It's landscape definition. Landscape Design IT

There were times when there was no information landscape. Or did they simply prefer not to notice it? Information systems were distinguished by unusual arrogance and reigned in a sterile airless space. Becoming more and more complex, they began to think about their architecture - how to provide themselves, loved ones, with such attractive qualities as endurance and performance. But they were still distinguished by unforgivable selfishness.

Each implementation project automated systems(AC) tried to pay less attention to what the enterprise already had in IT (its current IT landscape), which was quite justified from a commercial point of view. The concept of landscape in relation to IT was one of the first used by SAP, when on a difficult path of self-knowledge aimed at obtaining competitive advantage and profit, found that, it turns out, even it does not cover all the needs of a modern enterprise in automation. Looking around, she found many systems, systems and programs that the company does not want or cannot refuse. Known methods of transferring, as before, data from one system to another, whether by typing it from one system and entering it on the keyboard into another, by copying or writing to Excel and unloading from there, have long shown their inefficiency. Moreover, they are dangerous for the quality of information. All those efficiency and reliability that could be achieved by introducing not the cheapest software with the help of not the most simple and low-cost projects crumbled with this kind of “integration” or, rather, “disintegration”.

First, users began to grumble, followed by business customers. Users - because of the mechanical and meaningless work of information carriers, customers - because of the lack of real benefits that were promised by the suppliers of heavy systems and the companies implementing these solutions. This is where the era of "IT landscape design" came. It turned out that there are many speakers around, equipment, communications, living people with their habits and experience, as well as corporate standards which must be taken into account and with which to get along. In other words, the light AC building has to be built in the real, not in the abstract world. And then a skyscraper in a summer cottage becomes at least inappropriate, like a change house in the city center. Of course, bushes can be dug up, a puddle can be filled up, a small hill can be leveled, and a building can be destroyed. But the more significant elements of the information landscape, the more difficult it is to change them to suit a new architectural structure. And at the very least, these efforts and the corresponding costs must be taken into account when analyzing the project.

Landscape and architecture

So, when forming the AS architecture, it is necessary to take into account the surrounding landscape, but the question arises: “Why did you need to introduce the concept of landscape when there is already the concept of enterprise architecture (Enterprise Architecture)?” A system of ideas, standards and methods has already developed around it. It is difficult to say what guided the authors from SAP, but it can be assumed that they needed the concept of landscape to separate the construction of a new system being built from the reality that is already present in the enterprise. There is one more aspect. If architecture implies a certain openness, clarity, formalization, then the landscape can carry more secrets and lend itself to painting with an indefinite degree of certainty. But everything that is called legacy systems is almost always bad or not documented at all. Therefore, when enterprise architecture description methods are applied to the landscape, we must be prepared for many gray and white spots. Filling in these spots, where necessary, is a science in its own right, or rather an art. Quite often, the impossibility of formalization forces companies to abandon existing elements of the landscape and plunge them into costly projects with a high degree of risk. Therefore, as in nature, the information landscape can be fraught with threats and troubles. However, having mastered landscape design, disadvantages can be turned into advantages by skillfully embedding the structure in the surrounding “nature”, skillfully using existing IT tools.

The landscape of a modern enterprise depends quite strongly on the type of enterprise (see Table 1).

Table 1. Types of enterprises and their corresponding information landscape
Type of classification Impact on the landscape Landscape Features
Industry For banks, a special role is played by ABS, for telecom - billing systems, for mechanical engineering, shipbuilding, aircraft building - CAD / CAM / PDM, for manufacturing enterprises- APCS and MES, for selling companies - Sales Force Automation and CRM systems. In some industries, the landscape is built around the central, most significant AS for the enterprise.
Territorial device Distributed, highly centralized companies require centralized systems that support distributed architecture and are resilient to network failures. In addition, it imposes requirements on networking and the entire landscape. Territorially distributed companies with strong centralization are characterized by high quality communications and software reliability to network failures.
Ownership type State enterprises actively consume document management systems, commercial ones - analytical systems. The type of ownership determines the requirements for document flow and financial management automation tools.
Size For large companies performance and scalability of the system, its reliability are important. For small - ease of installation, configuration and maintenance. The requirements for enterprise architecture components are different, so the look of the landscape varies significantly.

Virtually all modern companies have a patchy information landscape, and what are called legacy systems play a prominent role in it. For a long time no one offers to cut all this down to the root and completely transform it in a single project. Therefore, not only any implementation, but most changes in any of the architectural layers require an analysis of how these changes will affect the entire enterprise landscape. Unfortunately, the quality of change management in most cases leaves much to be desired. Few people think about the landscape. It's good if they limit themselves to at least adjacent "bushes", or even plan to introduce AS, as in the good old days - without taking into account the existing information landscape.

Classification of landscape components

Let's try to look (Table 2) at the components of the information landscape, at all these bushes, trees, berries and flowers, and how they fit together.

