Richtlijnen design slides

Data Centre Networking Technology Update Fabrice Feyten - Christophe Crous 28 September 2009 Agenda • Data Centre Business Requirements • Some New Trends • Data Centre architectures • Overview Belgacom Data Centre Solutions • Why Belgacom ? Restricted use only 28/09/2009 Slide 2 “IT as a service” imposes new business requirements for the Data Centre (network) Availability  How can you achieve 100% uptime for your critical applications? Scalability: DC must support organisational growth  How can you keep up with the ever increasing demand for resources? (11% annual growth for servers and 22% for storage on average). Optimise use of resources  Whether due to regulatory compliance or the increasing cost and complexity of IT, organisations are consolidating and centralizing IT resources. Operational Efficiency  Companies are focusing on data-centre management and operations, attempting to cut costs, improve efficiency, and better align IT spending with business needs and service demands. This is part of a shift in IT culture to “IT as a service” with data centres as business-service delivery centres. (sources: IDC, Gartner, Nemertes) Restricted use only 28/09/2009 Slide 3 Agenda • DC Business Requirements • Some New Trends • Virtualization • Unified Fabric and DCE • Cloud Computing • Data Centre architectures • Overview Belgacom Data Centre Solution • Why Belgacom Restricted use only 28/09/2009 Slide 4 End to end Virtualization drives new DC architectures IT virtualization is the abstraction of IT resources in a way that masks the physical nature and boundaries of those resources from resource users. It is the most important and impactful trend in infrastructure and operations through 2012 Virtualization 1.0 = Consolidation (cost savings) Virtualization 2.0 = Agility (speed) Virtualization 3.0 = Cloud Computing (quality of service) Virtualization changes architectures, processes, cultures, markets, businesses — and sourcing. Restricted use only 28/09/2009 Slide 5 The future DC is at the heart of Service Oriented Infrastructure (SOI) Virtualization will change how organisations manage, purchase, deploy, plan and charge. It will also shake up licensing and pricing. Restricted use only 28/09/2009 Slide 6 Virtualization targets (1) • The Network • Known and generally deployed since many years • The Security • Virtual instances of security devices exist since years but slow adoption • Security on VM level arrived only very recently on the market • The Storage • Storage has already been virtualized, but primarily within the scope of individual vendor architectures. • Vendor-agnostic storage virtualisation tools are either not scalable, or not mature Restricted use only 28/09/2009 Slide 7 Virtualization targets (2) • The Servers • Driver: • Traditional model: 1 application = 1 server • 80% to 90% of computing capacity unused at any time (x86) • Uneffective usage of data center space, power and cooling. • Short term advantage of virtualization: • Unlock underutilized capacity • Long term advantage of virtualization : • faster deployments, • reduced downtime, • Easy disaster recovery, • variable usage accounting and usage chargeback, • holistic capacity planning Restricted use only 28/09/2009 Slide 8 End 2 End virtualization in the Datacenter source Cisco Datacenter Virtualization overview: •Front-end Virtualization: •Network: VPN, VLAN, VSS •Security: Virtual Firewall, Virtual Server Load Balancer, Virtual SSL •Server Virtualization •Back-end Virtualization: •VSAN/LSAN, CNA, FCoE Restricted use only •Storage 28/09/2009 Slide 9 The 3 hot issues related to Virtualization • Network management doesn’t stretch to the VM • Network administrators don’t see the whereabouts of the individual VMs • VMs versus Security Policies • Hardware cannot be shared amongst multiple security zones • VMs are Memory Intensive • Consolidating VMs on hardware is often limited by the memory bottleneck on todays server platforms Restricted use only 28/09/2009 Slide 10 VM & Network Management – the Issue With virtualization, VMs have a transparent view of their resources… Restricted use only 28/09/2009 Slide 11 VM & Network Management – the Issue …but its difficult to correlate network and storage connectivity back to virtual machines Restricted use only 28/09/2009 Slide 12 VM & Network Management – Solution Virtual Distributed Switch Aggregated datacenter level virtual networking APP OS APP OS APP OS APP OS APP OS APP OS APP OS APP OS APP OS Simplified setup and change Easy troubleshooting, monitoring and debugging Enables transparent third party management of virtual environments vSwitch vSwitch vNetwork Distributed