RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

Cisco UCS Unified Computing System (Cisco UCCX)

Product
The name of the base system (platform): Cisco Data Center Business Advantage
Developers: Cisco Systems
Date of the premiere of the system: March 2010
Last Release Date: September 2018
Branches: Telecommunications and communications
Technology: Virtualization,  Data Centers - Data Center Technologies

Content

Main article: Virtualization. Classification and applications

2022: Cisco UCS - SAP Streaming Temperature System Hardware Platform

On February 9, 2022, SAP announced that, together with partners Facemetric, Cisco Systems, Konica Minolta (MOBOTIX) and Skybuffer, it has developed a system for automatically measuring face recognition temperature. The solution integrates with the corporate HR system, mail and instant messengers. The collection of biometric data and user interaction takes place through a chat bot in Telegram, various scenarios are provided for informing employees and HR specialists, as well as the formation of statistical reports. The project was implemented by SAP Services CIS experts.

According to the company, the information is processed and sent to SAP BTP, where the final analysis of the data takes place and notification scenarios are created. The hardware platform is Cisco UCS. More details here.

2018: First Cisco AI and Machine Learning Servers

In September 2018, Cisco introduced the first server in its range, specially designed for processing artificial intelligence and machine learning processes. The American manufacturer updated the Cisco UCS family of unified computing systems by releasing the Cisco UCS C480 ML model.

The new product, clad in a 4U chassis, received the latest generation of Intel Xeon processors and eight Nvidia Tesla V100-32G graphics accelerators with NVLink support at the time of announcement.

Cisco UCS C480 ML

The most advanced configuration includes a pair of Xeon chips, 128 GB of DDR4 RAM, 24 SATA hard drives or SSDs, six NVMe SSDs, and four x100G converged network adapters. The system is capable of working with various Cisco servers and HyperFlex systems, reports NetworkWorld.

According to Cisco, the announced server can cover all stages of AI and machine learning applications. In particular, we are talking about collecting and analyzing data, preparing and training the system in the data center, as well as generating logical output in real time. The system will help customers extract more intelligence from their data and use it to make quick and productive decisions, the manufacturer says.

With new initiatives DevNet AI Developer Center and DevNet Ecosystem Exchange, Cisco provides data professionals and developers with tools and resources to build next-generation smart applications.

Cisco UCS C480 ML is designed to work with container applications and multi-cloud computing models for AI systems with datasets stored on multiple servers and services.[1]

2017: Cisco Released Fifth Generation Servers

On July 17, 2017, Cisco introduced fifth-generation servers for unified computing based on Intel Xeon Scalable processors, UCS M5 servers.

The memory of devices is doubled compared to previous generation systems and they can handle more tasks.

Cisco UCS B200 M5 Blade Server, (2017)

The Cisco UCS B200 M5 blade server is the base data center unit. With half the width of the module, the B200 provides the performance, versatility, and density required for tiered and distributed applications. Two GPUs are supported.

Cisco UCS B480 M5 Blade Server. Performance, versatility, and density indicators make the server suitable for a variety of tasks, from mission-critical, memory-intensive enterprise applications to virtualized tasks using distributed databases.

The Cisco UCS C220 M5 Rack Server is a generic dual-processor server for enterprise infrastructures and applications. The server has high performance and efficiency, designed for various tasks, including virtualization, collaboration and applications without virtualization (bare-metal applications).

Cisco UCS C240 M5 Rack Server. An enterprise-class server optimized for storage and I/O, the main application is big data analytics, software-defined storage, and applications without virtualization.

Cisco UCS C480 M5 Rack Server. C480 is based on a modular architecture and provides vertical scalability for in-memory databases, big data analytics, virtualization, VDI, and non-virtualized applications. Six GPUs are supported and 32 drives are available.

ON Workload Optimization Manager addresses the challenges of coordinating application needs and infrastructure capabilities in local and multi-cloud configurations, including through analytics technologies. Integration with Cisco UCS components (chassis, blades, I/O modules switchboards , fabric interconnect) enables customers to reduce costs by optimizing computing resources for virtualized and cloud environments. The latest version Cisco UCS Director 6.5 is integrated with the Workload Optimization Manager, which allows you to automatically create virtual machines and physical server configurations using UCS Director, after which the Workload Optimization Manager moves resources to ensure application performance and cost-effectiveness.