Table 2. Elements of the information landscape
Terrain element type Main characteristics Integration methods
Large corporate systems (ERP, CRM, ...) Wide range of functionality

Competent architecture

Implementation methodology

Good documentation

Well-developed APIs and well-thought-out technology improvements

Thoughtful integration techniques, interfaces, adapters described in XML

ESB, J2EE Application Servers, .NET, CORBA, XML, EDI

Ready systems narrow specialization(best of bread) (document management, project management, systems accounting …) Systems that occupy one of the leading places in their segment. They usually offer better functionality in their area than larger systems. Integration options are not as diverse as large systems, but the choice of technologies is usually the same.
Functional platforms (BPM, service support, …) Systems that offer convenient means of assembling speakers in a selected functional area. Most of this assembly does not involve coding, although these systems always have an API. Usually, ready-made solutions are also offered on these platforms, which can be further developed. Integration methods are determined by the technology on which the platform is based. The most common platforms are J2EE and .NET.
Software platforms - development environments (Eclipse, Visual Studio, NetBean) Software platforms offer developers a convenient means of assembling applications of wide functionality and usually do not carry elements focused on a specific functional area. Many platforms offer visual development tools that allow you to partially build applications without coding. Integration tools are determined by the technology used, in this case the programming language
Custom developments Usually, custom developments are at least somehow documented. Integration tools depend on what, in fact, was ordered, so if there are any integration plans for the development, it is better to immediately include the relevant requirements in the TOR.
Own developments Own designs can also be created on the platform. They are usually poorly documented or not documented at all and are often a black box that cannot be changed. The most difficult landscape elements to integrate

In the general case, none of the elements of Table. 2 is no better or worse than the other - it all depends on the specific company and situation. In particular, hard-to-integrate native programs in a company's well-established software development process may be more suitable for combination with other components than all other elements. In favor of the latter conclusion, one can cite the example of the largest Japanese telecommunications operator - Docomo. All software in this company is developed independently, for which Docomo keeps a considerable staff of IT specialists. They explain this approach simply: “For us, flexibility is the most important competitive advantage. And business flexibility requires the flexibility of our speakers. Therefore, we are ready to pay several hundred developers, invest effort and money in their training and motivation, but have “all the changes in our hands” for this. As far as one can tell, many Japanese companies take this approach, which, apparently, the best way corresponds to the Japanese business miracle.

Russian companies, on the contrary, are now at the stage of transition from their own developments to outsourcing. Many managers believe that in this case they can completely get out of managing the process of creating an AS and recover from an excruciating headache. Their illusion is that all other elements, except for their own developments, can not be controlled and the systems created by outsourcers, planted on the soil of their own enterprise, will be quite compatible with each other and with the surrounding landscape. How many times have they told the world: when choosing any component, the requirement of its compatibility should become one of the most important - and things, alas, are still there. It is clear that in order to put forward such a requirement, one must know what to ask - the information landscape must be described at least in some reliable form. And who has it? So unviable freaks are born - AS, which do not take root and cannot take root, and wither away in the corporate landscape under the indignant lamentations of those who tried to instill them.

Terrain Merge Tools

There are two "news": good and bad. The good news is that modern IT allows you to integrate anything with anything. It is a matter of resources (effort and money), the reliability of what will be the result, and the effort that will be required to maintain and maintain an integrated solution. However, the last question is clearly not studied. When they talk about the natural landscape, they remember that without proper care it will very soon turn into a wild thicket. When they talk about the information landscape, they believe that it is important to walk around it once, and then everything will be fine. Like many delusions, this lies in the depths of the subconscious, so it is difficult to fight it.

The bad news is that any integrated solution is an order of magnitude more complex, harder to maintain and develop, more expensive, and at the same time much worse than the one sold to the business. The last one is very significant. CIOs, integrators and consultants alike face this challenge. After all, this very integration cannot be shown and implemented separately from what is being integrated. And it is rather difficult to explain that now the information will be of a different quality, since the concept of information quality is not developed at all, unfortunately.

So there are integration projects attached to implementation projects, they are financed on a residual basis and done “on a living thread”, meaning that later, somehow, when the time and budget appear, we will do everything right. However, there is nothing more permanent than temporary. Therefore, the incompleteness of integration comes back to haunt the enterprise for many more years, not allowing either to receive the promised benefits or competently implement the necessary changes. As an illustration, one can cite the reasoning of one IT manager who persuaded the management to implement at least integration and thus show the success of the project: “Well, can we link the two systems? Ah, already something works! Well, let's show it!" Obviously, it was necessary to show only a virtual bridge between two systems that are not yet working.

Systems for combining elements of the information landscape can be divided into two groups: tape and umbrella.

Tape ones are usually built on the basis of the ESB bus, to which separate speakers are attached to “suction cups” - standardized interfaces, interacting with each other through the bus. These systems grew out of CORBA technology and messaging systems.

Umbrellas, which are also often ESB-based, access the disparate systems of the information landscape, covering them like an umbrella. In its simplest form, these are corporate portals that provide access to a variety of organizational systems. The BPM systems that are actively gaining popularity today are of the umbrella type.

Both tape and umbrella integrated systems most often require processing of integrated speakers, unless the latter have big amount predefined adapters or interfaces. Because in an integrated landscape world, this acquires everything greater value, many manufacturers are trying to provide their offspring with a set of convenient and reliable integration features. They believe, not without reason, that this is one of the important competitive advantages. I would like to see consumers, in turn, learn to appreciate these efforts.

On the other hand, manufacturers only care about their "bushes and trees" of the information landscape. It is hardly possible to demand from them that they act as gardeners of the entire information landscape of the enterprise - this is a field for system integrators, both in name and in essence. However, while domestic integrators do not seek to take on this honorable, but very difficult role. With the cloud threatening to take over large chunks of the IT market, system integrators must jump into this still half-empty business area.

If the Network unites people in different parts of the world, then it is also great for bringing companies together, for building what is called the Supply Chain - supply chains for organizing the interaction of suppliers and consumers, for joint work and crowdsourcing (attracting a "crowd" of voluntary employees to the production and sale). All this suggests a very different information landscape compared to the one we talked about earlier. Perhaps even the term "landscape" is not entirely appropriate here, but "information world" or "business Internet" is more suitable. Obviously, the requirements for integration tools will only grow.