Switch Cisco Nexus 1000V VMware vSphere™ vSwitch Restricted use only 28/09/2009 Slide 13 VM & Network Management – Solution VN-Link: Virtualizing the Network Domain VN-Link: Virtualizing the Network Domain Policy Based VM Policy Based VM Connectivity Connectivity Server VM VM #1 #1 VM VM #2 #2 VM VM #3 #3 VM VM #4 #4 VM VM VM VM#1 #1 #5 #5 Mobility of Network & Mobility of Network & Security Properties Security Properties Server VM VM VM VM#2 #2 #6 #6 Non-Disruptive Non-Disruptive Operational Model Operational Model VM VM VM VM#3 #3 #7 #7 VM VM VM VM#4 #4 #8 #8 Cisco Nexus 1000V Cisco Nexus 1000V VMW ESX VMW ESX VMW ESX VMW ESX VMs Need To Move • VMotion • DRS • SW Upgrade/Patch • Hardware Failure Restricted use only Virtual Center VN-Link Property Mobility • • • VMotion for the network Ensures VM security Maintains connection state 28/09/2009 Slide 14 VM & Security – the Issues • No security policies can be applied between VMs on same Host •Consequence: •Virtual Machines that are member of different Security Zones (DMZ1, DMZ2, Internal Network…) cannot reside on the same hardware •Hardware Consolidation is limited to 1 security Zone •No security policies can be attached to a VM •Consequence: •Virtual Machines that move from 1 network to another require reconfiguration on security level Restricted use only 28/09/2009 Slide 15 VM & Security – the solution vShield Zones Capabilities Bridge, firewall, or isolate VM zones based on familiar VI containers Monitor allowed and disallowed activity by application-based protocols One-click flow-to-firewall blocks precise network traffic Benefits Well-defined security posture within virtual environment Monitoring and assured policies, even through Vmotion and VM lifecycle events Simple zone-based rules reduces policy errors Restricted use only 28/09/2009 Slide 16 VM & Security – the solution Introducing VMsafe™ Security VM  HIPS  Firewall  IPS/IDS  Anti-Virus Security API ESX  Creates a new, stronger layer of defense – fundamentally changes protection available for VMs running on VMware Infrastructure vs. physical machines  Protect the VM by inspection of virtual components (CPU, Memory, Network and Storage)  Complete integration and awareness of VMotion, Storage VMotion, HA, etc.  Provides an unprecedented level of security – “Virtual is more secure than Real” Restricted use only 28/09/2009 Slide 17 VM & Server Hardware – the Issues • Virtualization & Consolidation require: • Lots of Cooling and Power (1) • The consolidation that result from virtualization leads to high power consumption within the Datacenter Racks • Identical hardware & wiring (2) • To allow VMotion • Same capabilities (I/O) on different host-servers • LAN, SAN, Clustering, Management • Sufficient CPU & Memory (3) • CPU is readily available - so no issue here • Memory is often the bottleneck to increase the consolidation ratio Result of (2) and (3): Large database servers are often kept out of the virtualization scope which prevents uniform deployment and, consequently, operational optimization Restricted use only 28/09/2009 Slide 18 VM & Server Hardware – the solution • Cisco Unified Computing System unifies: • Network virtualization • Storage virtualization • Server virtualization • (1) 1/3th less infrastructure: • Less power/cooling • Less mgmt points • (2) Unified Fabric based • All blades are wired for everything • 320 blades in 1 system (7 racks) • (3) 4x More Memory-scalability compared to similar servers • 384GB per DP Blade Restricted use only 28/09/2009 Slide 19 Agenda • DC Business Requirements • Some New Trends • Virtualization • Unified Fabric and DCE • Cloud Computing • Overview Belgacom Data Centre Solution • Data Centre architectures • Why Belgacom Restricted use only 28/09/2009 Slide 20 Unified fabric increases efficiency and simplifies operations in the DC Mgmt Network Front-End Network Backup Network Unified Fabric Storage Network Back-End Network Use of DCE Restricted use only 28/09/2009 Slide 21 Virtual Lanes and Priority Based Flow Control are important new features in Data Centre Ethernet Priority-Based Flow Control (PFC)  Enables lossless Fabrics for each class of service  PAUSE sent per virtual lane when buffers limit exceeded  Network resources are partitioned between VL’s (E.g. input buffer and output queue)  The switch behavior is negotiable per VL Restricted use only 28/09/2009 Slide 22 Virtual Lanes An example Up to 8 VL’s per physical link Ability to support QoS queues within the lanes VL2 - No Drop Service - Storage DCE CNA VL1 – LAN Service – LAN/IP LAN/IP Gateway VL1 DCE CNA VL2 VL3 DCE CNA Campus Core/ Internet VL3 – Delayed Drop Service - IPC Storage Gateway