SibIntel created a server cluster of high readiness on the basis of the systems of Fujitsu

Problem and premises of the project

Owing to active development vertically the integrated SibIntel holding had a need for creation own highly productive IT infastruktury which would provide needs of all participants of holding for computing resources and also performed support of infrastructure IT services. Hardware resources of a new failover cluster were supposed to be used for such tasks as ensuring work of accounting of holding, the internal portal and file storage and also creation of infrastructure of virtual machines (VDI). Considering the round-the-clock operation mode of a part of divisions of holding, it was necessary to provide permanent availability of all IT services on demand.

As a result as a hardware basis of a computing cluster for IT infrastructure of JSC SibIntel the Fujitsu PRIMERGY RX200 servers and a storage system of Fujitsu ETERNUS DX80 were selected.

The used equipment

Fujitsu PRIMERGY RX200 S8 is the rack server providing high performance thanks to processors Intel Xeon E5, expansibility to 1536 GB of memory and energy efficiency up to 96%. The PRIMERGY RX200 S8 server is the ideal solution for virtualization, cloud computing, small databases and also for high-performance computing. It has the maximum capacity and scalability in the body 1U high saving the place counting upon future requirements.

Fujitsu ETERNUS DX80 is the universal disk platform for data storage with high degree of scalability expected the intensive growth of amount of data and ensuring the highest performance. The free choice of modular disk storage modules and the latest technologies for connection within all line of ETERNUS provides the maximum flexibility of configuring taking into account individual requirements of the customer to performance and total cost of ownership. Existence of reserve elements, the RAID functions and the built-in means of protecting of data provide the highest degree of reliability of a system which, in turn, guarantees maintenance of continuity of business processes.

System Selection

In selection process of the server hardware for creation of a failover cluster representatives of SibIntel company drew great attention to such criteria as reliability and non-failure operation of a ready system; a system scaling option "on the fly", trouble-free service maintenance on site installations and the minimum attention to the started system in use. Besides, the IT department of the enterprise was faced by a problem of transition to failsafe infrastructure which would have a considerable stock of building of resources (OZU and the CCP). The storage system was also selected taking into account a stock for significant increase in capacity and at cost ideally entered the planned budget.

According to the results of comparison of the equipment with models of the competing vendors a choice was made for benefit of servers and storage system of Fujitsu as they as much as possible meet all qualifying standards of customer company.

Solution

The ready hardware and software system for SibIntel consists of switches, six Fujitsu PRIMERGY RX200 S8 servers, a storage system of Fujitsu ETERNUS DX80 and the software for virtualization of the server Microsoft Hyper-V resources. Each subsystem of a complex executes a certain role:

• The server subsystem is a source of computing resources for infrastructure.
• The storage area network saves files of virtual machines and provides a pool of disk space for file storage.
• Software of virtualization of Hyper-V virtualizes computing resources of servers and creates virtual machines on their base.
• The subsystem of back supply including the UPS and specialized software which provide smooth functioning of a failover cluster.

The general scheme of interaction of subsystems of a cluster looks as follows. The storage system is connected to the Fujitsu PRIMERGY servers and on them the environment of virtualization (hypervisor) in which templates of future virtual machines are created is unrolled. Templates are stored in a storage system, on virtual machines the services of holding are deployed.

On the basis of systems of virtualization Microsoft Hyper-V, Microsoft Windows Server 2012 R2 OS included in structure creates the cluster of virtual machines placed on two servers. With the help of the DFS mechanism file access for the applications requiring protocols of CIFS/NFS is organized.

The failsafe complex of Fujitsu services all crucial business processes SibIntel and runs smoothly in the round-the-clock mode of operation.

The software products unrolled based on a server cluster of Fujitsu:

• Microsoft SharePoint
• Microsoft Hyper-V
• Microsoft System Center 2012 R2
• Autodesk AutoCAD
• 1C: Accounting

Project Results

During creation of a server cluster of high readiness on the basis of the server systems of Fujitsu the management of customer company remained completely enough project. It was succeeded to raise considerably indicators of continuity of work of infrastructure services and services, IT services were consolidated in a uniform pool of virtual machines that considerably simplified service and improved quality of providing services on demand for the staff of the companies which are a part of holding.

"As a result of project implementation of creation of a failover server cluster of Fujitsu at the enterprise high degree of fault tolerance of IT infrastructure, a possibility of fast deployment of virtual machines for new IT services of the company was provided and also the controllability and extent of scaling of infrastructure in general is significantly increased. We received the reliable managed failsafe system for the "correct" money with excellent opportunities of accumulation of computing power, flexible configuration and low ownership cost", – Alexander Shtachenko, the head of department of automation says SibIntel.

Further SibIntel is going to expand the existing infrastructure by increase in memory in servers and increases in disk space at DWH. This project is a sample for creation of failsafe infrastructure in remote branches and divisions of holding. The planned terms of implementation of new projects – 2015-2016.