However, so far, the interaction of companies, surprisingly, is still often based on the ancient EDI, when one AS put the file, and the second took it away. Another thing is that now these files are in XML format, in particular, it is on EDI that the interaction of the largest automobile concerns with component suppliers is usually built. By the way, for those who have not yet realized that integration has already ceased to be something that can be easily dismissed, you should know - without providing operational information about the passage of the order, without integration with the AU used by the auto giant, the company will not even be considered as a supplier of auto components.

Integration, issues of creating and maintaining the information landscape are increasingly capturing banks and telecoms, manufacturing and retail companies. In any case, the future is about an information landscape that makes it easy to add new elements, remove and replace existing ones, and treat enterprise IT as a whole.

Marina Anshina([email protected]) - Chairman of the Standards Committee Russian Union IT directors (Moscow).

The introduction of ERP or CRM systems is impossible without building an information landscape that combines the concepts of deploying engineering, network, computing infrastructures, backup and recovery systems, management and administration. The concept of the IT landscape is not only the basis for the development of IT services, but also forms approaches to building architectures that meet the present and future business needs of growing companies.



With rich experience in systems integration and business consulting, Inkom Corporation sees the IT landscape as a fundamental basis for automating business processes. On approaches to architecture design and practical implementation system landscape PC Week/UE talks with Sergey Tikhonov, Head of Server Solutions Implementation and Maintenance Department, Dmitry Kruchinin, Director of Storage and Processing Systems Department, and Eduard Savushkin, Director of IT Consulting.

PCWeek / UE: Who usually initiates the issue of building a landscape - the customer or the integrator? Does the customer always understand the importance of landscape development?

Dmitry Kruchinin: Both the customer and the integrator are equally involved in the process of designing and building a landscape. For customers who encounter ERP systems for the first time, the concept of “IT landscape” may seem new. In this case, we explain the necessity and importance of the correct construction of the landscape. We tell that from the interaction of servers, network equipment, control software - those components that are today the core of the business of many companies - their competitiveness and success depend. Since the ERP system is becoming a critical element of the entire business, it must be treated with the appropriate attention.

Sergei Tikhonov: With customers who have already encountered ERP systems, the conversation is conducted in the same language. In any case, we, as integrators, present our vision and approaches to the client, after which we discuss the specifics and wishes.

PCWeek/UE: What size companies are interested in landscaping?

Eduard Savushkin: Building an IT landscape is essential for companies of all sizes. Even in solutions that are designed for 20 people, implementation starts with landscape design. It is very important at the first stages to define the concept of the vision of the system at the top level, to form a vision - a strategic document that defines the rules for building the company's IT infrastructure, the main architectural solutions and standards, the model and requirements for management processes. In fact, the concept of construction is an idea of ​​the future of the system, which determines the ways, methods and plans for its development, allowing to evaluate necessary resources. The design itself takes place taking into account the four phases of implementation - development, testing and stabilization, commissioning and further support.

PCWeek/UE: According to your concept, what are the components of the IT landscape?

E. S.: At the logical level, the infrastructure model consists of basic architectures:

. network architecture;
. application architecture;
. architecture of management systems and IT services;
. architecture of security systems;
. storage architecture.

The principles of building an IT landscape are the same for everyone. The architectural approach professed by Inkom is maximally adapted to our conditions - we try to get down to earth and speak with the customer in the language of practitioners. In many ways, our approaches are based on the Microsoft System Architecture, on the basis of which a system of components is created that describes:

. approaches to building information infrastructure;
. IT services are technological systems that solve business problems;
. logical model of the IT infrastructure.

For example, when considering a network architecture, we can talk about the physical layer that describes channels and equipment, and about services that include encryption, routing, mail, and so on.

PCWeek/UE: Where does the practical filling of the landscape begin? Is there specificity depending on the type of activity of the company or its structure?

D.K.: Our first task is to find out from the customer his expectations from the ERP system, to determine what kind of equipment will be required to build certain layers of the landscape. At this stage, consultations are held at various levels, business and technical requirements.

S. T.: Then we proceed to the sizing procedure - based on the data on the number of operations, the amount of functionality, the number of users, the requirements for software and hardware are formed. The sizing procedure is performed according to generally accepted methods developed by software manufacturers. They are agreed with equipment suppliers who can answer which set technical means will be required for the functioning of the system with the given parameters of performance and reliability. For example, SAP issues requirements adapted to the hardware of the world's major A-brands. After the load becomes clear and the hardware necessary to ensure the operation of the systems is determined, we begin to form the content of the landscape.

E.S.: The IT landscape is designed as a complex. Some customers try to start right away with the network, without considering other aspects. As a result, after the deployment of services, it turns out, for example, what is missing bandwidth channels. Therefore, at the first stage it is very important to analyze the system, to understand how its architecture as a whole will look like. Only after the information flows and structure are clear, the IT services provided are formalized and measures for their availability, security and manageability are provided, it will be possible to start implementing one or another part of the landscape.

PCWeek/UE: What criteria do customers use when choosing a hardware manufacturer?


DK: According to the requirements formed as a result of the sizing procedure, A-brands offer solutions that are approximately equivalent in essence. At this stage, the integrator, together with the client, proceeds to the selection of equipment, taking into account additional parameters - service, response time and recovery.

Since many customers have already encountered not always adequate vendor service in Ukraine, today more and more attention is paid to the integrator's ability to provide high-quality service and technical support implemented system. The response time of the vendor, in which the client will receive a call and sympathize with the problem, may be quite acceptable, but the recovery time is not. In such a situation, an integrator comes to the rescue.