2016: UCS S-Series Servers

On November 2, 2016, Cisco introduced the UCS S-series server line as part of the Cisco Unified Computing System (UCS) product portfolio. Servers are optimized for storage and are designed for applications that operate in large volumes: Big Data, streaming and collaboration Deploy software-defined storage, object storage, and data protection solutions.

The UCS S-Series is a solution for working with fast-growing volumes of unstructured data generated on the Internet of Things and mobile networks, with applications for video processing, collaboration and analytics. The UCS S-Series enables enterprises to quickly access and process data for real-time analytic information.


UCS S3260 Storage Server

The UCS S3260 Storage Server features versatility, scalability, and performance, all you need to leverage your data for business intelligence. Using UCS Manager modular architecture and automation, the S3260 helps reduce total cost of ownership by more than 50% over the public cloud, enabling you to configure your infrastructure to perform your specific tasks efficiently.

S3260 Storage Server, (2016)
Rear view of the S3260 Storage Server, (2016)


Using UCS S3260: Reduces investment and transaction costs (CapEx) by 34 per cent; Reduces management costs by 80% Reduces cable infrastructure costs by 70% occupies a smaller area by 60%; consumes less power by 59%.

The UCS S3260 has the flexibility to select configurations:

  • storage capacity up to 600 TB per system,
  • seamless scaling by UCS Manager to petabyte volumes
  • Efficient caching and unified I/O for any type of storage connection.

With these resources, the vendor believes, organizations can move from passive data in storage and cloud to dynamic applications on a cloud-scale platform to enable real-time data analysis and decision-making.

2015: UCS servers have been tested by Oracle PeopleSoft

February 24, 2015 it became known about the performance record of Cisco UCS servers when executing the Oracle PeopleSoft Human Resources Management Systems (HRMS) 9.1 FP2 Self-Service (online transaction processing, OLTP) for self-service and batch payroll processing.

The self-service test demonstrated the record performance and linear scalability of Cisco UCS servers. The results show to what extent it is possible to scale existing Cisco? Unified? Computing systems SystemTM as the number of users in the HR system grows.

Oracle PeopleSoft HRMS 9.1 FP2 Self-Service Test Results, 2015

Oracle PeopleSoft HRMS 9.1 FP2 Self-Service and Payroll Batch Extra-Large Data Model, 2015

The Oracle PeopleSoft HRMS test is designed to obtain verifiable performance indicators when performing various tasks Human Resources Management in the PeopleSoft HRMS system.

2014: UCS Mini

In September 2014, Cisco introduced a new UCS Mini server as part of the Unified Computing System portfolio.

UCS Mini

The system is designed for small offices and peripheral enterprise configurations. A compact, easy-to-mount, easy-to-use, all-in-one solution.

UCS Mini view (Russian-language subtitles)

UCS Mini Interaction Diagram in Cisco UCS M3 Blade Server and Cisco UCS C240 M3 Rack Server

2013: Cisco UCS for Big Data

Because of the Cisco UCS unified computing environment's balance between performance, manageability, and scalability, UCS stands apart from competing solutions. Let's name the three main factors that make Cisco UCS a platform for Big Data analytics.

Factor one: unification of server management of different types and independence of management functions from the dimensions of hardware systems

Cisco UCS creates a single management point for the entire infrastructure (blade architecture for enterprise applications and rack architecture for Big Data analytics), including diagnostic, monitoring, and disaster warning features. This allows the customer to monitor system performance proactively and reduce running costs.

In other words, the customer can manage the Cisco UCS rack servers in the same way as the UCS blade servers and freely transfer jobs from the rack servers to and from the blade servers. This simplifies management tasks and eliminates the need to install new stand-alone control systems in the data center. The Cisco Unified Fabric unified fabric with a single control channel and Cisco Unified Management Solution (Cisco Unified Management), including a UCS Manager system with service profiles, provides the independence of management functions from hardware systems.