Restricted use only Storage Area Network 28/09/2009 Slide 23 Data Center 3.0 Infrastructure Portfolio Restricted use only 28/09/2009 Slide 24 Challenges with regards to Unified Fabric Brocade and Cisco are launching products but interoperable, standards-based solutions in this immature market are at least two years away. (source Gartner) Customer Organisations and their governance are not ready. In most enterprises, the server, storage and networking teams are independent, do not fully trust one another and demand exclusive control of their network components. Limited Distribution channels: Brocade has yet to establish credibility as an Ethernet vendor and Cisco’s influence with server administrators is limited. The rest of the vendors in this industry (such as Juniper, Foundry, Nortel, HP, NEC, Hitachi, EMC and IBM) have yet to reveal their converged-core-data-centre product plans. Designing a Unified Fabric, meaning a collapsed LAN & SAN, proves very challenging and requires new architectures in the DC. To simplify designs, it’s best to think in terms of a Fibre Channel overlay on top of an Ethernet network Restricted use only 28/09/2009 Slide 25 Ethernet, a historical perspective Ethernet •The L2 network is a communication pipe •Amorphous pipe, amorphous end device relationships •East-west vs. north-south traffic ratios are undefined •Maximum flexibility Switch ? ? ? ? ? ? Switch Switch ? ? Network designs fill the void •Give shape to device roles, client/server relationships, availability semantics ? ? ? ? ? Client/server relationships Restricted use only ? ? ? 28/09/2009 Slide 26 Ethernet example: a basic IP subnet Default gateways Gx •High availability built with FHRP (HSRP/VRRP) G2 G1 DNS DNS DNS •Purely a server Clients •Most traffic is north-south (through G1 and G2) •Some applications might require east-west traffic Switch Switch C6 Switch C5 The role of broadcast/multicast •Enables service discovery C0 C1 C2 C3 C4 Restricted use only 28/09/2009 Slide 27 Fibre Channel, a historical perspective Fibre Channel •The L2 network embeds most services and provides end device connectivity •Well defined end device relationships (initiators and targets) •Only north-south traffic, eastwest traffic mostly irrelevant •Tailored to fit one function (limited flexibility) I0 I1 I2 I3 I4 T0 T1 T2 DNS Zone FSPF FSPF Zone Switch RSCN Switch DNS DNS FSPF RSCN Switch Zone RSCN I5 Network designs build scale and enhance availability •Everything else is predetermined Client/server relationships Restricted use only I(c) T(s) I(c) 28/09/2009 Slide 28 Ethernet vs. Fibre Channel ? ? ? ? ? Switch Switch T0 T1 T2 ? ? DNS FSPF FSPF Zone Switch Zone RSCN DNS Switch DNS RSCN Switch FSPF ? I0 Switch Zone RSCN I5 ? ? ? ? ? I1 I2 I3 I4 ? ? Bandwidth and services are separate layers, offered by separate entities Bandwidth and services are collapsed, all offered by switches I(c) T(s) I(c) 28/09/2009 Slide 29 ? Restricted use only Problem statement: let’s be pessimists Converged operations? Trust VSAN isolation? Life’s so easy More merge options How can we design a SAN on top of a LAN? Restricted use only 28/09/2009 Slide 30 Problem statement: let’s be pessimists Converged operations? Trust VSAN isolation? Life’s so easy More merge options How can we design a SAN on top of a LAN? Restricted use only 28/09/2009 Slide 31 Agenda • DC Business Requirements • Some New Trends • Virtualization • Unified Fabric and DCE • Cloud Computing • Data Centre architectures • Overview Belgacom Data Centre Solution • Why Belgacom ? Restricted use only 28/09/2009 Slide 32 Virtualization Enables New Delivery Models Changes that will make virtualization critical to most enterprises during the next few years: • Processor evolution • Power outpaced requirements of most applications • Cost decreased  overhead virtualization not an issue anymore • High TCO per Server remains: • Space, power, cooling • Installation, integration and administration -> independent of whether a resource is 10% or 90% used. • Lack of required resources: • People as well as space & power become scarce to individual organisations • Unpredictable workloads: • Due to Web access, workload evolved from relatively predictable to spiky • Enterprises forced to overprovision.  