If you try to evaluate on a 10-point scale various factors that influence the choice of the customer, then I would give 6 points to the service, 3 points to the price, unique technical features — 1.

PCWeek/UE: In practice, the implementation of ERP-systems in the enterprise is often carried out in stages. How in this situation to correctly estimate the requirements for the equipment, if the final scale of the system is unknown?

S. T.: Systems are always designed with future development in mind - large customers represent the scale of business growth in the foreseeable future. At once, of course, no one buys power for many years to come, however, the design of the landscape and its constituent systems is initially designed with a certain margin for expansion.

D.K.: A-brands offer different implementations of horizontal and vertical performance scaling. To put it quite simply, vertical scaling means installing a large server, to which, over time, boards with processors, memory, etc. can be added. Horizontal, on the other hand, implies the installation of several systems of the same class, i.e. a similar one is placed next to an existing device (server, storage, etc.). Blade servers are a typical example of a horizontal approach. At the same time, the purchase of a blade system in a minimal configuration will cost more than a set of several servers, however, the potential, manageability and ease of expansion of such a solution will allow the customer to subsequently save significant funds. The system integrator offers for consideration to customers different ways scaling, but final choice, made by the client. Some prefer and are able to make a large initial investment, while others prefer to gradually expand the system by adding new modules. This is the age-old dilemma: save today but pay tomorrow or invest today and save tomorrow.

PCWeek/UE: Is it practiced to use the client's existing terrain elements?

E.S.: Often the client can provide systems for development and testing, but a productive system is almost always purchased new. Adaptation of the existing fleet of machinery to a productive system is not practiced, since the operation takes place under high load. In some cases, we can only talk about using legacy technology for other tasks related, for example, to building hierarchical storages, using backup sites, etc.

According to our estimates, in about 30-40% of cases, the customer needs to build a system from scratch. Such projects are now characteristic of retail chains that are actively developing regional chains, and Western companies that are opening operations in our country. In other cases, it is necessary to build a system by adapting existing elements of the landscape - this is relevant when introducing new management systems into existing structures. Such customers usually already have some kind of network infrastructure, less often engineering.

PCWeek/UE: Is it possible for terrain to exist without one of the layers?

D.K.: This is one of the questions that sometimes takes a very long time to discuss with a client. Companies have different levels of decision making. IT professionals understand the need to deploy all layers quickly enough and perceive our arguments. At the financial level, they begin to count money and try to abandon certain systems. At the highest level, we can meet either complete understanding - if the manager realizes the importance of ensuring the continuity of the business - or rejection. Each level needs its own arguments.

S. T.: Theoretically, SAP, Oracle, Siebel, and other systems do not need a backup and recovery layer to function. Systems will work without it ... until a certain point, after which disaster will come. Without building backup and recovery systems, which are not the result of sizing software manufacturers, work in practice is impossible. Understanding this comes to clients during the discussion of such important points as the cost of information processed in information systems.

PCWeek/UE: Isn't geoclustering a one-stop solution for maximizing security and availability?


E.S.: The fundamental questions are what, how and where to store. A specific implementation of a storage system can be, among other things, a geocluster. It is important that the customer is aware that the cost of a backup is the cost of the hardware, and the cost of a restore is the cost of the business. That data loss is a crash. Information management concepts throughout life cycle, the very ones that A-brands love to talk about, are now implemented in real systems.

Customers need to properly build fast backup and archiving systems, form and practice backup recovery procedures. For existing systems It is critical to match the value of enterprise information, applications, and data with supported technologies and service levels in complex data centers. Defining rules and policies for the presentation and implementation of ILM, improving the quality of service can reduce the cost of ownership of the system and optimize the use of resources. An integrated approach to managing business information resources is focused on directly supporting the most important goals of the organization. Well-designed backup and disaster recovery systems will make it possible to resume the operation of systems even in the event of physical destruction of storages in one of the data processing centers.

D.K.: ITIL says - create backups, practice restoring data. For data centers, it is generally recommended to perform a service failover exercise every six months to make sure that the procedures adopted are working or not. You can not do this, but when a critical situation comes, it turns out that this was not written in vain. If the client refuses to deploy this or that layer, he himself must bear responsibility for this.

In general, disaster-tolerant systems are no longer a rarity and no longer a fad. Ukrainian companies have come to understand that data is a key business asset. The first geocluster system was built by Inkom specialists back in 1999 at Aval Bank using the equipment of those years. Channeling equipment today has become so powerful and capable of providing such speed and functionality that it has become quite simple to build a geocluster and implement replication between sites.

PCWeek/UE: How willing are customers to spend money on security?

E.S.: At the pre-audit stage, it often turns out that customers themselves do not know what equipment they use. An inventory is carried out, a map of the customer's infrastructure is drawn up.

Then the system architecture and its components are designed, for example, a communications system: mail, voice, portals, and other services. After and in parallel with these processes, the deployment of a security system begins, the formation of requirements for personnel, processes and technologies.

D.K.: Clients whose business is built on information - banks, insurance companies, financial institutions - are well aware of the importance of security. To a lesser extent, as practice shows, retail chains and production think about security. Unfortunately, our market has not yet formed an understanding of integrated security, and there are not always generally accepted approaches to building such systems.

E.S.: After the formation of the general concept of security, a conversation begins about the software part, about ensuring collective work on documents, etc. Often on the periphery, the staff of companies lacks qualifications, so they turn to us on management issues, down to elementary things - how to unify software and management of workstations.

