What to take into account when choosing an average server?

An average server is a much more complex system than one might think at the beginning. When choosing a system for an installation, there is a wide number of issues that must be carefully considered. Matt Allard, senior director, solution marketing, in Grass Valley, analyzes in this tribune some key aspects that, from his point of view, we must take into account when choosing an average server.

The best average servers of today take advantage of the basic technologies of the general IT industry, and then optimize and improve those technologies specifically for real time requirements and high availability of television broadcast and video production. The best suppliers build systems that use standard file systems, storage and networks.

These suppliers offer added value through optimization and adjustment of IT technologies for medium applications in ways in which IT suppliers cannot or would not. The implementation of the latest technology produces more profitable solutions.

An average server must incorporate a variety of the latest technologies. This includes 8 GB fiber trunk networks with 15/7.2K SAS discs in the topology of RAID-5 or RAID-6. Storage components alone will not provide a satisfactory solution, but must also be integrated with complete infrastructure. A IT server component must be included to act as a data bridge and manage dynamic and deterministic bandwidth requirements. For optimal performance, an operating system of 64 bits and file system are convenient. One way to provide scalability and profitability is to adapt the ISCSI protocol on 10 GB/1 GB Ethernet to provide a deterministic and real -time performance for the average movement.

Technologies and infrastructure should be combined to control bandwidth. Bandwidth is a primary factor to determine the reliability with which a system will work and what capabilities can provide. A server system must be designed to manage the multiple level width levels that may be necessary for deterministic operations. This can be provided by a shared file system that includes service quality mechanisms.

Three bandwidth levels must be managed:
• The first level with a higher priority is for the average in real time where customers are designed to never exceed the bandwidth set. This real -time performance is a service level not needed by most regular IT systems.
• At the other end of the spectrum we have the shared bandwidth managed for operations that are not in real time, such as file transfers.
• Between these two levels there should also be a bandwidth level reserved for
Critical time production activities, such as the edition, where each individual client is assigned a bandwidth that cannot be exceeded.

The three levels must be available at the same time. In many systems there is only one level of bandwidth management and you cannot depend on it to perform all tasks reliably and consistently.

An appropriate mean server system has integrated redundancy, buffering, and multiple levels of processors for guaranteed performance. The provider of said system characterizes storage performance for several production uses, and measures bandwidth yield by Luns, individual servers and controllers.

Another unique aspect of media servers is file management. Unlike the files found in most regular IT server systems, medium server files are large ... very large. The system, including storage controllers and file system, must be especially tight for this type of files.

In a standard IT server system there is a large number of small, simple files, with a rapid reading/writing and from storage. It is not necessary to read during the writing process; Delays and latencies can be tolerated.

On the average, delays and latencies are not acceptable, since they translate into black frames and sound losses. On the average, working with large files means that storage has many simultaneous readings during the writing process.

For adequate average file management, the server system must provide specialized buffering and prebuscass, and greater errors correction on the march, stock container gathering of multi-pista content files and optimization of the application layer for reading/writing efficiently and deterministicly. When evaluating a server system for average production, this is an area that must be examined very critically.

Many systems that claim to be appropriate for Media do not offer these optimizations, and therefore are not up to the constant performance necessary for the delivery of content.

Platform flexibility

An average server must be versatile so that you can offer different solutions adapted to a variety of unique requirements. You need to climb any dimension of performance and redundancy. The expansion capacity needs to climb to multiple and independent vectors. A system may need to increase medium channels, storage capacity, file transfer bandwidth, or operational bandwidth for production needs such as reproduction and edition in variable speed.
A Broadcast server and video production must be equally skilled as a media server and as a file server.

There are no second attempts during the delivery of average data, the redundancy at multiple levels is necessary. The system must provide redundant data routes among media customers, media servers, and storage, a design that does not have a single failure point. In terms of implementation, it means that specific practices, such as redundant RAID controllers, network redundant controllers connected through multiple servers and switches must be available. At a level of devices, there must be redundant food sources, redundant cooling and redundant storage units.

To interconnect the average, a standard PC with some processor cards is simply not reliable or robust enough. To provide a great availability reproduction and recording service for 24/7 environments, a device for broadcast and production, built for this purpose, must be used, which incorporates selected IT technologies.

With this approach, a variety of benefits can be achieved. The device can be reinforced for use even in mobile environments. Modular components allow easy access and service. A dedicated device can use incorporated specialized operating systems and in real time, together with multi-level personalized processors and general purpose to offer unique and complex characteristics, with reliable performance. Some of these features include multi-relations channels that can manage audio/video recordings, audio/video emission, super slow camera recording, recording and emission of 3D left eye/combined right eye, recording and video emission combined, multiple compression formats, and upper/lower/lower conversion resolution.