These evolutions opened the way to new delivery models Restricted use only 28/09/2009 Slide 33 What Is Cloud Computing? Gartner's Definition Gartner defines "cloud computing" as a style of computing where massively scalable IT-related capabilities are provided "as a service" to multiple customers using Internet technologies. It would be easy to add a refining statement to this definition, such as "where the service consumers need only care about what the service does for them, not how it's implemented." Related thoughts: • • • • Restricted use only Private clouds vs the public cloud Buying products vs contracting a service Elasticity vs Scale - Pay-as-you-Use vs SoD Global class — not just Enterprise class 28/09/2009 Slide 34 The Cloud: Private or Public? Private Clouds will de-couple enterprise apps from IT infrastructure Inter-Cloud will emerge, based on Cloud Standards and Interoperability Private Private Private - AUTOMATION - DYNAMIC CAPACITY MANAGEMENT Server Storage Network Public Cloud Public Cloud VIRTUALIZATION Private Private Private Private “Cloud solutions must be implemented and managed across multiple geographies, service providers, networks, users and enterprises.” Yankee Group Source: Cisco IBSG; Restricted use only Slide 35 Cloud computing: 5 Styles are emerging SaaS Browser Cloud Desktop Browser Personal application Personal application Client hardware Server application Server/ storage Salesforce Restricted use Source: only Gartner Cloud Desktop Infrastructure Cloud Server Cloud Infrastructure Personal application Client hardware Server application Server/ storage GoogleDocs, Web search Browser/ Personal application Browser/ Personal application Client hardware Server application Server/ storage Desktone, (VDI "in the cloud") Client hardware Server application Server/ storage Client hardware Server application Server/ storage 28/09/2009 Slide 36 Cloud Computing on the Hype Cycle visibility Green IT Social Computing Platforms Microblogging 3-D Printing Video Telepresence Cloud Computing Surface Computers Augmented Reality Mobile Robots Basic Web Services Behavioral Economics Public Virtual Worlds Web 2.0 SOA Service-Oriented Business Applications Virtual Assistants RFID (Case/Pallet) Context Delivery Architecture Erasable Paper Printing Systems Tablet PC Electronic Paper Wikis Social Network Analysis Idea Management Corporate Blogging Location-Aware Applications Solid-State Drives As of July 2008 Technology Trigger Peak of Inflated Expectations Trough of Disillusionment Slope of Enlightenment Plateau of Productivity time Years to mainstream adoption: less than 2 years Restricted use only 2 to 5 years 5 to 10 years more than 10 years obsolete before plateau 28/09/2009 Slide 37 Source: Gartner The network: the final frontier for cloud computing Restricted use only 28/09/2009 Slide 38 Agenda • Data Centre Business Requirements • Some New Trends • Data Centre architectures • Overview Belgacom Data Centre Solution • Why Belgacom Restricted use only 28/09/2009 Slide 39 Answers to the questions below will result in a preferred architecture 1. Business Continuity plan • How do we design the IT infrastructure in order to support the the BC-plan of the customer’s organisation in case of primary site disaster 2. RPO (Recovery Point Objective) • In case a certain incident causes an application service to go down, how much data/transaction loss can the customer’s organisation accept after restart 3. RTO (Recovery Time Objective) • In case a certain incident causes an application service to go down, how long can the customer’s organisation accept the absence of this service 4. Service availability • How much % of the time, a certain service needs to be available 5. Application service protection • Which protection mechanisms are already available on application level? Restricted use only 28/09/2009 Slide 40 Some guidelines (1) RPO <24u • In case of low RPO (up to zero) we need to build-in replication-mechanisms to ensure all data to be present on 2 site • Synchronous replication (RPO zero) or asynchronous replication (RPO < 24u) can both be investigated based on the inter-site connectivity possibilities RPO >24u • In case of high RPO, we can limit data-security to a daily backup to offline media • Replication is not mandatory Restricted use only 28/09/2009 Slide 41 Some guidelines (2) Low RTO • A low RTO implicates the need for fast restore-times. This can be realised by implementing backup-to-disk solutions (eg SnapShotting) • Virtualisation solutions (eg. VMware) also deliver many advantages for fast restore operations (eg. HA, SRM) High RTO • In case of limited RTO requirements, a backup to offline media (eg Tape) might be sufficient We always advice to implement an automated remote backup-restore architecture that include a second site without the need for manual handlings as these typically are error-prone Restricted use only 28/09/2009 Slide 42 Some Guidelines (3) Develop DR procedures, (part of Business Continuity Management BCM) Test, test, test,… High speed MAN design (CWDM, DWDM, LRE,…) Synchronous mirroring