Implementation of server virtualization on JSC Navashinsky Plant of Building Materials

Task: Optimization of a server framework

Solution: Creation of a failover cluster of hypervisors, migration of crucial IT services in virtual environment and implementation of management tools the virtualized infrastructure.

"Implementation of technology of virtualization of Hyper-V and Microsoft System Center Virtual Machine Manager 2008 allowed us to solve the main task – to consolidate and optimize IT - infrastructure, having at the same time increased its reliability, scalability and flexibility necessary in the conditions of constantly growing business": Shustov Konstantin Vasilyevich, Chief of data-processing center, JSC Navashinsky Plant of Stroymateria

Existing at the time of the beginning of an IT project – infrastructure of the Customer did not satisfy to required characteristics. First of all, the fault tolerance of business processes of the enterprise was not provided and there was no centralized system of data storage. For this reason the decision on upgrade of an information system was made.

The booked audit of a server framework of JSC Navashinskogo zavoda stroymaterialov revealed the following weak points:

• Lack of technology of fault tolerance. In infrastructure of the Customer there were no technologies of reservation, both a hardware, and program component of a server framework that in turn at failure of server components could bring to long idle times in rendering IT services and to considerable material losses.
• Use in IT - infrastructure of an obsolete equipment with the expired guarantee period and supports by the producer. So for the organization of work of crucial applications the server of the Kraftway trademark was used. This server is constructed based on the outdated Intel platform. Now spare parts to it are not made.
• Consolidation in one operating system of several IT services at a high load on them.

Proceeding from the above shortcomings of infrastructure, the main objectives when implementing the project were:

• Increase in fault tolerance of a server framework;
• Failure from use of the equipment which was outdate and not covered with a guarantee, and also further reduction of the used equipment.
• Service migration of the Active Directory on the new platform.
• Increase in effective management of information infrastructure by means of implementation of the intellectual software;
• Creation of a basis for implementation of the modern systems of collective work and information exchange (informational portals, mail systems, means of document flow and so forth)

For achievement of these tasks, specialists "Sonet of NN" and JSC Navashinsky Plant of Building Materials made the decision on creation of a failover cluster of hypervisors, migrations of all services in virtual environment and implementation of instruments of monitoring.

In the project progress the following tasks were carried out:

For project implementation the storage system of HP of StorageWorks HP P2000 G3 SAS was used. Use of this storage allowed to consolidate information resources of the enterprise in one highly available reliable place. In storage 12 disks of SAS 10000k of 146 Gb SFF are set.

Acts as a basis of a server framework fault-tolerant two - a nodal cluster. As nodes of a cluster were used a HP server of DL 360 G7. These servers conform to the most strict requirements to fault tolerance and performance. Therefore their application as nodes of a cluster is justified.

As a program component of the solution for the organization of a cluster of hypervisors Microsoft Windows Server 2008 R2 Enterprise is used. As the basis of infrastructure is formed by a cluster from the Hyper-V nodes, among improvements of the operating system it is possible to note separately:

• improvements of a hypervisor of virtual machines;
• support of live migration by architecture of the general volumes of a cluster;
• increase in fault tolerance of connections of nodes of a cluster;
• advanced testing aid of clusters.

For fault protection of the server hardware and the software two nodal cluster of hypervisors are used. At design of a cluster the redundancy of all components of a cluster was provided.

At the enterprise the Active Directory directory service based on Windows Server Small Business Server 2003 was unrolled, but in connection with its restrictions in the project progress Windows Server needed to migrate a directory service on full. During the course of performance the project the directory service was postponed for the new Windows Server platform 2008 R2, and the server under control of Small Business Server 2003 was taken out of service.

In implementation process of the project for the DBMS system the separate virutalny machine was selected.

During the course of performance the project in a server framework were implemented a management system and monitoring based on the products Microsoft System Center. For placement of management tools the separate virutalny machine working running the Windows server 2008 R2 operating system was created.

In the project progress basic backup strategy which use allows to protect IT services from logical errors and also from the errors caused by a human factor was developed and implemented. Further it is going to upgrade a backup system, by implementation of the specialized software allowing to make the bigger number of backup copies, to archive data at the level of services and also to save disk space due to technologies of deduplication.

Result

• The efficiency of the project is defined with many factors, such as reliability, convenience of management, housekeeping overheads, high-speed performance and security.
• In terms of security, services of new infrastructure are more protected from compatibility issues, and all infrastructure in general is more protected from harmful actions.
• High reliability of new infrastructure is caused use of technologies of fault tolerance and duplication of principal components of infrastructure.
• Response time of the main services did not increase in comparison with old infrastructure, and there were also free computing resources for implementation of new services in the future. It is reached thanks to modern high-end servers and the storage area network using the latest version of the protocol of SAS.
• The controllability of infrastructure increased thanks to use of one server platform for support of all information services, and also thanks to use of instruments of monitoring and management.

The updated infrastructure is more adaptive and intellectual, and also answers recent trends in creation of data processing centers.