Let’s take a look at the root causes of this maintenance predicament and some other recent cost trends you might be seeing in your data center. Today, the data center makes up 44% of overall IT spend. The traditional model of data center investment was to build in silos—often with dedicated resources for lines of business—and to design them for peak environments. Unfortunately, this model had some inherent inefficiencies because resources couldn’t be shared and, for the majority of the time, applications weren’t maximizing the capability of the server platform, which drove up costs via underutilization of resources A few other things jump out. First, more than 50% of the cost is the combination of people and software, with energy and facilities as the next largest chunk (contributing no new value-add to the IT department—it’s required to simply keep things running). One may come to the conclusion that infrastructure doesn’t matter that much, but when you actually analyze the root cause of the spiraling costs in people and software, one quickly realizes that they are related to the infrastructure – and the software to manage it. For instance, if we double click on servers, while the cost of hardware has been flat or declining for the past 15 years, the cost of managing and operating servers has been growing steadily and now represents 2/3 of the total spend associated with servers. The problem has been exacerbated by virtualization because higher server utilization has come at the price of increasing complexity that shows in the rising Opex costs. Before we do that let’s look at some of the trends going on within CapEx and OpEx at the bar chart. Note that server and power and cooling costs seem to be remaining steady. Even with the inherent efficiencies of server virtualization—companies continue to need to invest in servers because of the growth in project demand they continue to experience. But where increase pain is really being felt is in the area of management and administration costs associated with your virtual servers. This stems from the problem of the traditional approach to management in so many IT organizations. Because a wide variety of traditional servers, software systems, storage have been utilized, and implemented in silos—most system management functions are complicated—designed to try to integrate solutions that were never designed to work together. Virtualization, of course, means there is increasing movement of virtual machines—and the need to ensure key information and adequate security moves with them. Without a simpler approach to management, administration costs will continue to escalate.
Virtual Switching: Server virtualization allows multiple OS images to transparently share the same physical server and I/O devices. As a consequence, it introduces the need to support local switching between different virtual machines within the same server. The most common way to solve this problem has been to implement a software based switch inside the physical host, inside the hypervisor Each VM gets a vNIC from the hypervisor and the vNIC is associated with a virtual port (or vETH) on the vSwitch. Traffic between VMs on the same host is switched locally within the host but traffic to a VM on a different host needs to still go thru the external physical switch. Palo is a Cisco ASIC processor. As such it represents a sustainable competitive edge over any thing a rival might produce. And it provides unrivaled flexibility in terms of supporting multiple network protocols while maintaining consistently high levels of performance Drivers currenty available for Vmware and KVM, soon for Citrix XEN, next Open XEN, later Hyper-V
The Cisco UCS B22 M3 Blade Server is an entry-level, efficient and affordable blade server for IT/web infrastructure and HPC workloads. Building on the success of the UCS B-Series blade servers, the UCS B22 M3 server further extends the highly differentiated capabilities of the Cisco Unified Computing System. It uses Intel's latest Xeon E5-2400 processor product family, 12 DIMM slots and 2 x 10 GbE LOM slots to deliver new levels of entry-level performance and efficiency. The Cisco UCS B22 M3 Blade Server balances cost, simplicity, performance, and density for general business applications, entry-level virtualization, network infrastructure, HPC, collaboration and web serving workloads. The server is a half-width, 2-socket blade server with cost-optimized, throughput and scalability. Building on the success of the UCS B-Series blade servers, the UCS B200 M3 server extends the capabilities of the Cisco Unified Computing System. It uses Intel's latest Xeon E5-2400 Xeon product family, 12 DIMM slots and 2 x 10 GbE LOM (LAN on Motherboard) slots to deliver new levels of affordable performance and efficiency. Add HPC
As you can from the left slide is an example of performing a In-Memory consistency check and manual procedure within the technical operations guide for SAP. *Imagine now, those steps are represented as process flows steps that can be orchestrated again all of the touch points I showed in the last slide.
A couple of key features that SAP ITPA delivers out the box is the ability to consume alerts from CCMS or Solution Manager and automate the application incident response process to deliver higher IT operations efficiencies including reducing mean time to repair (MTTR), and to increase IT personnel effectiveness by automating repetitive tasks associated with IT operations processes. One can be quickly model resolution content or use predefined corrective processes based on SAP best practices for issue resolution. Advanced IT Process Reporting delivers auditing and reporting capabilities that are essential for managing IT operational processes and SAP best practices, and to fully report on how the processes are executed in line with established policies and service levels. It has a Visual Configuration Environment that provides a rich drag-and-drop environment for creating process flows. It includes over 400+ out-of-the-box flow activities and processes that can be easily reused across your SAP applications, including SAP HANA. Its modeling language reduces the need for scripting and enables workflow creation that is optimized for modeling best practices content for implementing and managing IT operations processes. It also has a visually guided mode that has the ability to incorporate human approvals, governance check-points and guided IT operational processes. All of this capability is deployed with no agents, a small footprint and integrates with existing monitoring tools, thus allowing IT to leverage existing investments.