over SAN extensions up to 125 km. Restricted use only 28/09/2009 Slide 43 Active-Active approach • The customer has production servers on 2 sites. • Data is replicated synchronously (either on application or on storage level) • Clusters can be stretched over 2 locations • Both DCs act as DR site of the other • Secured IAS (Internet Access Street) is stretched over 2 locations • LAN is stretched over 2 locations • Backup is crossed • Only switch over for the defect layer Restricted use only 28/09/2009 Slide 44 Datacenter Blueprint (1) LAN Si LAN Si Si Si XWDM BU server BU server XWDM Re m SAN o te t ap eB SAN ac ku p Synchronous mirroring Backup to Disk Restricted use only 28/09/2009 Slide 45 Active-Active-DR approach • The customer has production servers on 2 sites. • Data is replicated synchronously (either on application or on storage level) • Clusters can be stretched over 2 locations • Backup is crossed • Both DCs act as DR site of the other • Asynchronous replication to 3rd DC outside region • Third site has (optional) standby server capacity that can serve as (degraded) production environment in case of primary/secondary site disaster • Secured IAS (Internet Access Street) is stretched over 3 locations • LAN is stretched over 3 locations • Switch between DC1 and DC2 for a defect layer • Switch to DC3 for an regional disaster in DC1 and DC2 Restricted use only 28/09/2009 Slide 46 Datacenter Blueprint (2) WAN LAN Si LAN Si Si Si XWDM BU server BU server XWDM SAN SAN LAN Si Si Synchronous mirroring As c yn on hr ou s g in or irr m SAN Restricted use only 28/09/2009 Slide 47 Active-Passive approach • The customer has production servers on 1 site • Data is replicated (Synchronous or asynchronously) • Second site has (optional) standby server capacity that can serve as (degraded) production environment in case of primary site disaster • DC2 acts as DR site of DC1 • Secured IAS (Internet Access Street) is stretched over 2 locations • LAN is stretched over 2 locations • Backup is unidirectional (site 1 to site 2) Restricted use only 28/09/2009 Slide 48 Agenda • DC Business Requirements • Some New Trends • Data Centre architectures • Overview Belgacom DC Solution • Systems • Connectivity (LAN & SAN) • Storage & Backup • Access • Management • Physical environment • Why Belgacom ? Restricted use only 28/09/2009 Slide 49 Overview of our Data Centre Solution Security Back UP S A N Storage External Access APPS Systems L A N Internal Access Management Remote Operations Business Continuity & Disaster Recovery Physical environment (Cabling, Cooling, Power…) Restricted use only 28/09/2009 Slide 50 DC Networking Security Back UP S A N Storage External Access APPS Systems L A N Internal Access Management Remote Operations Business Continuity & Disaster Recovery Physical environment (Cabling, Cooling, Power…) Restricted use only 28/09/2009 Slide 51 DC design has to take new connectivity options into account In the DC this has to be robust, flexible, scalable and Secure. • Local Area Network (LAN) based on IP • IP is unreliable (drops packets, out of order arrival, corruption,…) • High Speed connectivity (10 Gigabit uplinks) • Integrated blades in high end solutions: load balancer, firewall, IDS, … • Multi chassis viewed as one: virtual chassis • Close integration with BAPAS Restricted use only 28/09/2009 Slide 52 Belgacom Application Acceleration Services Get ready for the Real Time Enterprise  How do you optimise if you don’t know where the problem is?  How do you design and build application fluent networks?  How do you improve the application experience in remote offices ?  How do you optimise your web-based applications ?  How do you guarantee application quality when security threats are not under control ?  How do you get End-To-End Performance SLA visibility about the delivered Application Performance ? Restricted use only 28/09/2009 Slide 53 WAN Optimization – Why ? Market trends Globalization • Introducing WAN latency • Management challenges Result: costly branch offices, poor performance Centralization & Consolidation • Servers and applications • Increased WAN bandwidth consumption Result: slow response times, low end-user satisfaction New application landscape • Additional bandwith • Prioritize business critical applications Result: additional investments, limited improvements Restricted use only 28/09/2009 Slide 54 WAN Optimization Concept • Optimize bandwidth usage Objective: Increase available bandwith without upgrading  Compression & data redundancy elimination  Increase available bandwidth without upgrading • • TCP Optimization Application Acceleration Objective: Reduction in round trips Objective: Overcome