Enable Data Center Infrastructure Management Via Critical Monitoring: Ten Proactive Strategies to Deploy Today to be Ready for DCIM Tomorrow
Emerson Network Power: The global leader in enabling Business-Critical Continuity TM
Automatic Transfer Switch
Paralleling Switchgear
Fire Pump Controller Surge Protection Uninterruptible Power Supplies & Batteries Cold Aisle Containment
Row-Based Precision Cooling
Cooling
Rack
Extreme-Density Precision Cooling
Integrated Racks
Rack Power Distribution Unit KVM Switch Monitoring
Power Distribution Units Data Center Infrastructure Management
UPS
Presentation topics Emerson Network Power Overview Advancing Infrastructure Monitoring to Enable Infrastructure Management in the Data Center Bay Area Internet Solutions’ Building Management and Electrical Monitoring Systems Question and Answer Session
Advancing Infrastructure Monitoring to Enable Infrastructure Management in the Data Center
Matt Zieg, Manager, Product Marketing, Liebert Monitoring, Emerson Network Power
Top data center challenges External forces changing the business climate
Virtualization, Cloud IT Outsourcing
Infrastructure Management
Consolidation
Heat Density Efficiency & Green initiatives
Regulation Compliance
Facility Challenges
Availability
Energy Efficiency Higher Density
Increasing Demand Reduced Budget
Business and technology forces pressing on the data center
Source: Data Center Users’ Group Survey
Power Density
Top data center concerns Spring 2008
Spring 2009
Spring 2010
Heat Density (Cooling)
Heat Density (Cooling)
Adequate Monitoring / Data Center Management Capabilities
Power Density
Energy Efficiency (Energy Costs and Equipment Efficiency)
Heat Density (Cooling)
Availability (Uptime)
Adequate Monitoring / Data Center Management Capabilities
Availability (Uptime)
Adequate Monitoring / Data Center Management Capabilities
Availability (Uptime)
Energy Efficiency (Energy Costs and Equipment Efficiency)
Energy Efficiency (Energy Costs and Equipment Efficiency)
Power Density
Power Density
Space Constraints / Growth
Space Constraints / Growth
Space Constraints / Growth
Source: Data Center Users’ Group Survey
The Emerson approach Reducing the Cost of Design & Operation
Drive lower 1st cost Ease of deployment Lower engineering costs Energy efficiency Maximum leverage of existing resources Power & Cooling Equipment Infrastructure Services
Reducing the Cost of Management
Diagnostics for availability Performance optimization Easing change management Improved asset utilization Dynamic, real time infrastructure management Software & Communications
Data Center Infrastructure Management optimized performance A car’s performance is optimized through – Sensors – Monitoring dashboard – Real time diagnosis and data
You know how much fuel you have left Alerted if you’re overheating You know if you’re running efficiently or not
Data Center Infrastructure Management optimized performance Battery Monitors
Server Control
A data center’s performance is optimized by
KVM Switch
– Sensors and networked devices – Monitoring dashboard – Real time diagnosis & data
UPS Web Cards
Cooling Control
Power Meters
Managed Rack PDU
Temperature Sensors
Leak Detection
You know how much capacity you have left Alerted if you’re overheating You know if your facility is operating efficiently or not
Systematic approach to DCIM Measurement
Control Centralized Management
Building foundation for DCIM  Improve data center efficiency, availability and capacity utilization through infrastructure monitoring
1. Sensing temperatures 2. Monitoring power usage 3. Monitoring rack conditions 4. Detecting fluid leaks 5. Intelligent control of precision cooling
6. Intelligent control of critical power 7. Managing alerts and alarms 8. Monitoring energy efficiency 9. Monitoring batteries 10. Monitoring and managing remotely
Measurement
If You Can’t Measure It, You Can’t Control It
Data collection throughout the data center ecosystem
 Opportunities throughout the data center to collect data to enhance management of the entire ecosystem
What do I measure?
Sensing temperatures – Precision cooling return air – Rack inlets – Rack exhaust More precise control of cooling capacity 10 degree increase in server inlet temperature equals 21 percent reduction in cooling energy costs
What do I measure?
Rack conditions – Access visibility Minimize human errors – Where do I have space? Connectivity Best practice: Segment racks by density and business critical needs
What do I measure?
Fluid leaks – Top cause of unplanned downtime* – Water pipes – Condensation – Adjacent spaces (i.e. restrooms) Best practice: Around all cooling systems and under rack rows
* 2010
Ponemon Institute Study
What do I measure?
Fluid leaks – Top cause of unplanned downtime* – Water pipes – Condensation – Adjacent spaces (i.e. restrooms) Best practice: Around all cooling systems and under rack rows
* Source:
2010 Ponemon Institute Study
What do I measure?
Power usage – Data center input – UPS output – Floor PDU – Rack PDU Required for capacity management and efficiency – PUE Best practice: Intelligent Rack PDU’s
Data collection throughout the data center ecosystem
 Strategic use of data points enable holistic management of the entire space
Control
Enhance Reliability, Improve Performance, Increase Efficiency
Intelligent control Precision Cooling
Critical Power
Efficiently manage temperature set points, humidity levels and airflow Coordinate operation of multiple units as a “team” – Mitigate over- / under-cooling Best practice: Intelligent controls in room, row and rack cooling systems
% Load
FAN SPEED CONTROL SENSOR
Optimize UPS system performance – Mitigate conversion “losses” “Intelligent paralleling” can enable more efficient operation and energy optimization Best practice: Integrate intelligent controls into all UPS systems
75-80°F Variable Fan Speed
3 Units @ 25% Load Each = 90% Efficiency
70-72°F
% Load
72-75°F
2 Units @ 38% Load = 92% Efficiency TEMP CONTROL SENSOR
Centralized Management
Move Beyond Just Collecting Data– Use it to Improve Performance
Managing alerts and alarms  Centralized alarm management system can help prevent problems before they happen  Allows for faster, more effective response to system problems  Enables analysis of equipment operating trends and development of more effective preventive maintenance programs
Monitoring energy efficiency Reduce energy consumption while increasing IT productivity Track total data center consumption and automatically calculate and analyze PUE Best practice: Rack-level monitoring provides most accurate picture of IT equipment power consumption
Monitoring batteries Weakest link in UPS system #1 cause of unplanned downtime Preventive Maintenance visits are an important first step Dedicated battery monitoring provides early notifications and ensures cell deterioration is identified early enough to schedule a PM visit Emerson survey found that customers using battery monitoring experienced half as many battery failures Best practice: Implement monitoring system that connects to and tracks health of each battery within a string
Monitoring and managing remotely  Remote monitoring can lift burden of onsite infrastructure monitoring  Creates efficient use of human resources  Enhanced data analysis and specialization reduces downtime
Benefits of infrastructure monitoring Reduce power and cooling energy costs Increase system efficiencies Support higher capacities Eliminate stranded capacity Enhance data center availability Eliminate common causes of data center downtime Gain visibility and control required for data center optimization
Importance of data center infrastructure has elevated Building Management
Data Center Infrastructure
IT Management
? Companies lack cohesive management strategy for data center infrastructure
Bay Area Internet Solutions Building Management and Electrical Monitoring Systems BAISTM, Inc.: Leading Innovation in Silicon Valley’s Co-Location Datacenter Facilities
www.bayarea.net
Tom Wye President & CEO Bay Area Internet Solutions
Bay Area Internet Solutions BAIS, Inc Founded in 1995 Located in Silicon Valley, California Facilities Based MSP, offering
o Co-Location o Managed IT Solutions o High Speed Internet Access
www.bayarea.net
2050 Martin Ave Santa Clara, CA 95050
State-of-Art Facility Built in 2009
www.bayarea.net
13.5 MW Tier IV Datacenter 83,000 sq. ft. facility N+1, Fully Redundant 2,800 Tons of Cooling 8,000kVA of UPS Cold Aisle Containment 30 inch raised floor Outside Air Economizer Design Comprehensive Monitoring & Building Management System (BMS) with integrated network, mechanical, electrical and security systems
The Need for Monitoring & Management to Meet Customer SLA Requirements
Comprehensive BMS System o Temperature, Humidity, Air Pressure control Chiller Plant control
o Economizer Management Fanwalls Data Center CRAC VFD speeds Chiller Plant staging
Complete Electrical Monitoring o Power usage o Battery Cell Voltage o Alarms and Trending
www.bayarea.net
PUE Monitoring & Trending
Network Operations Center Manned 24/7 NOC • • • •
www.bayarea.net
Building Management Electrical Monitoring Security and Access Control Fire and Smoke Detection
BMS System Environmental Control Systems: • Chiller Plant • CRAC units
www.bayarea.net
BMS System Economizer Management
www.bayarea.net
BMS System Environmental Trends
www.bayarea.net
Liebert SiteScan Data Center • CRACs • PDUs • RPCs
www.bayarea.net
Liebert SiteScan Electrical Room Monitoring • Utility Feeders • 12kV-to-480v • Generators and Fuel Levels • Paralleling Gear • ATS • UPS • Battery • STS
www.bayarea.net
Real-Time PUE & DCiE Calculation
www.bayarea.net
Mechanical vs. IT Load Mechanical • UPS • PDU
www.bayarea.net
Monitoring and Measuring Energy Savings
Outside Air Economizer Design (30-90%) Cold Aisle Containment (25-95%) CRAC units with EC plug fans (10-30%) VFD upgraded Chillers and Cooling Towers (10-30%) High Efficiency UPS Systems (3-5%) High Efficiency Power Distribution Units (2-5%) Energy Efficient Lighting (30-50%)
SVP Energy Innovator Award Winner of 2009
www.bayarea.net
Summary BAIS Achieving Corporate Standards
Meeting Customer SLA Requirements o Proactive monitoring and management of all our colocation facility systems o Prevention of outages Multiple level of threshold alerts, e.g. early warning vs. critical
Efficient Control and Management of Resources o Automated Data collection, analysis, alerting and reporting Mechanical system controls and failure recovery
www.bayarea.net
o Keeping electrical cost down by managing PUE using trending data o Preventive equipment maintenance using alarms and trending data o Capacity planning
Q&A
Matt Zieg, Manager, Product Marketing, Liebert Monitoring, Emerson Network Power
Tom Wye, President & CEO, Bay Area Internet Solutions (BAIS, Inc.)