A data center attack does not necessarily mean data loss or a long-term outage for users. Backup sites, data replication, and a prepared disaster recovery plan are key. Serhiy Kolesnichenko, COO of Ukrainian cloud operator Ucloud, told what can happen to a data center after an attack and what determines how quickly digital services will return to work.
A data center attack does not necessarily mean data loss or a long-term outage for users. Backup sites, data replication, and a prepared disaster recovery plan are key. Serhiy Kolesnichenko, COO of Ukrainian cloud operator Ucloud, told what can happen to a data center after an attack and what determines how quickly digital services will return to work.
The consequences of a strike depend on the location and strength of the impact. It can be the physical destruction of servers and network equipment, damage to power and cooling systems, disruption of communication channels, or fire.
According to Kolesnychenko, quoted by KP.UA, even if the server room is not directly affected by the fire, the equipment may fail due to combustion products. Soot gets inside the equipment and makes it unsuitable for further use. The worst-case scenario is the complete loss of equipment at a specific site.
However, a power outage in itself is not a critical scenario for a professional data center. Such facilities have backup power lines, UPS, and generators. With fuel and a functioning engineering infrastructure, the data center can continue to operate even during a prolonged power outage.
In a data center, according to Kolesnychenko, servers are only part of the system. Cooling, fire extinguishing, physical security, monitoring, and communication systems also operate there. They are the ones that ensure the continuous operation of the IT infrastructure.
Physically, data on a destroyed site may be lost. But in a properly built cloud infrastructure, it doesn’t have to exist in just one physical location.
For example, Ucloud reserves the site’s infrastructure in other locations. This allows, in the event of the loss of one data center, to restore client services in another.
After losing one of its sites, Ucloud did just that: the company restored infrastructure and client services at another location within 24 hours. The systems were launched in stages according to a defined priority.
According to Kolesnichenko, the speed of recovery depends on whether the company has backups, a second site, data replication, an alternative data transmission route, and a pre-developed Disaster Recovery Plan.
For the average user, the consequences may appear as the unavailability of familiar digital services: websites, mobile applications, banking transactions, online stores, logistics systems, or government services.
There is no universal scenario for which service “goes down first.” If the servers of a particular system are physically destroyed and it has no backup site, it will become inaccessible, regardless of whether it is a bank, taxi, online store, or other digital product.
In fact, the work of almost all business sectors today depends on data centers — from banks and retail to logistics, telecommunications, medicine, manufacturing, and government institutions.
The main principle is not to keep all the infrastructure in one place. It is about geographical redundancy, infrastructure dispersion and, where possible, placing key nodes in protected or underground premises.
Kolesnichenko also believes that the state could help the market with access to suitable underground structures owned by state and municipal authorities for the placement of critical IT infrastructure.
But even a backup infrastructure is no substitute for backups. If data exists in only one copy at one physical location, it can be lost with it. Therefore, backups should be a basic part of data protection — for both businesses and consumers.