PCWeek/UE: Are landscape engineering infrastructure design and construction services in demand?

D.K.: No one will make infrastructure for its own sake. Serious customers understand that systems need to be built as a complex, and not separately - power supply, SCS, server system. Otherwise, it is impossible to predict the mutual requirements of the systems and provide the necessary expansion margin. The demand for building the right infrastructure is huge - these questions are important both for companies setting up a business from scratch, and for those who are looking to expand or modernize their systems.

S. T.: Some companies attach third-rate importance to the issues of building engineering infrastructure - power supply, air conditioning, restricting physical access to premises, fire safety, radiation. We have to argue the importance of these components. It is not uncommon for equipment to fail due to dust, overheating in the server room, and unstable power supply. All these problems are solved by the correct organization of the engineering sector. A whole structure of requirements has been developed, which branch and determine the degree of protection of the engineering layer and, accordingly, data from external influences. The cost of organizing a server room is sometimes higher than the cost of a director's office with all the expensive furniture and an aquarium.

PCWeek/UE: Are there any peculiarities of the IT landscape for different business segments - manufacturing, telecommunications, banking?

E. S.: Features of the landscape are determined by the architecture of the implemented control system. An important role is played by the number of users who will work with the system, business requirements for fault tolerance, availability of data, services. Specific specifics are determined not by the industry in which the company operates, but by the requirements of the business.

D.K.: We can talk about certain specifics in the banking sector. It is conservative and focuses mainly on Unix systems. A segment retail to some extent it is located to Windows-solutions. However, the choice of platforms is mostly individual, it is difficult to talk about any predetermined positions here.

E.S.: In the banking sector, the issue of security of retail services is relevant. It is necessary to serve hundreds of remote points of presence, ensure the safe operation of various services in the field, and create a system of centralized storage and access to data. Some customers prefer to store data in-place, but the majority is still leaning towards consolidation. This approach allows you to significantly increase the level of security and provides more tools for organizing business continuity, reducing the cost of data storage while increasing the efficiency of the storage infrastructure. There is a noticeable interest in building a corporate infrastructure in accordance with international standards (British Standard 7799 and its development International Organization for Standardization (ISO) Standard 17799). Policies developed in accordance with these standards can provide the necessary level of requirements for personnel, processes and technologies to ensure that authorized users use assets correctly.

PCWeek/UE: Which industry companies are the most technically equipped?

DK: We consider the banking segment to be the most technologically mature. The NBU has long forced commercial banks to switch to electronic system payments, to close the balance every day. It is impossible to imagine a bank that has not closed its balance sheet at the end of the day, which cannot be said about enterprises that may not know their own balance sheet for half a year. Banks often have well-developed layers related to management and administration. The telecommunications sector, retail chains are pulling up, where they realized that data management is a competitive advantage.

PCWeek/UE: Is there a specificity of building layers in geographically distributed companies?

E.S.: Retail retail chains put forward increased requirements for the width of the channels. Depending on the location of the point of presence, cable, radio or satellite channels are used. In Ukraine, there are a number of companies operating through a satellite channel, which is due to the underdevelopment of cable networks. If the enterprise is growing rapidly, then often a satellite channel is the best solution. For example, a retail chain opens a store, but it is impossible to lay the cable in time for the opening. The satellite channel will come to the rescue. It will not turn out to be superfluous and subsequently - as a backup.

PCWeek/UE: How do the integrator team and the ICS service interact during the implementation phase? When does the division of teams take place?

E.S.: After the end of the sizing procedure, the customer is determined with the equipment - servers, network devices, data storages, etc. The equipment “arrives” in the server room. After that, the integrator proceeds to implement the system for development, and at the same time, the transfer of knowledge to the customer's specialists begins. The further the implementation project progresses, the more functions are transferred to the customer's specialists. The integrator conducts trainings, the client sends his specialists to courses on the vendor's products, on the ERP system. This is often stipulated in the terms of the contract. The customer receives knowledge on the ERP system, hardware, network, DBMS and other components.

Often the customer installs a productive system on his own, the integrator only observes and provides assistance if necessary. After the completion of the project, a documentation package appears, which describes the system in the most detailed way and fixes its final state.

PCWeek/UE: What factors determine the composition of the implementation team? How many people does it usually include?

S. T.: First, the ability of the enterprise to allocate the necessary specialists. Secondly, the scale of the project and the required speed of implementation. Thirdly, the plans of the enterprise for the further development of the system. In some cases, it is enough that the administrator only maintains the built system, while in others the customer intends to expand it independently. Sometimes the enterprise does not have the opportunity to allocate personnel for the development of the system - in this case, this work is transferred to us.

PCWeek/UE: Can we talk about the average time to build a landscape?

E.S.: Optimal time from the beginning of the project to the commissioning of the system - from 3 to 9 months. But it must be understood that the development of the landscape never stops. The enterprise is growing, new business processes are emerging, which require new technology, infrastructure, control modules of the system. An ERP system must always meet the requirements of the business.

Over the past few years, the Zvezda machine-building enterprise has been consistently building an integrated information system. Zvezda produces high-speed diesel engines for shipbuilding and railway transport, as well as emergency standby power plants for various industrial and defense facilities. The company has a full production cycle - from development to service. There are about 40 thousand parts and assemblies in production, and the number of technological stages reaches 80-100. “The main task of an enterprise information system is to operate with all these data in their interconnection,” says Pavel Plavnik, CEO of Zvezda OJSC, who spoke at the Eyes Wide Shut section of the CIO Congress White Nights 2009. The most interesting and important points this speech, as well as the interview that followed, we bring to your attention.