Factor Two: Fast and Easy Implementation with UCS Service Profiles

The Cisco UCS computing architecture takes service profiles into account and abstracts from the hardware device state. Cisco UCS service profiles are a powerful tool for optimizing the management of modern data centers. They allow you to quickly provision resources for servers and their associated network connections and storage. In addition, these profiles support consistent operation of all elements of the unified computing environment.

The service profile contains all information about application, job, and hardware settings. When installing this solution, the customer determines the hardware profile, writes it to the template, and then uses this template to configure each node in the cluster. As a result, it takes a few minutes to set up multiple nodes. In the process of scaling, the customer does not need to configure each new node separately: you can use the same templates to configure new nodes, which gives the customer flexibility and high speed to enter the market.

Factor Three: Excellent Scalability with Cisco UCS Central (Up to 10,000 Nodes)

When a company builds up its Cisco UCS infrastructure, the issue of robust infrastructure management with global policies is of particular importance. Cisco UCS Central can manage multiple instances (or domains) of Cisco UCS in global distributed data centers. As a result, IT can easily manage computing, network resources, and storage in a unified manner.

In a single UCS domain, Cisco UCS Manager can host up to 160 servers (10 racks of 16 C240 M3 UCS servers each). Cisco UCS Central supports global policy and pool definitions, allowing you to flexibly allocate resources across distributed data centers. This allows an administrator to reuse one-time jobs in their infrastructure and monitor up to 10,000 sites. This capability is especially important for large Big Data systems.

2012: Platform Development

The third generation UCS unified computing platform Cisco includes processors of the new Intel Xeon E5-2600 family servers and different models with different sizes. This platform supports the industry's highest server density and efficiency. They now outperform UCS-based servers eight times in memory and four times in I/O performance. Cisco UCS Manager enables IT administrators to manage blade servers and rack servers as one. This solution extends management to include thousands of servers installed in data centers worldwide.

According to customer feedback, Cisco UCS provides the following benefits:

  • 30% reduction in infrastructure costs,
  • 90% reduction of implementation time,
  • 40% improvement in application performance and
  • 60% lower power and cooling costs.

Unified management

Advanced integration with Cisco? Unified? Fabric enables Cisco? UCS Manager to unify blade and rack server management in a single domain for the first time in the industry. As a result, server connectivity costs are reduced, resource allocation flexibility is increased, and IT administrators are able to increase data center capacity seamlessly.

In the second half of 2012, Cisco intends to announce the technology to support large-scale UCS implementations with centralized management of many domains, including thousands of servers installed in one data center or distributed throughout the planet. This will be the industry's first cloud automation and coordination solution of its kind.

Integrated networking and virtualization

The built-in 2204XP I/O module provides optional support for 80Gb/s and 160Gb/s speeds on each chassis to handle peak traffic spikes. This module also supports the port channeling function, which allows balancing the load on all ports and improving the efficiency and reliability of communication through optimal use of channels and bandwidth.

Cisco UCS 6296UP Fabric Interconnect doubles the switching capacity in a UCS environment (from 960 Gb/s to 1.92 Tbit/s) and reduces overall latency by 40 percent. As a result, system performance rises to the industry's best level. This solution supports the flexibility of a large-scale infrastructure with unified ports, and improves energy efficiency by reducing each port's power consumption by 36 percent.

The Cisco UCS 6200 solution, combined with the Cisco Nexus ® Fabric Extender solution, further increases the benefits of Cisco UCS Manager in large-scale UCS implementations based on blade servers and rack servers .

Next Generation Cisco UCS M3 Servers

Cisco has added the Cisco UCS product family with another blade server and two rack server models with the Intel Xeon processor E5-2600.

2011

Performance indicators

In annual terms, UCS profit Cisco increased to 650 million (dollars USA for comparison: at the end of the first quarter of fiscal year 2011, it amounted to about $500 million). Thus, it can be said that Cisco's technology strategy for the data center virtualizations cloud computing is delivering more and more impressive results, and you can expect to continue this trend, especially since Cisco has been the largest provider of architectural solutions for almost two decades. DPC

At the end of Q2 2011, the number of unique Cisco UCS customers came close to 4,000. For comparison: in the previous quarter there were 2.800 such customers. Every quarter, Cisco UCS has new customers, helped in no small part by the fact that Cisco's unified computing environment significantly reduces total cost of ownership. It is noteworthy that Cisco UCS distribution goes about the same in all market segments and industries. The fastest spread of this technology is in the USA and Canada, but also in Europe, as well as in the Asia-Pacific region and in countries with fast-growing markets, Cisco UCS sales are growing quite rapidly.