All this can be increased with the creation and management of average of low resolution proxy. At the same time, this type of device incorporates all the advantages of a standard storage platform, control interface, graphics and network connectivity to facilitate integration and reduce costs.

An integral part of any average server system is storage. Broadcast and production environments have different needs, so a server system must offer a range of storage solutions.

The simplest implementation is of SCSI (SAS) units connected in series that are placed internally to the average customer. To simply increase the capacity, external RAID chassis with SCSI units connected in series can be connected to the average customer through a bus adapter.

To obtain the highest yield and flexibility, an average server must support a storage area network (SAN). A SAN is a dedicated storage network that provides access to consolidated storage. Network storage devices are accessible to servers in such a way that it seems as if they were connected locally in the operating system, allowing direct requests for reading/writing to storage discs.

The server component acts as a bridge between the Ethernet and the fiber channel connectivity. It is also the seat of a shared file system to provide an abstraction of files with which they can interact applications and networks. A SAN provides scalability, connectivity, bandwidth and storage to allow a wide range of production scenarios.

As a server platform provides the essential infrastructure for file -based production, there are a series of features that must be provided so that interoperability is easy and efficient. The files must be stored as elementary flows on the disc for direct edition and metadata operations. There must be an importation and export in parallel with other network storage operations, removable supports, editors and file. The average files must support standard video compression formats such as MPEG-2, MPEG-4, DV, and AVC-Intra. These files must be available for the exchange in common wrapping, such as MXF and Quicktime.

Other types of files, such as WAV for Audio, Quicktime 32 for graphics and flow for transport video/MPEG program must also be compatible. The files must be transferable using standard FTP protocols and CIFS, as well as through direct ISCSI connections. To help automate the process of moving files, there must be supervision folder services to carry the files inside and outside the server platform.

As content owners wish to reuse material, creation, saved, and metadata exchange must be associated and managed with average files by the server system using an open standard, such as XML.

The management of the server system and its capabilities, as well as the maintenance of operating costs and the total cost of low property must be planned.

To simplify the use and maintenance of the systems, there must be a set of integrated tools and profits to help implement, maintain, configure and update the system.

Integration of applications

A server system is as useful and interesting as the applications with which it can be used. To have complete solutions, a server provider must also enable applications that people need to use. They must be applications that allow people to use a server system for live production flows, news, entertainment, and emission. A server systems provider needs to create applications that allow innovative and affordable solutions.

Some of these applications include the intake, repetition, video storage, edition, channel control, and broadcast. Application services must always offer proxy operations, lists/reproduction lists of common editing decision, keywords of common metadata and markers, and content management. Having these available applications shows that the supplier understands the details of the complexity that the necessary solutions must have to work through the total Broadcast infrastructure.

The user experience and comments are what must boost the design of applications. Applications must focus on reducing learning curves and making their use simple. The ease of use must also be applied to updates and maintenance. All applications must have common user interfaces and terminology. They must share compatibility of common files format, proxy, metadata and list information/edition decision -making list.

No average server provider can produce all applications that could be necessary for production users. Therefore, it is necessary to have an API available for other applications to be developed by third parties and even end users. The API should withstand common programming environments that do not require proprietary languages.

The traditional capacity of an API is for the control of the average channel. With the transition to file -based production, the API must also incorporate content management capabilities. API must provide direct access to average assets. It should be able to access through network services for the entire asset system with the ability to start file transfer through FTP and CIF protocols. All new features must be openly supported through the API. To create sophisticated application structures, there must be an API that supports a layer of architecture abstraction to services (SOA) to provide content management services, proxy, intake, emission, search, transfer and metadata.

With a well -executed API, an average server provider must have a team of development support engineers to promote a diverse community of active developers. They will help the supplier develop a wide range of support applications.

Test and validation

In order for an average server to offer all these functions and integrations, an extensive research and development work must be carried out long before any solution is delivered. Although the development of careful software and hardware design at the engineering level is expected, there are some other critical technical processes that must also be incorporated.

With the aim of having a close integration of a variety of standard IT components and subsystems, rigorous validation has to be performed.

Some of the technology evaluation criteria that must be included as part of the validation:
• Does a certain technology meet what promises in its characteristics and specifications?
• Can technology be used effectively for the planned application?
• Can technology also work as a fundamental pillar for the entire system?
• Is the development cycle to integrate technology predictable?
• Will technology climb as expected and operate under different loads?
• What will be the operational behavior in error conditions?