Eyes wide shut management

Today, one of the key tasks of the enterprise is to create an integrated information system. What benefits will we get? Is it possible to estimate the economic effect of IT? We do not analyze in detail economic efficiency each of the blocks of the enterprise information system, since the benefit, from my point of view, is obvious.

I will bring small example. Previously, no special fixtures were required to control the processing of the part on the machine, because the metal processing speeds were not very high. The control was carried out simply visually, and even protective screens as an obligatory element appeared only towards the end of the last century. But today's machine tools process metal with special ceramic tools at very high speeds. Naturally, the processing process is already going on in a closed chamber and at such processing speeds and coolant pressures that it is practically impossible to control it visually. It is necessary to use sensors and display information on a special scoreboard.

It is the same on the scale of the entire enterprise - "eyes wide shut" today simply cannot be managed. There are many ways to transfer information, but their effectiveness is different. So much so that there is a critical amount of information and the necessary speed of its processing, in which IT technologies cannot be dispensed with. In today's environment, we can only ensure the manageability of the enterprise if we have IT tools.

About production planning

Back in Soviet times, while working as the head of the main dispatching bureau, I was engaged in the organization of labor. After all, the situation at all factories was the same: one did not fill in the blanks, one did not issue any invoices on time, this one did not provide the technology. The constant stress of poor work organization made me want to build a planning system. After all, the task is extremely clear - taking into account the existing standards for every detail, a plan that is approximately repeated from month to month, etc. And I built this system, having spent quite a lot of time. The calculation was actually carried out manually - using a calculator (Excel did not exist then).

The task is solvable, but, unfortunately, it will no longer be possible to take into account today's large amount of information in this way. However, now IT has reached a qualitatively different level. New products have appeared that take into account the first negative implementation experience. They allow solving production planning problems, and it is impossible not to use their capabilities - this is completely obvious to me. Otherwise, we will constantly manage backdating- plug holes. You can not drive the car, looking only in the rear-view mirrors. Management is always looking ahead.

On the quantitative assessment of the effect of implementation

In order to absoluteize the quantitative parameters that we would like to achieve from the implementation of this or that IT system, we need to have the right conditions for obtaining them. Alas, without having built a clear and detailed accounting in real time, it is impossible to achieve the required accuracy of these parameters. Without an ERP system, it was very difficult to unambiguously assess the economic efficiency of certain management decisions.

And the second point is the optimization criteria, the reliability of which is also necessary. For example, what should be the level of work in progress? In Soviet times, there were industry instructions, scientific calculations, but with our demand dynamics, they do not work. Then what to rely on?

About "inoculation to IT"

In the 1990s, the company made an unsuccessful attempt to implement an integrated ERP-class management system. The project failed for a number of reasons: a fairly large and complex software product and high dynamics of legislation. As a result, it was not possible to combine the production management system and the accounting system, taking into account all the features of tax reporting and the dynamics of its changes. The accounting systems of economic services, accounting and production, differed too much.

Yes, we have invested in IT without getting the desired effect. But as a result, this project has become a kind of stepping stone - the preparedness of employees to work with such systems has grown. Even if the project does not enter the trial operation stage, we nevertheless create the groundwork on the basis of which we can grow further. Including thanks to this work, the preparedness of the environment, with some stabilization legislative framework and a new level of IT development, turned out to be sufficient to reach the level of practical benefits from IT in the following projects.

We have a kind of "inoculation to IT". The experience of the failure of the implementation of the previous system was taken into account as much as possible. With the new implementation, we paid special attention to the tasks of docking production and accounting. Even so, there were great difficulties. And it’s not always possible to rise to a new level from scratch - that’s why unsuccessful projects are also useful.

IT base of an industrial enterprise

After the unsuccessful implementation of the ERP system in the 90s, we took a different path, which was based on the idea that for industrial enterprise the product data management (PDM) system shall be used as the reference. Today, the entire structure of our products is fully described both from the design and technological side (standards for the consumption of materials, labor costs, etc.) in the information system. This is the basis that allows you to work with information structure our products (see pic.). We chose the Lotsman PDM system, which copes with our tasks of designing and preparing new models of equipment for production.

Enterprise IT landscape

The most expensive and at the same time the central element of an enterprise information system is the SyteLine ERP system (see fig.). It is one of the heaviest IT systems in the enterprise. Its implementation was not easy and took about two years, the project, as they say, drank a lot of blood. But from the point of view of efficiency, today it is the most important tool that allows you to quickly collect all the information and comprehensively plan the activities of the enterprise. The system allows real-time control of processes in all departments and greatly simplifies the procedure for monitoring production, purchasing activities, planning the timing of production of each order separately and accounting for costs - including separately for each of the orders.

All of the above information is collected quite easily, in an automated mode. Thanks to this, today we can clearly plan labor resources, which we need under production program six months ahead. Without an ERP system, it was impossible in principle to calculate in advance with accuracy the labor resources for each site and workplace. Although I'm not very pleased with the interface of this product - in my opinion, it needs to be refined and refined to meet the needs of managers, making it as visual as possible.

We also use the interesting OrgMaster system, which allows you to automate the management of business processes in an enterprise. The structure of the enterprise processes was completely rebuilt during the reengineering project several years ago, and the software package made it possible to consolidate these processes by describing the activities of each department down to the workplace level, as well as continuously monitor all changes occurring in the processes. In order to solve the tasks of monitoring the performance discipline at the enterprise, increasing the speed and quality of internal and external communications, we have implemented a system electronic document management NauDoc. It has been working at our enterprise for about five years, it is constantly developing and allows almost instantly to process both the processing of incoming information and the movement of documents within the enterprise, including control of executive discipline. It significantly speeds up the preparation of correspondence, organizational and administrative documents, contracts, and at the enterprise in any document (including during the preparation of a contract) electronic signature equates to physical.

