The review is focused on the project in which we have transferred the IT-infrastructure of our partner-company system integrator ITSG. Customer’s requirements specification:
The company's staff consists of more than 40 employees, who have used two outdated physical servers. The client has relatively few basic corporate services: ERP, corporate email and file sharing server. In order to create virtual machines on a physical server, we have used a high-tech VMware ESXi hardware hypervisor. The resource pool in the cloud should consist of specialized powerful server hardware with a high workload. We have suggested the client to deploy 5 following virtual machines on the Windows Server 2016 platform:
Hardware configuration of the cloud-solution:
All the virtual machines are located in the same local network. Access to servers from the office is accomplished using a secure encrypted VPN tunnel which uses L2TP + IPsec technology. The tunnel connects the company's router (network router) and the cloud router. External connections are available with the help of client VPN connections (the same L2TP + IPsec).
The terminal server on the Windows Server 2016 platform is the user's working environment. It sets up access to corporate resources: ERP, files, and folders (according to access rights), mail-client and Microsoft office suite. To improve the runtime performance of ERP, we have transferred file databases to MS SQL and placed them on a separate virtual server with SSD disks. We have chosen Microsoft Exchange 2016 as a mail server to implement additional functionality (calendars, mail archiving, etc.). The entire server part is deployed on the Microsoft platform using an Active Directory domain controller. For corporate documents, we have used a separate virtual server that serves as a file server. We have created a folder structure and distributed rights for employees. The server back-up is made every night, without any downtime or runtime performance problems. Using the backup copies for 14 days, it is possible to restore data or the entire server of any period. Estimated server recovery time is up to 6 hours, depending on the amount of data. Availability of corporate resources has been achieved through the implementation of the server architecture best practice and its deployment in the cloud, built in a reliable West German data center of TIER III level, with the services’ availability 99.9%.
We selected the hardware configuration based on the results of the services’ usage intensity audit and possible scalability, as well as on our experience in implementing such projects. If it is necessary, you can easily increase or decrease hardware resources with minimal server downtime. On some servers (such as terminal and database servers), we have added SSD disks which have significantly improved runtime performance. The project has been implemented in 5 working days with a slight downtime of operating services. After we have completed the settings, the servers were transferred to the service of the client’s system administrator.
LIKE THE ARTICLE? SHARE VIA: