Data Center Design Utilizing Data Center Infrastructure ... · Data Center Design Utilizing Data Center Infrastructure ManagementData Center Infrastructure Management (DCIM) Chuck
Post on 30-Jul-2018
322 Views
Preview:
Transcript
Data Center Design Utilizing Data Center Infrastructure ManagementData Center Infrastructure Management
(DCIM)
Chuck Kramer - RCDD/NTS/CDCP/CDFOMSenior Solutions Consultant
Emerson Network Power
What is DCIM?
The monitoring, collection and analysis of real-time, continuous, accurate information from an intelligent infrastructure that allows for immediate action tightlyinfrastructure that allows for immediate action . . .. tightly integrated with IT operations . . . maintain high availability and control costs. . .
- Real-Time Data Center Infrastructure ManagementInfrastructure Management
What Is DCIM To Technology Industry Leaders?Industry Leaders?
Documentation of Information in the Data Center Infrastructure Devices IT Devices Cabling
Management of Infrastructure and IT Devices Capacity Consumption Change Management Change Management
What’s The Value of DCIM To The Enterprise?To The Enterprise?
• Real-Time Data Integration& li di– Power & Cooling usage, trending,
monitoring, alerting– Server Service-Processor Monitoring– Virtualization mapping to physical devices– Leak Detection
• Automated notification of warning/failure• Automated notification of warning/failure alarms
• Automatic data capture for trending and capacity analysiscapacity analysis
• Interface to existing BMS systems
Traditional Management Framework
Deliberate Design Stable Operation
Framework
APPLICATION
Plan Design Deploy Operate Maintain
LAYER
IT Managed
IT LAYER
PHYSICALLAYER
Facilities Managed
DCIM Tools TodayyDocumentation of Information in the Data Center Infrastructure Devices IT Devices Cabling
Typically Spreadsheet based orConfiguration Management Data Base Tools
Management of infrastructure and IT Devices Capacity Consumption Change Management
Building Management Systems (BMS)Other proprietary management systems Change Management Other proprietary management systems
DCIM In The Future• Closer integration between facilities and IT elements• More real-time data collection and integrating that data with
inventory• Device “discovery” and automated changeDevice discovery and automated change
detection/notification• More integration with virtual server management systems
(tracking virtual to physical)• Better interfaces for extracting and analyzing collected data• More customization tools in the hands of users vs vendor• More customization tools in the hands of users vs vendor
provided professional services
Preparing For The Futurep g• Process and Policy improvements:
– Change control– Device, cable, cabinet, room naming and
labeling– Centralizing who makes changes in the
computer room• Documenting the existing environment
– Inventory– Connections
• House CleaningDe-commission/remove “zombie” servers &– De-commission/remove zombie servers & devices
– Remove unused cabling
Preparing For The Future (cont.)p g ( )• Install “smart” devices when doing a technology refresh• Stay current with hardware and software maintenancey
– Continued reliability– Able to upgrade technology when it comes available
• Listening– Attention to current trends and directions– Vendor application/development road mapsVendor application/development road maps– Emerging products and technologies
• “Cord-mounted’ ATS’s• Monitored power cables• New or enhancements to existing DCIM Products
Initiatives in the Data CenterAGILITYAGILITYMobile Enterprise Strategy
SERVICESERVICEVirtualizationVirtualizationVoiP and UCC
Strategic PlanningData Center Data Center
Modernization andModernization andQUALITYQUALITY Strategic Planning Modernization and Modernization and ConsolidationConsolidation
Cloud ComputingCloud ComputingMicrosoft
ITIL and ProcessITIL and ProcessImprovementImprovement
Negotiating Vendor Contracts
MicrosoftWindows 7and Office
VendorIT Cost
OptimizationInfrastructureECONOMICSECONOMICSManagement
pInfrastructure Protection
Business Continuity
Source: Gartner (March 2010)
Scale Aggravates EverythingInfrastructure Costs Impact of Outage # of Activities
x100
TIME INFRASTRUCTUREDATA POINTS
x10
CLOUD
x10
VIRTUAL SERVERS
x1
PHYSICAL SERVERS
Results of Traditional Approaches
107minutes95% “ of companies have experienced an107minutes the average length of time for each outage95% . . of companies have experienced an unplanned outage within the last 24 months.”
80% t bl“In general, failed equipment, data center mishaps and insufficient resources exacerbate the frequency and duration of unplanned outages ”
$505,502 “ . . .average cost of each incident.”
80% believed the incidents were preventable
duration of unplanned outages.
Source: Ponemon Institute, Calculating the Costs of Data Center Downtime, February 2011
Analyst Perspectivey p“In order to achieve HIGH EFFICIENCY, HIGH HIGH EFFICIENCY, HIGH AVAILABILITYAVAILABILITY and LOW TOTAL COST OFLOW TOTAL COST OFAVAILABILITYAVAILABILITY, and LOW TOTAL COST OF LOW TOTAL COST OF OWNERSHIPOWNERSHIP, datacenter operators need to be able to
UNDERSTAND EACH PIECEUNDERSTAND EACH PIECE of the system asUNDERSTAND EACH PIECEUNDERSTAND EACH PIECE of the system, as well as how the pieces fit together. Datacenter infrastructure
management tools are designed to PULL TOGETHER PULL TOGETHER ALL THE RELEVANT DATA ALL THE RELEVANT DATA and
TRANSFORM IT INTO INSIGHT TRANSFORM IT INTO INSIGHT that operators can use to make decisions.”