Some of the evaluation criteria for storage validation should include:
• Comparative performance evaluation for several days during reconstructions, and during the execution of various server input and output configurations
• Price compared to performance compared to latency
• Delimited latency with increased storage sub-systems that respond within a certain period
• Exchange storage controllers at full load, live firmware updates to full load
Some of the evaluation criteria for the validation of the file system should include:
• With a San Shared, how is access to the simultaneous access of
real time and not real applications?
• Automatic file system profile management so that defragmentation is not required, even in operation 24/7
Some of the evaluation criteria for validation of network data movement should include:
• Network data and data structures management, so that transfers are effectively without loss
• Characterization, configuration and management of the initiators and objectives of the devices, network interface cards, and network switches so that they work as a clean connection without clean end -to -end loss
• Analysis of the internal architecture of the network switch, port configuration, and microcode versions

Once a design has been completed and the software created, a component validation/ Multi-Etapa system and the test process must be completed. This process must be implemented on a long -term base and be very scalable. For continued support and new versions, the functionality and behavior of each piece of technology, and of the entire system, there must be extensive regression tests.
As technology progresses, there will be following generations of components. As they are incorporated, the system must be tested again with the new versions of the storage units, RAID controllers, file system, data servers, network interface cards, Bus Host adapters, and switches, to ensure that the system behaves as specified.

To put the necessary effort here there are some examples of what should normally be done with each generation of components to guarantee the operation of the system.
- Raid storage:
Six months with two engineers at the average server system supplier and three engineers in the supplier to validate. Three months for firmware version, regression tests, quality control, analysis of configuration tools.
- Series of shared storage units:
Four to six months with two engineers to develop firmware, validation and test. Part of the validation is for extended service for several years, thus storage can be expanded with future generations of units and have units, units groups, reconstructs all work within the planned performance levels.
- Series of internal units:
Four to six months with two engineers to develop firmware, validation and test. Performance and validation analysis, failure analysis, special management of the page for the use of average, integration of management and ease tools. As with shared units, not all families of units or units suppliers meet the necessary standards.
- Media networks:
After years of initial development, nine months with three engineers to develop the latest average network tools based on 10 Gigabit, comparative performance evaluation, failure analysis, supplier requirements, and integration. Six months of release tests, scale and analysis of the configuration tool.

Services and support

After a media server has been selected, the interaction with the supplier has to
Enter a new phase. Now, the supplier must implement experience in services to effectively design and implement a system that meets its business needs, and that supports the system efficiently during its useful life. The evaluation of suppliers in these areas is essential to perceive the benefits of the new technology, minimizing the risk and controlling the total property cost.

The design of an effective system is essential to perceive all the benefits of technology of
Media server. In this step, supplier system architects must participate in a discovery process to meet detailed technical requirements, and then translate those requirements in the system design. An experienced systems architect will ask you basic questions about your needs, such as the compression and speeds of preferred bits, the estimated amounts of storage, and the numbers of intake and emission of channels.

An authentic advisory approach to the best suppliers in your class must also explore the future potential expansion of the system, interface points with other company systems, long -term data continuity planning, and total tolerance to system failures. These considerations should influence the design process as well as simple bits, bytes, and channel data, which often have priority outside an advisory commitment.

If an effective system design is vision, then reality is built during the implementation of the system on the site. Here, the Team of Project Managers of the Media Server Supplier, field service engineers, and the trainers combine to commission new systems in real world environments. The system must be configured and tested for optimal performance, and then users must be trained in operation and maintenance procedures.

Each average server provider must have these basic capabilities. For systems of greater complexity and size, such basic implementation capabilities must be orchestrated by a detailed project management methodology for project control, schedules and risk.

In any supplier evaluation process, each supplier must be asked to explain its project management methodology, including the work declaration, project programming, communication between supplier/vendor, system documentation and problems monitoring.

An adequately designed and implemented average server system can provide years of operation such as the heart of various average delivery infrastructure. However, all systems, inevitably, require technical support to solve problems, maintain the updated system with compatible software versions, replace defective hardware components, and general maintenance of system operation. Potential suppliers must demonstrate their capacities and investment in these essential areas to ensure that capital investments are protected.

Beyond the basic technical support capabilities, medium servers suppliers with an advanced customer service infrastructure should be able to provide high availability services such as:
• 24/7 technical support by phone
• Remote system diagnosis
• Continuous access to software launches and associated installation services
• Advanced hardware replacement with delivery to the next working day
• Field Service Engineering

Media server suppliers that can offer such high availability services are capable of drastically reducing any system inactivity time, while guaranteeing the profitability flow of average assets for the entire company, air emission, and online operations.

In addition, first -level customer support organizations are capable of offering close relationship services such as the exclusive technical accounts management, 24/7 remote monitoring system and engineering support, even external.

Often, these services are offered in the context of a support agreement to the system established with response time parameters. Not only support agreements force the system supplier at a specific level of performance, they also definitely control property costs. A complete evaluation of possible media servers suppliers must include a review of their customer service, high availability services, and systems support agreements.