR CONTROLLERHYPERVISOR CONTROLLERHYPERVISOR
VM VMVM VM VMVM VM VMVM
HX Data Platform
VM VMVM
CONTROLLERCONTROLLER
…and Efficient Capacity and Network Utilisation
HYPERVISOR HYPERVISOR HYPERVISOR
VM VM
1 2 3
VM
DATASTORE
CONTROLLER CONTROLLER CONTROLLER
Balanced Space UtilisationNo Data Migration on VM Migration
Less Stress on Network
High Resiliency, Fast Recovery
Platform can sustain simultaneous 2 node failure without data loss; replication factor is tunable
If a node fails, the evacuated VMs re-attach with no data movement required
Replacement node automatically configured via UCS Service Profile
HX Data Platform automatically re-distributes data to node
CONTROLLERHYPERVISORCONTROLLERHYPERVISOR CONTROLLERHYPERVISOR CONTROLLERHYPERVISOR
VM VMVM VM VMVM VM VMVM VM VMVM
HX Data PlatformHX Data Platform
Non-Disruptive Operations
DATASTORE DATASTORE
• Stripe blocks of a file across servers
• Replicate one or two additional copies
to other servers
• Handle entire server or disk failures
• Restore back to original number of copies
• Rebalance VMs and data post
replacement
• Rolling software upgrades
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
CONTROLLERVMHYPERVISOR
VM VM VM
File.vmdk
D1 E1A1 B1 C1B2 A2 A3C2 C3 D2D3 E2E3 D1E1 B3 B3
EDCBA
Data Protection and High AvailabilityData Protected by Replication of Data Across the Cluster Nodes
VM
Replicate
Write IO
Replication Factor = 3 Replication Factor = 2
Default Is Replication Factor = 3 Set at Cluster Creation Time
Every Block Is Written to 3 Different Nodes in the Cluster Every Block Is Written to Min of 2 Different Nodes in the Cluster
Higher Availability to Survive Multi-Point Failures;
Higher Device ProtectionCan Survive a Single Node/Device Failure
Reduces Raw Disk Capacity to 33% More Usable Capacity; Uses 50% of Raw Disk Capacity
Access Policy Is Strict (Default); Can Be Modified Via CLI Access Policy Is Lenient
Continuous Data Optimisation
BEFORE
Inline Deduplication
20–30% space savings
Inline Compression
30–50% space savings
No Special Hardware
No Performance Impact
Log-Structured File System Yields More Efficient Data Optimisation
Integrated Management and Data Services
• Clones and snapshots are pointer-based and space-efficient
• Creation and deletion do not impact performance
• Instant provisioning, cloning and snapshotting of virtual machines from within vCenter
• No separate console, no learning curve
• UCS Manager familiar to 50,000 customers WW
• Server and network deployment settings in pre-configured Service Profiles
Fast and Flexible Native Snapshots for Backups
• Pointer-based snapshots
• Space-efficient
• Fast creations and deletions
• Fine-grained or coarse-grained
• VM-level or VM folder-level
• VAAI-integrated
• Quiesced and crash-consistent
• Use vCenter Snapshot Manager
• Policy-based
• Schedules, retention period
Native VM Clones for Rapid Provisioning
• Batch creation GUI
• Apply unique names
• Use customisation spec to apply IP
• Powerful tool to rapidly setup a large set of VMs using just VC (without scripting or View composer); Up to 256 clones in parallel per job
• Golden/Base VM can be a template, powered on or powered off
• Pointer-Based Writeable Snapshots (Instantaneous Clones)
• VAAI integrated
• VM-level granularity
Management Built for Server Admin100% vCenter-Based Management
• Extends virtualisation
management seamlessly
• No switching between
management consoles
• View storage alerts/alarms alongside
with ESX alerts/alarms
• Command line interface for automation
vCenter Plug-in: HX Data Platform Summary
vCenter Plug-in: Datastore Provisioning
Capacity Reporting in UI
• Capacity: Total formatted with RF
• Free: unused capacity
• Total Savings:Combined % savings of dedup+comp
• Compression: % recovered from compression
• Deduplication: % recovered from deduplication
• Provisioned Capacity:Amount of space allocated in DS
• Thin provisioned
• Used Capacity:Total of physical space consumed in DS
• Available Capacity: Provisioned minus Used
Cloud-based auto-support
Daily reports
Periodic heartbeats
Instant alerts on critical events
Cluster visibility
Support use cases
Real-time monitoring for critical events
Proactive and advanced warning
Preliminary troubleshooting and investigation support cases
Access currently limited to support
Cloud-Based Monitoring and Analytics
Infrastructure for Next Generation Hyperconvergence
HCI Solutions Combine Computing, Networking, Storage for rapid
deployment
Elimination of management silos within and between data centres
and remote sites simplify operations
Unique capacity scaling and pricing models allow you to scale as you
grow
Participate in the “My Favorite Speaker” Contest
• Promote your favorite speaker through Twitter and you could win $200 of Cisco Press products (@CiscoPress)
• Send a tweet and include
• Your favorite speaker’s Twitter handle
• Two hashtags: #CLUS #MyFavoriteSpeaker
• You can submit an entry for more than one of your “favorite” speakers
• Don’t forget to follow @CiscoLive and @CiscoPress
• View the official rules at http://bit.ly/CLUSwin
Promote Your Favorite Speaker and You Could Be a Winner
Continue Your Education
• Demos in the Cisco campus
• Walk-in Self-Paced Labs
• Table Topics
• Meet the Engineer 1:1 meetings
• Related sessions
Q & A
Complete Your Online Session Evaluation
Learn online with Cisco Live!
Visit us online after the conference
for full access to session videos and
presentations.
www.CiscoLiveAPAC.com
Give us your feedback and receive a
Cisco 2016 T-Shirt by completing the
Overall Event Survey and 5 Session
Evaluations.– Directly from your mobile device on the Cisco Live
Mobile App
– By visiting the Cisco Live Mobile Site http://showcase.genie-connect.com/ciscolivemelbourne2016/
– Visit any Cisco Live Internet Station located
throughout the venue
T-Shirts can be collected Friday 11 March
at Registration
Thank you