Let’s examine the solution. SAP IT Process Automation by Cisco is a platform with the ability to leverage, model and orchestrate IT operational flows that implement best practices across all management disciplines. The core platform is comprised of the client, database and server components and support for non-SAP operational process activities. It c an interact with all types of infrastructure elements such as applications, databases and network through adapters. The platform contains automation options that contain over 300 predefined automation packs for quickly adopting and implementing best practices for technical operations. The Option for incident response allows you to quickly answer the question: “I have an alert, what do I do?” Automated correlation of SAP alerts and non-SAP events helps to quickly identify and isolate incidents. SAP troubleshooting and non-SAP troubleshooting content can be combined to quickly identify E2E incident resolution actions though a generated Incident Response Automation Summary Report. And diagnostic tools can be directed for further analysis and problem resolution. The Options for IT task automation provides the ability to answer the question: “If I can repeat it, how can I quickly automate it?” SAP and non-SAP tasks can be bridged for automating functional maintenance and landscape optimizations. Common administrative checks and corrective actions across SAP and non-SAP landscapes can be automated. The Option for SAP NetWeaver BW and BW Accelerator automates many manual technical operations for Business Warehouse functions for significant time savings. The following automation capabilities are currently in ramp-up: The Option for SAP Solution Manager provides support for E2E alerting. The Option for SAP HANA delivers support for automating HANA manual technical. And integration with SAP Landscape Virtualization Management delivers the SAP standard for system copy and refresh with support for quickly orchestrating and modeling custom pre- and post-operations.
Five years ago, RISC/UNIX was the right choice but today, there are challenges with these platforms Unable to Meet Business Flexibility/Agility Needs Complex system management Inflexible environment from both a server and cabling perspective Very few build private or public clouds on RISC – inflexible and poor economic model Complexity and high costs makes it difficult to upgrade/migrate applications to another RISC/UNIX platform Transitioning with corporate mergers/acquisitions is difficult because open, standard platforms are not being used New projects coming online and user growth leads to increased need for greater scalability, capacity and better utilization (should this be targeted more towards the need for private clouds and virtualization with appls moving around the data center) Skills shortages – Few graduates come out of school with RISC/UNIX experience – primarily Linux/x86 and Windows/x86 – so the current population with RISC/UNIX background is aging and very expensive. High Costs while IT department budgets are shrinking High environmental, support, and maintenance costs with current RISC/UNIX infrastructure Extremely high capital costs to upgrade to another RISC/UNIX platform and high operational costs with greater environmental and maintenance costs. High software licensing costs High levels of complexity leading to high admin costs Too many different platforms with different management paradigms High cost of trained staff/talent and difficulty in finding them Poor price-performance Current installations have insufficient performance, and are not keeping pace with the market, i.e. the latest Intel Xeon processor performance capabilities. Upgrading to the latest RISC/UNIX platform is VERY expensive Out-of-band high costs for small modular capacity increase Environmental Factors Large footprint High power and cooling requirements RISC/UNIX server sprawl leading to need/desire to consolidate Customers Uneasy About Future Plans of Oracle for SPARC/Solaris Not a hardware company - Uncertain roadmap Already has a Linux offering and may be positioning x86 as a SPARC replacement Long release schedules Vendor lock-in Support policies that require bundling: difficult to do business with Many long term Sun partners no longer on-hand as trusted advisors BUT, you are Hesitant to migrate mission-critical apps to new OS/Platform Possible multi-year pain of migration Lack of experience/trust of Services/OS/Platform vendor Unknown after-effects in operational processes Presumed lack of staff knowledge and the costs of retraining (RASS) Need 4/5-9’s availability, not sure x86 platforms can deliver
97% unit volumes are x86 architecture-based 89% are Intel Xeon Processors and only 60% of the server revenues Only 2% unit volumes are RISC-based, but make up for 19% of server revenues! Businesses are paying a high premium for their RISC-based systems. Gravitational effect on software developers and technology innovators The mission-critical application market developing first for Intel Xeon processors Leading enterprise applications optimized for Intel Xeon processors Economic efficiencies are forcing many applications vendors to focus development Intel Xeon Processors. Enterprise applications on Intel Xeon include - Oracle - SAP -
Cisco works with you to Provide a proven migration process from RISC-based Unix environments, to Cisco’s industry-leading Intel Xeon processor-based Cisco Unified Computing System platform. Comprehensive cost-effective approach to migration initiatives using best-of-class migration methodologies, in-depth analysis tools, a robust planning process, and design and implementation. Designed to provide a flexible collaboration depending on the complexity and criticality of the applications to be migrated. Cisco leverages strong relationships with many ISV vendors including Oracle and SAP and have performed numerous migrations with most name brand applications, as well as highly customized applications. Cisco’s RISC/UNIX Migration Services include Introductory Workshop and Proof of Concept Service where we work with our customers to explore current applications and migration candidates including off-the-shelf and custom applications. Migration complexity is assessed and a joint next-steps document is created. Cisco RISC/UNIX Migration Service works with you to develop a migration strategy blueprint and high-level migration roadmap, detailed migration planning is performed evaluating application service level requirements, a RISC/UNIX to Cisco UCS transfer map for workloads, and risk analysis. A detailed migration plan including low-level design and functional and non-functional test plans is produced. The creation of a gold image, the physical migration, and system management framework integration. Onsite operational support including mentoring and knowledge transfer to your staff.
Transforme, automatice y acelere su negocio con Cisco y SAP
Transforme, automatice y acelere sunegocio con Cisco y SAPCarlos CamposCisco UCS Consulting Systems Engineer – Latinamerica