Cisco UCS Q2 2011 Order Structure by Market Segment:

  • Small and medium business = 40%
  • Large enterprises = 28%
  • Government institutions = 22%
  • Carriers = 10%

Cisco UCS Q2 2011 Order Structure by Geographic Region:

  • USA and Canada = 65%
  • EMEA (Europe, Middle East, Africa) = 17%
  • Asia Pacific and Emerging Markets = 18%

Cisco UCS is being implemented in both existing and new data centers. The growth of UCS sales in the data center market is determined by a number of factors - from reducing costs, in turn, associated with the consolidation of infrastructure in order to save energy and workspace, to strategic investments in accelerating the slowed down virtualization process, increasing the level of automation of information technology and preparing for the implementation of cloud computing. At the same time, part of the market is increasingly interested in rapidly deployable pre-optimized turnkey solutions, and Cisco provides such solutions with the help of its strategic partners: NetApp, EMC, Citrix, VMware and several other companies. These joint solutions are also successfully distributed in the market, providing Cisco with additional revenue in addition to the revenue from the sales of Cisco UCS solutions themselves.

As a result, by the end of the 2nd quarter of FY 2011, compared to the same period of FY 2010:

  • Cisco revenue from data center solutions increased 59 percent;
  • revenue from Nexus 5000 solutions increased by 56 percent;
  • Revenue from Nexus 2000 solutions increased by more than 150 percent;
  • Revenue from Nexus 7000 solutions increased by almost 100 percent and reached an annualized level of $1 billion;
  • Cisco UCS technology revenue rose more than 700 percent to above $650 million (year-on-year).

Market interest in convergence and cloud computing has reached a critical mass, and from talking about these technologies, the market is moving to their practical implementation. According to Cisco, its data center architecture and market-leading product portfolio promise excellent prospects. Cisco believes that its data center solutions meet the immediate critical customer requirements. The company is ready to support customers who intend to deploy their own data centers and cloud computing solutions in the coming years.

Cisco Unified Computing System - Advanced Data Center Solutions

The Cisco UCS hardware line is a platform that combines compute and network resources, storage access, and virtualization. The introduction of this product is associated with the need to solve such problems as increasing the complexity of business applications, ensuring the continuity of business processes, their recovery in emergency situations, saving power and data center space.

Such tasks can be solved using data center virtualization, using virtual machines. However, there are other challenges - security issues, limited manageability, I/O limitations, and a host of other challenges facing past-generation data centers. The joint impact of these factors creates serious barriers to increasing flexibility and accelerating application deployment, realizing the full potential of virtualized data centers.

Cisco has many years of experience in developing data center products. These developments include transmission and storage networks, traffic balancing and optimization systems, information security solutions, advanced technologies, and new protocols. The manufacturer works with data center industry leaders such as Intel, VMware, EMC, and BMC to enable it to offer customers a single, unified Cisco Unified Computing System that can address these issues. Cisco has developed two lines of UCS server equipment - the B-series (blade servers) and the C-series (rack servers).

Blade Servers (B-Series)

Servers of this type are characterized by a simple and thoughtful system design, using the capabilities of unified fabric consolidation based on Data Center Bridging and Fibre Channel over Ethernet (FCoE) technologies. These technologies significantly reduce the number of components and accordingly reduce the cost of network infrastructure elements (many network adapters, LAN switches, SANs, various network cables, etc.), which ultimately reduces the cost of power and cooling.

One administrator can manage all components of a system that includes up to 40 chassis using UCS Manager. Software Cisco UCS Manager implements the concept of role-based and policy-based management using service profiles and templates. Information about power supply, cooling system parameters, as well as equipment status, network environment configuration and network Data Storage is contained in the service profile. The implementation of Stateless Computing technology, or the transfer of service profiles, allows you to minimize the maintenance time of the system and easily adapt it to rapidly changing requirements.