This picture will be supplemented with new elements in the future. In particular, in the near future we will be implementing a system for customer relationship management, as well as a system for automating maintenance and repairs.

Advantages and disadvantages of integrated automation

The main prerequisite for the formation of a complex information system is the possibility of forming and updating significant information resources. The effectiveness of such a system lies precisely in its complexity. It is used by production workers, sales managers, and economists. All these functions are important, and it is wrong to place an increased emphasis on any one. I personally work not so much directly in the system as with ready-made reporting forms. But I have a strict requirement - that handmade were not included in these reports. Naturally, such a complex information system has pros and cons.

pros Minuses
The release of the "army" of accountants, planners, accountantsInevitable "adjustment" of the control system to existing software products
Reducing the role of the "human factor" in the reliability of management information. "Transparency" of the enterpriseThe need for careful software selection
Get accurate, just-in-time informationThe need for the most severe administrative efforts to implement
Possibility of modular scaling up in the manner and within the time frame determined by the administration of the enterprise"Transparency" of the enterprise for regulatory authorities

IT and rigidity of thinking

Active use of IT inevitably requires adjusting the management system to the capabilities of software products. For 70 years, certain stereotypes and standards of work have developed at the enterprise, and, like in many machine-building enterprises, there is some inertia and inertia of the old thinking. And this is one of the serious problems that have to be overcome by adapting to the capabilities of the information system. But as a result, we only benefit from it. We are significantly simplifying many things, for example, the wage system. Prior to the implementation of the system, we had approximately 20 different payout mechanisms for various groups workers. By introducing a standard software product that does not have the functionality of such a complex calculation wages, we abandoned the old payroll system and unified these processes.

On anti-crisis measures

In 2009, the budgeting automation project was completed, which, although financed before the crisis, became a good anti-crisis measure. Now, of course, the requirements for efficiency and the level of analyticity of information are increasing. Our budgeting system based on the Plan Designer software package just allows you to perform motion analysis Money, tracking various debts by time and other factors, multivariate analysis of development prospects and much more.

About interaction of the chief information officer with a management

Implementation of a software product is not an initiative of the CIO. In particular, at our enterprise the main initiative came from managers various directions- for example, the chief accountant or the chief mechanic. And for such complex systems as ERP, the main initiative belonged to key leaders- me and the financial director. Therefore very important aspect IT development in the enterprise is effective communications IT directors with management different categories. It would be very useful to inform managers in a timely manner about the possibilities of modern software products, to convert people to their faith. This is the main piece of advice I wanted to give to CIOs - to communicate more with employees who directly use IT systems.

Moreover, this is very important not only during the start of the project and the choice of the product, but also at the stage of its operation. Since the presence of unique IT systems does not guarantee their effective use, not everyone is very happy to use these products. This is one of the big problems. And the authority of the CIO should be used to the maximum to enable information technologies into the actual operation of the enterprise.

  • Project management ,
  • Product management
  • What is everyone's problem Russian companies? The business has its own understanding of what is happening, its own requirements and tasks, but it is very difficult to translate them into the language of specific IT actions, and these requirements do not always reach IT in an understandable form.

    The second traditional problem is legacy IT. There have been a lot of chaotic purchases and implementations in the past years, which creates a lot of complexity in the IT landscape. Any IT manager who comes to the company takes a lot of time to understand the current situation. On the other hand, businesses invest money in IT and want to see where it goes, what projects and initiatives it takes, and how it moves the company toward strategic goals.

    In Russia, the direction of corporate architecture has not yet been developed. Companies are in great demand for the development of architectures for specific solutions, but no one is seriously engaged in linking the components of the architecture within the organization. At best, spreadsheets are used for this purpose, where information is stored in an unformalized form, and when a manager leaves the company, knowledge goes with him.

    The most important consequence of this approach is the suboptimal spending on IT, sometimes even their direct waste. That is why companies abroad are increasingly demonstrating a strategic approach to IT management, and the vector is increasingly leaning towards the streamlining and standardization of corporate IT landscapes.

    Words into action

    The problem is obvious: it is necessary to connect business with IT, build a bridge between them, ensuring transparency. The strategy needs to be translated into requirements and actions that need to be communicated to the people. Each department must understand the essence of the work and the general direction of IT development.

    HPE Enterprise Maps is designed to solve this problem. Importantly, this is the company's own development, and not an acquired solution, which has recently become a rarity. That's good: HPE, as a system vendor, came to understand what the market wanted and needed, and took the initiative to develop such a product. As an in-house development, Enterprise Maps seamlessly interacts with and enriches data from other HPE systems to form a cohesive, powerful solution.

    Rather than reinvent the wheel, Enterprise Maps uses the TOGAF methodology, which is understandable to architects and shows how to approach enterprise architecture management. ArchiMate 2.0 notation is used as a language for describing the interaction of architectural elements with each other.

    The main objectives of the developed system are to reduce the risks and confusion associated with IT, to ensure compliance with the IT strategy. It is extremely important to be able to translate “beautiful” words into concrete actions. In addition, meaningful management of the “zoo” of existing systems becomes a non-trivial task: there are a lot of vendors on the market with their own policies for licensing, updating products, and withdrawing old solutions from circulation.

    In almost any company, you can find software that pays for support, although it is not really needed or, moreover, the system has already been withdrawn from support. It also happens the other way around: the company plans to build a new system, not realizing that the platform used is very outdated.

