"Airplane" has built a fault-tolerant corporate IT infrastructure

2024: Creation of a fault-tolerant corporate IT infrastructure

Group "Airplane" transferred all corporate IT systems to a fault-tolerant model that guarantees operability and full data protection of losses in the event of an emergency. The infrastructure availability level is 99.95%. The company's information systems are reserved on the basis of two data centers (), DPC combined into a metro cluster. In total, these are more than 400 systems and applications that are used to ensure the daily operation of the Aircraft, and about 2.4 PB of data - financial data, and - HR CRM systems, by data construction process, information about clients. The Samolet Group announced this on April 3, 2024.

The project to create an infrastructure solution took one year. Since the specialists of the "aircraft" were faced with the task of creating a fault-tolerant system online, the choice was made in favor of building a metro cluster.

The first stage was design, data centers network infrastructure, infrastructure, data storage selection of technologies and equipment. The architecture of the metro cluster is implemented on the basis of solutions and, Hitachi Huawei virtualization made on the basis. The VMWare organization DSS used technology SAN that brought the solution of tasks, reservations data recovery and scaling to a consistently high level due to the high speed of data transfer and minimal loss of performance. In parallel with this, we worked out the placement of data centers in the constellation, which met our technical requirements, the main and backup channels between communications data centers and the central office were carried out. Both data centers are active, and their resources are available normally.

At the last stage, the capacities between the data centers were increased and equalized so that, if one second failed, it could ensure the normal functioning of the Aircraft's product services. During peak periods, the company will be able to maintain high availability of its services without increasing the waiting time of employees and customers. In the event of a failure, switching within the virtualization cluster takes several seconds, without data loss and, as a rule, unnoticed by the user. The reasons for the failure of the data center can be different - from an accident on a power line to a human factor, but a properly built infrastructure must ensure uninterrupted operation.

Another important point is the hybrid infrastructure, which is built so that capacity can be added as needed: within the reserve, they can be provided within an hour, new capacity can be organized within four hours.

The aircraft received a safe and fault-tolerant IT infrastructure. The basic value of infrastructure availability for us for April 2024 is 99.95, but certainly the goal is 99.99. For our projects, we select the best practices and solutions in the market.

For example, we use SAN technology, which provides simultaneous and synchronous copying of data; the possibility of cross-code use of capacities has been implemented, a mandatory backup of data for all production systems has been set up. The immediate plans also include the introduction of financial management practices in IT (ITFM), which will allow us to correctly assess the cost of services that we provide to internal users and business units, and effectively plan economic aspects, "said Alexey Semenov, head of the IT infrastructure and support services" Airplane. "