now, what comes next? Henrik Leerberg Data Centre Expo, October 2014 1
Data center trends affecting the space (from a CIO perspective) > Capacity lock-down > Transparency > Connecting the data center with the business > Agility and speed of deployment > Right-sizing of resources > Increasing outsourcing of all kinds 3 Source: 451 Group, NEXT-GENERATION DATACENTER MANAGEMENT, From to DCSO, January 2014, Uptime Institute Data Center Industry Survey, 2013
THE DATA CENTER AS A BUSINESS 4
The data center as a corporate business Cost center P/L business Resources Services +/Break-even 5
Workloads will drive the business in the future 6
Choosing the right deployment strategy An oversized data center Capex for future opex savings! Determining cost/service Other Factors Criticality Privacy Security Scalability Seasonality Data center deployment strategy Public cloud Hybrid cloud Private cloud On premise Fact-based decision 7
Tracking cost per service Cooling? Floor space? Network? Power? a SERVICE Storage? Security? Software? Rack space? Labor? Compute? CIO needs Transparency Insight Analytics Operations Agility & speed Effectiveness & efficiency Accountability Think service 8
THE SOFTWARE-DEFINED DATA CENTER 9
Software-defined IT Today Software-defined IT (Compute, storage, network, perimeter etc.) SHARED resources SHARED resources 1 server = many applications 1 data center = many data centers Tomorrow Software-defined Physical Infrastructure (power, cooling, space etc.) -defined data center DevOps will treat the data center as a a utility 10
Software-defined data center Services ITSM SLA Applications VMs Hypervisor Server Physical Infrastructure 2,000 workloads = 340 VMs = 52 physical servers Efficient power, cooling...compute, storage, network & perimeter 11
Software-defined data center understands service needs Service 1, SW-Def DC 1 Service 2, SW-Def DC 2 Will be able to create many small data centers Will match physical infrastructure requirements for services Continually reserves, maintains and understands how physical resources align with the required services Service 3, SW-Def DC 3 Predicting service availability 12
Balancing utilization in the software-defined DC An example Services SLA Applications VMs ITSM Resource allocation Reservation Protection Reduces risk & costs 98% hotspot hours 41% cooling energy Hypervisor Server Physical Infrastructure Reserve power, cooling and rack space Protect Dynamic cooling control 13
extra Balancing utilization in the software-defined DC An example Services SLA Applications ITSM Resource allocation Reservation Protection Power on/off & Power capping Enforce global cap Increase rack loading VMs Hypervisor Server Physical Infrastructure Reserve power, cooling and rack space Protect Dynamic cooling control Dynamic power control Power capping 14
Balancing utilization in the software-defined DC An example Services SLA Applications ITSM Resource allocation Reservation Protection IT asset utilization - identify: Underutilized servers Comatose servers Consolidation candidates VMs Hypervisor Server Physical Infrastructure Reserve power, cooling and rack space Protect Dynamic cooling control Dynamic power control Asset optimization 12% average IT utilization across any size data center Source: Schneider Electric study 15
Converged infrastructure Services SLA Applications VMs ITSM Vendors Vendors Hypervisor Server Physical Infrastructure - Open architecture - Restful APIs / web services - Public interfaces / ETL - Bi-directional data flows 16
Benefits of the converged infrastructure ITSM + Converged infrastructure CIO > Transactions linked with $$$ > Deployment strategies > Elasticity (dynamic loads) DevOps/IT > Focus on service, SLAs and applications > Treat the data center as a utility ( blackbox ) - Provides speed and agility > Automatic consolidation of workloads Facilities > All facility & IT assets owned by facilities > Dynamic power & cooling environment control - Automatically shutdown infrastructure > Reduced number of moves, adds and changes - 1 server = many applications 17
is a core piece of the puzzle Mapping from services... Services SLA Applications VMs Economic pressures Business relevant IT Understanding the cost Data center deployment strategy Software-defined data center Converged infrastructure IT asset ownership changing Hypervisor Server Physical Infrastructure...to rack, power and cooling... 18
A NEW? 19
New? + Server monitoring Server power capping Server Access (vkvm) Capacity management Network topology Virtualization management Run-book / Orchestration IT Service Finance Service management SLA management Performance management Availability management Cloud management Power & Cooling monitoring Power & Cooling control Building Management Systems Space management Sensors, Access & Security Asset management Capacity management Facilities Cost of service Energy procurement Financial management Business planning & analytics Risk management Charge-back management Compliance 20
What 451 Group predicts DCSO incorporates components The components of DCSO are built on the foundation of Source: 451 Group, NEXT-GENERATION DATACENTER MANAGEMENT, From to DCSO, January 2014. 21
in the future > is on most CIO/IT Infrastructure Execs roadmaps for 2015 > The deployment risks around are considered low > Based on: - Architecture fit - Marketplace maturity - Security - Support 22 Source: CEB Emerging Technology Roadmap, 2014-2017
in the future 2014 is managing only 7% of racks in global data centers Primarily within service provider data centers (incl. colocation & telcos) By 2017 tools will be deployed in more than 60% of the larger data centers in North America. 23 Source: Gartner Magic Quadrant for tools, September 2014
Your choice! 24
Thank you! Questions? 25