- John Stanley, 451 Group
14
Reasons to Adopt DCIMp
• Reduce operational costs• Accurate Data for Design• Maintain high availability• Maintain high availability• Manage energy efficiency• Increase agility• Drive value across the
Availability + Capacity + Efficiency
businessToday’s high-density computing environments
i t l l f i ibilit d t l threquire a greater level of visibility and control than ever before
Stages of DCIM Adoption
What and where are
Data Capture and Planning
How do I extend
Analyze and Diagnose
How are my
Monitor and Access Recommend and Automate
How do I anticipate assets in the data center?
How are they interconnected?
Do we have space
the life of thedata center?
How do I reduce mean time to repair (MTTR)?
assets operating?
Am I getting real-time notification of alarms and alerts?
How do I get my
potential failures and automatically shift compute and physical load to eliminate downtime?
Do we have space, cooling and power to meet future needs?
How can I efficiently commission or decommission?
( )
How do I sync infrastructure with virtualization automation?
How are we doing
How do I get my server back upand running?
Can I populatemy planning tools with actual
How can I optimize efficiency across my data center?
decommission?
Improved Planning
How are we doing against SLAs?
Reduced MTTRand Effort
with actual performance data?
Early Warning(Reactive)
Availability at Optimal Performance
(Proactive)(Proactive)
Each Stage Provides Positive ROI Allowing Varying Adoption Approaches and Speedpp p
Data Capture and Planning
AutoCAD Drawings
Multiple Excel Spreadsheets Do this
1. Get an accurate inventory2 C lid iDrawings
Individuals
2. Consolidate to one repository3. Create elevation views4. Track tasks and projects5 Model current capacity and
Visio
5. Model current capacity and growth
Get this
Vendor Dependent
DatabasesGet this
1. Align inventory with needs2. Improve planning3. Identify underutilized, Vendor Dependent
MonitoringEmail and Meetings
ydecommissioned and excess equipment
Eliminate Disparate Tools – Move to One Inventory Mgt System
Monitor & Access EquipmentDo this
1. Instrument the infrastructure2. Implement one alarm
management system3. Establish remote access to servers
Battery Monitor
s
Server Control
KVM Switch
Get this1. Real-time health of the
infrastructurePower M t
UPS Web CardsCooling
Control
infrastructure2. Proactively manage issues3. Faster MTTR4 Reduce risk unplanned downtime
Meters
Temperature Sensors
Managed Rack PDU
Leak Detection
4. Reduce risk unplanned downtime
Manage the Facility & IT Subsystems
Analyze & DiagnoseDo this
1. Document processes2. Establish best practices3. Incorporate real-time data into
planning process (if you have not t)
AlarmMonitor yet)
Get this1 Increased operational control
Monitor
Access Plan 1. Increased operational control2. Identify areas of concern before
alarms are issued3. Accurately forecast capacity
Access Plan Manage
3. Accurately forecast capacity
Defer or Eliminate CapEx and Reduce OpEx
Traditional Solution : Add Buffer To Increase AvailabilityIncrease Availability
Data Center Power Allocation
Actual CapacityA di E l l i if
Buffer Limit
y
Pow
er Stranded CapacityDynamic Changes
According to Emerson calculations, if every U.S. data center could utilize
10% MORE of theirPower Usage
TimeDynamic Changes
10% MORE of their
CAPACITY, the need to build more than 2 million square feet of
“Data center power capacity includes buffers intended to absorb spikes in power use caused by peaks in resource utilization. These buffers are typically based on either nameplate or
St tiS P C li
new data center space could be eliminated each year . . . . SAVING MORE THAN $10Bnominal server power consumption or power
consumption measured at peak utilization with specific workloads.”
StaticSpace Power CoolingMORE THAN $10B
Source: Intel White Paper - Increasing Data Center Efficiency with Server Power Measurements
Recommended Approachpp“More and more organizations are seeking to eliminate the inefficiencies created when IT and facilities systems are managed
t lseparately . . . . DCIM solutions that integrate IT and facilities systems into a single management platform ill be essential tomanagement platform will be essential to optimizing the performance of the data center in the future ”center in the future.”
- David Cappuccio, Gartner
Optimize Technology Spend & Resources
Asset ManagementDCIM
• Provide data to more efficiently & effectively manage technology assets and optimize technology spend p gy p
• Establish a single global repository, using consistent processes, functions, and data – to provide a consolidated view of infrastructure assets
Bridges the gap between finance and data center operations and aligns business process by eliminating duplication of efforts
Avoid Collisions & Reduce MTTRConfiguration MgtPhysical Infrastructure
• Map the relationships of the data center assets to business services
L ITSM t h l i t h th i t f d t t k d• Leverage ITSM technologies to show the impacts of data center work orders across all layers
P id h l l f i i d f h b iProvide the level of service required for the business at a competitive cost
Next Generation: Closed-Loop Control
Requires the ability to:Requires the ability to:• Understand what is going on
in real time• Make the right decisions on• Make the right decisions on
what to do and plan our the actions to take
• Take action to correct or• Take action to correct or improve and validate the result
24
Summary• DCIM provides dynamic access into resource utilization for both
Facilities and IT
• Understanding real-time resource consumption can provide accurate data for right-sizing a designaccurate data for right-sizing a design
• Maintenance of DC inventory and resources allows for precise di d d d di f f itrending data and an understanding of future requirements
• Cost savings have proven to be substantial for customers who g padopt DCIM processes, by eliminating multiple IT and Facilities point products used for inventory tracking, monitoring & alerting, MAC work future design requirements and reportingMAC work, future design requirements, and reporting
top related