The memory expansion technology enables 48 DIMM slots in a separate server and provides up to 384 GB of RAM on a dual-processor system at a maximum frequency of 1333 MHz, creating an excellent base for virtual applications.

File:marvel cisco unified computing system 02.jpg

The Cisco Virtualized Network Adapter has exceptional features - it can create up to 58 virtual adapters and support Hypervisor Bypass mode. Support for VN-Link technology extends the edge of the network to the virtual machine and erases the differences in network infrastructure management for physical and virtual servers. All network connections are configured and managed centrally, without providing an additional switching layer for virtual environments. I/O port configurations and network policies move between virtual servers, increasing efficiency and reducing complexity.

The Cisco UCS architecture is based on 10 Gigabit Ethernet technology and the widespread Intel x86 server architecture using the Intel Xeon processor 5600, 6500, and 7500 series. All traffic switching in the blade system is performed through a single central switch using network modules with standard SFP + cable connections. Traffic is then transmitted from the network modules to the servers via adapters connected to the modules via a passive internal bus. Thus, the principle of one-time sneaking is applied here. This allows the cables to be mutated once and further through the management interface to configure the number, types, and characteristics of adapters for the data network and storage network. If you change the connection type of the server to the network, you do not need to install additional adapters or route new cables.

File:marvel cisco unified computing system 01.jpg


The technology capabilities of the Cisco UCS B-Series allow you to solve any problems in a virtualized as well as in a conventional environment.

Rack Servers (C-Series)

Cisco offers its customers a budget option - rack-based UCS. Many of the developments specific to the B-series are implemented here, including the use of UCS Manager software, memory expansion technology, and Intel architecture. All of this enables the customer to consolidate servers, networks, storage access, and virtualization into a single infrastructure that reduces total cost of ownership and increases business flexibility. You can manage C-series servers centrally from UCS Manager or by using the Cisco Integrated Management Controller-CIMC (graphical or command line) management interface built into the servers.

The C-Series covers all customer needs, from low-cost 1U servers to powerful 4-socket servers. Many Cisco and non-Cisco applications are certified for UCS servers. In particular, the popular Unified Communications Manager (Unified CM) application, starting with version 8.0 (2), works with VMware software on UCS servers. In terms of performance, such a system is much more efficient and has many more capabilities than previous versions.

The Cisco Unified Computing System is a next-generation data center platform that includes networking, computing, and virtualization support, as well as centralized management of all of its elements. Cisco UCS significantly reduces data center costs and increases business operational efficiency with Cisco-specific quality and reliability.

2010: Cisco Unified Computing System Description

The Cisco Unified Computing System is part of the Cisco Data Center 3.0 portfolio of solutions that help you move your data centers to a 10 gigabit architecture, implement virtualization, use a unified communication matrix, cloud computing, and clean technologies.

Cisco Unified Computing System combines computing, networking, storage, and virtualization into a single, energy-efficient system. This reduces the cost and complexity of information technology infrastructure, optimizes the use of capital assets and increases business flexibility. Based on industry standards, the Cisco Unified Computing System is a new computing environment with integrated management and a unified switching matrix. The result is a standard computing platform that can optimize virtualization, reduce the total cost of data center operations, and support dynamic resource allocation that increases business flexibility.

Cisco Unified Computing System:

  • Can reduce total cost of ownership, with capital cost reductions of up to 20 per cent and operating costs of up to 30 per cent;
  • Improves IT productivity and business flexibility. Application activation takes only a few minutes, not days, which allows the customer to do less to support existing systems and pay more attention to innovation;
  • increases scalability without increasing complexity. Cisco UCS is managed as a single system, regardless of the number of servers (1 to 320) and virtual machines (up to several thousand);
  • Improves energy efficiency by drastically reducing the need for energy supply and cooling;
  • can provide compatibility, interaction and protection of investments by means of the infrastructure based on industry standards.

2009: Cisco UCS Announcement

In the middle of March, 2009 the Cisco company announced the new platform for the data-processing centers (DPC) under the name Cisco UCS. Cisco Unified Computing is the industry's first unified architecture for a virtualized data center, enabling you to optimize your IT infrastructure while reducing the cost of complex and expensive hardware.