DATA Intertech s.r.o

 Terms of Reference

Enter the project was to create a virtual infrastructure on VMware to run virtual servers. In the implementation should be maximized resources existing IT infrastructure. Along with creating a platform for virtual servers was implemented disaster recovery system to move these virtual servers to a backup location.

In the first phase of the project is expected to deploy about 20 virtual servers running MS Windows 2008 R2. There is a large requirement for further expansion of this number.

Problem definition

The problem which had to solve was obsolete hardware infrastructure and the need for server consolidation because of low levels of spatial capacity of the server room. The second problem was the impossibility of continuous testing (for reasons of audits) disaster recovery plans to move to a backup location.

 Solution

The solution was to deploy VMware vSphere 5, which created a stable platform for operation of consolidated virtual servers with central management. The existing physical servers were transferred to a new virtual servers were already installed directly as virtual servers.

  Used hardware

The solution was based on reliable components from IBM. Specifically used IBM x3850 servers X5, which were specifically designed for the needs of virtualization, mainly due to their potential capacity expansion. Implementation of a new SAN infrastructure was carried out on Fibre Channel, used IBM FC switches and storage subsystems DS3524, which allowed replication of data between them, which was necessary for the functionality of the backup location. The connection to the backup site was carried out by two independent FC and two independent lines, Ethernet lines.

  Application

The main pillar of the entire software infrastructure are products from VMware vSphere 5 family. First of ESXi servers, as hypervisors installed on each physical machine and also VMware vCenter Server for central management of the entire infrastructure. To ensure the migration to the backup site was deployed VMware Site Recovery Manager, which allows creating automated testable and recovery plans.

Other equipment can include software license MS Windows 2008 R2, for which he was elected Datacenter licensing model that allows you to create unlimited number of virtual machines over the zalicencovanou infrastructure.

  Setting up the infrastructure

The actual setup and test the functionality of the solution was already pending on the spot with the client. The first step was to set the hardware, especially SAN arrays and replication between the primary and backup site. It was also carried out a complete set vSphere infrastructure, created a basic template for the deployment of additional Windows servers and disaster recovery plans for the rehabilitation of infrastructure in the backup location.

 Problems in implementation

The problem that emerged during the implementation and caused problems, latency was different between FC lines to a backup location. This difference results in latency caused problems in data replication. The problem was eventually resolved by the provider of connections, so that FC optimize other line.

 Current status

Current status of the project were 3 physical servers with installed ESXi hypervisor and disk storage at the primary location and one ESXi backup and storage in the locality. Network and data link was implemented fully redundant.The main benefits are in addition to saving energy, time administrators and maintenance costs for hardware virtualization brings include:

 

• Increase application availability
• Faster backup and recovery
• Centralized management environment
• Greater flexibility in the allocation of resources
• Improved disaster recovery plans

Further development

In the near future, virtualization will allow for more physical servers and should be a gradual replacement of all physical servers as virtual.

To improve the management options in the backup location and increase the capacity of the backup location is calculated by adding an additional server and creating a HA cluster for high availability and live migration and backup location.

 Overall evaluation of the project

The project is evaluated as successful. Budget or deadlines are not exceeded and, with minor problems in the implementation of the entire implementation ran smoothly. Further collaborate with the client, especially in providing support or consultation on further development.