application protocol ineffeciency  Common applications include : - Windows Mail (MAPI) - Web, Intranet (HTTP(S)) - File Sharing (CIFS, NFS) - Transfer (FTP) -… Restricted use only 28/09/2009 Slide 55 Wan Optimisation – Riverbed based Steelhead Appliances • • • xx50-series new generation On-demand upgrade capability Best-of-breed solution, 85% of Telindus installed base Steelhead Mobile (SoftWOC) - Concurrent license model - Alternative for micro-branches, upselling opportunity for GPRS/UMTS mobile users Steelhead Appliance Steelhead Mobile Controller BRANCH OFFICE WAN DATA CENTER MOBILE WORKERS (GPRS – UMTS) HOME USERS (xDSL) Restricted use only 28/09/2009 Slide 56 Wan Optimisation – Juniper based Juniper WXC appliances • New generation appliances - 1800-, 2600- and 3400 models • WXC Client (SoftWOC) - Windows OS compatible - Integration with Juniper SA SSL VPN Solutions •WXC Central Management System - Central reporting and configuration management Small Branch Office Intranet/Extranet Data Center WX client Mobile Users & Telecommuters WX client Partners and Contractors WX client Restricted use only 28/09/2009 Slide 57 DC design has to take new connectivity options into account In the DC this has to be robust, flexible, scalable and Secure. Local Area Network (LAN) based on IP • IP is unreliable (drops packets, out of order arrival, corruption,…) • High Speed connectivity (10 Gigabit uplinks) • Integrated blades in high end solutions: load balancer, firewall, IDS, … • Multi chassis viewed as one: virtual chassis • Close integration with TAPAS Storage Area Network (SAN) based on FC (or iSCSI) • FC is reliable, based on the SCSI command set • High speed connectivity to storage (1, 2, 4, 8 Gbit/s) • Consolidation using virtual SANs Restricted use only 28/09/2009 Slide 58 Consolidation of SAN islands by using Virtual SANs • Virtual fabrics support the need to consolidate numerous SAN islands • Fabrics can be migrated from physical to virtual implementations • New fabrics are provisioned through switch commands, not physical adds, moves and changes • Fabrics are provisioned as a service Restricted use only 28/09/2009 Slide 59 DC design has to take new connectivity options into account In the DC this has to be robust, flexible, scalable and Secure. Local Area Network (LAN) based on IP • IP is unreliable (drops packets, out of order arrival, corruption,…) • High Speed connectivity (10 Gigabit uplinks) • Integrated blades in high end solutions: load balancer, firewall, IDS, … • Multi chassis viewed as one: virtual chassis • Close integration with TAPAS Storage Area Network (SAN) based on FC (or iSCSI) • FC is reliable, based on the SCSI command set • High speed connectivity to storage (1, 2, 4, 8 Gbit/s) • Consolidation using virtual SANs Infiniband • For grid computing through RDMA Upcoming new connectivity method is Data Centre Ethernet (DCE) with support for Fibre Channel over Ethernet (FCoE) Restricted use only 28/09/2009 Slide 60 Management of the DC infrastructure : FCAPS • Fault Management Fixing what is broken • Configuration & Change Management Controlling the operational parameters of something, so it works the way you want • Accounting Management Knowing who is using how much of what, and maybe billing them for it • Performance Management Making sure it all works acceptably quickly • Security Management Controlling who can do what Restricted use only 28/09/2009 Slide 61 FCAPS: Belgacom Solutions Our solutions cover the FCAPS model with - Tools from best of breed vendors - Our expert services to assess your needs & implement the right solution(s) Tools : Fault Management : Nimsoft , HP Open View Configuration & Change Management : HP , Cisco Works Accounting Management : Fluke Networks Performance Management Nimsoft , Fluke Networks , Compuware Security Management SIEM ( Security Information & Event Management ) : RSA, Juniper, Arcsight Restricted use only 28/09/2009 Slide 62 Agenda • Data Centre Business Requirements • Some New Trends • Data Centre architecture • Overview Belgacom Data Centre Solution • Why Belgacom Restricted use only 28/09/2009 Slide 63 Why Belgacom We have a network practice , a systems/storage practice and a security practice We have an end to end strategy /portfolio : architectural view balanced with specialised knowledge We have 4 state of the art DC’s used for housing/hosting We offer the choice between hosted/shared solutions and dedicated solutions We offer our solutions at different engagement levels ( from reselling to outsourcing ) Restricted use only 28/09/2009 Slide 64 Questions? Restricted use only 28/09/2009 Slide 65 Thank you! Restricted use only 28/09/2009 Slide 66