    For IT and business

    Who might be interested? If there is such a position in the company as a corporate architect, then he will become the main stakeholder.

    Typically, drawing programs are used to describe the architecture - for example, Microsoft Visio. It is important to note that Enterprise Maps is not about drawing, but about creating models related to the real world, and various policies will be monitored during their development. It is impossible to create an object divorced from reality.

    Architects are an audience that understands the advantages of the system and is able to appreciate its capabilities. However, unfortunately, they most often do not manage budgets, and decisions will be made higher.

    The next important user group is CFOs and CIOs. They have other tasks. Let's say financial director may not understand too much what is happening in IT, but it is important for him to see where the investments are going and how high a priority this area is in terms of strategy. He will receive information about this not in the form of a report edited by employees, but with the help of a real slice of information. For the CIO, the transparency of the activities of the IT department and the additional justification of their decisions to top management are important.

    Of course, I am interested in solutions of this class big business, who has reached a certain maturity - people who understand that it is necessary to put things in order in the economy. Most importantly, Enterprise Maps, unlike heavier solutions that require a fundamental approach, offers a “quick wins” path: quickly implement a separate scenario to prove itself, and then gradually increase the power of the solution.

    For example, the script “Five steps to bring an application to the cloud” is “hardwired” in the system. Having such ready-made scenarios is very important for the implementation of quick projects that show the business the value of the solution.

    Not only reporting

    There are three typical areas covered by Enterprise Maps. The first is to combine the available data (architecture models, business strategies) in a single repository, where all elements are related to the corporate architecture, identify people responsible for information slices, and set up synchronization with various data sources.

    Second, seeing current situation, you can build a transition plan to the target state by identifying the missing elements using Gap analysis. This is the key difference between the solution and configuration management databases (UCMDB), which reflect only the current situation and only at the infrastructure level. The third important direction is the standardization of the approach to IT assets, systems and functions.

    At the output of the system, the company receives slices of information for various users. Each of them will see the information in which they are interested. With Enterprise Maps, you can answer the most various questions: for example, it will be possible to find out how much the maintenance of this or that application costs, whether the investments correspond to the strategy, where the funds go, what can be done with the allocated budget.

    Usually, when they see the possibilities of Enterprise Maps, many people ask: is this a reporting system? Yes, the resemblance is great. But the solution is not only responsible for presenting data in a beautiful graphical form, but also becomes the main working tool of the architect - a means of creating and maintaining enterprise architecture models.
    It is important that inside the system it is possible to lay the policies for the development of the architecture (both the infrastructure part and the applications) that will be monitored. It will not be possible to perform prohibited actions, and further compliance with the rules will be rechecked. Thus, the architect does not just “draw pictures”, but quite deliberately approaches the solution of problems.

    Data sources

    In order for the system to be operational, it must be filled with data. This can easily be done with Excel tables and files, and this is enough for most customers - this is the form in which information about corporate architecture is usually stored. More advanced customers have UCMDB systems, which also become a valuable source of data.

    Another important source is the project portfolio management system. This is where goals, programs, projects come from. Based on these data, it is possible to show the place of ongoing projects in the overall strategy.

    Finally, to fill the systems, one cannot do without modeling tools - for example, Sparx Enterprise Architect, one of the most popular systems due to its low cost. Moreover, in some cases, the use of such specialized tools is preferable. If a new system is being developed that becomes a major architectural element, it is better to take design tools designed for this and familiar to users, and then upload the built models to Enterprise Maps, where they will be associated with current systems, infrastructure, plans and project activities.

    Objective picture

    An important part of the system are built-in reports - slices of information that become a valuable source of knowledge when making decisions about the development of IT. Among them are several key ones.

    One of them is the Application Portfolio Business Opportunity report. Based on his data, one can tell where the company should invest in accordance with its strategy, highlighting strategically important decisions, or vice versa - identify candidates for the transition to the cloud. The next step is to identify critical business applications that support key processes and therefore require special treatment.

    The Strategic Investments report allows you to see the gaps between real investments and business priorities. For example, based on financial information from a project portfolio management system, excessive investment in non-priority areas can be identified. This gives the financial transparency that businesses want so much. An overview of the cost of applications also provides food for thought about their business value and actual costs.

    "Platform Usage" is a report that can also be very helpful. It allows you to understand, for example, that a third of applications use an unsupported platform, and if you plan to introduce a new system, then you should think about it.

    One of the most important features is the creation of a "dependency tree" showing the relationship between specific infrastructure elements and business services. It is not uncommon for a business-relevant system to be supported by a single server, which reduces its fault tolerance, or vice versa - a non-critical application runs on unreasonably expensive hardware.

    Tidying tool

    Among the companies most interested in enterprise architecture are quite mature financial institutions. The reasons are obvious: they constantly have to be “at the forefront of IT”, capturing new markets, which is reflected in the IT architecture. They are much more likely to need transparency, including in the process of achieving goals. Finally, in financial sector competition is fierce, and major IT errors can be costly.

    Do not think that enterprise architecture management systems are solutions exclusively for giants. If you can show the business that there are platforms that you can no longer pay for, or business requirements that are not given enough attention, the result can be very impressive even at a small company size.

    Nevertheless, a reasonable limit on the size of the company still exists. Minimum purchase- 10 licenses, that is, the organization must have at least 10 people who are interested. A few architects is already a fairly large organization.

    But the main thing is that the company must be ready, having bought a lot of equipment and applications in recent years, to put things in order in its IT economy. There are few of them so far.