Impact on infrastructure
During a recent blackout that lasted nearly five hours, the vshosting data center did not experience any technical complications. "The data center infrastructure continued to operate normally without any service restrictions," explains Martin Sláma.
The incident confirmed the correctness of the backup power supply architecture.
Backup power supply architecture
The vshosting data center uses multi-level power redundancy:
UPS systems: Vertiv APM modular units
Diesel generators: CAT and Visa
Distribution: Each supply has two PDU branches, each branch with double redundant backup.
"UPS constantly balances the voltage. This means that in the event of a slight drop or surge, the output voltage remains correct. We have DG in hot standby and they start up within 30 seconds."
Testing and maintenance protocols
Backup systems undergo regular testing four times a year. The testing procedure includes controlled power disconnection and validation of automatic procedures.
"During the stress test, we deliberately disconnect the power supply and monitor how the backup systems take over operation according to precisely defined scenarios. Every step is coordinated—our specialist is in each technical room with a radio, monitoring the test in real time and communicating with the control room. Regular testing of all critical systems allows us to verify that in the event of an actual outage, everything will run smoothly and without interruption of services."
“The generators ran as expected, and we monitored them continuously. Internal systems ran at full capacity without any restrictions.”
Fail-safe mechanisms and redundancy
In case of failure of the primary backup systems, there are additional levels of protection:
- Backup diesel generators in case of failure of the main units
- Automated procedures for various failure scenarios
- Escalation procedures, including manual intervention by technical personnel
"We have backup diesel generators. The automatic control system has programmed procedures for how to proceed in the event of the unavailability of certain elements, and last but not least, we have trained technical staff who are on call 24/7 at the data center and will immediately begin to resolve the situation."
Risk assessment and continuous improvement
"The best safeguard for these cases is a properly configured automatic system. If everything works as it should, there is nothing to worry about."
Historically, the data center has experienced occasional short-term outages lasting fractions of a second, and the systems always respond according to specifications. The July incident was exceptional in terms of its duration, not the complexity of the solution.
Customer impact and service continuity
During the blackout, all customer servers continued to operate without interruption. Customer support received only two inquiries regarding the autonomy of backup power sources and service availability during the outage. We also received praise from clients on social media that while other services failed (payment terminals, mobile networks, hospitals switched to backup power sources, and some industrial operations had to halt production, etc.), we continued to function, and with us, websites, applications, and e-shops.
"Thanks to ongoing measures, everything worked for our customers. And our services were fully operational."
Operations perspective
From an operational perspective, the incident proceeded as standard. After detecting the outage, the team validated the correct functioning of the automation and backup systems, monitored the cause, and estimated the duration of the outage.
"As soon as we detect an outage, we immediately check that the automatic systems have responded correctly and that the backup systems are working as they should, and we determine the cause and duration of the outage. The operation of the server room remained completely unaffected throughout the outage.
The alarm system worked as expected: "This is nothing unusual for us, and we are very well prepared for such situations. We were just as surprised as everyone else by the extent of the outage, but it had no impact on operations," adds Martin Sláma.
Lessons learned and conclusion
The incident confirmed the wisdom of investing in redundant infrastructure and automation. "Everything worked exactly as it should, and it confirmed that our team knows exactly how to behave in such situations," summarizes Martin Sláma, Head of HW/DC at vshosting.
The July blackout showed that backup systems designed for critical infrastructure work reliably in real-world operations. The combination of properly designed architecture, regular testing, and prepared personnel ensured service continuity even during an extensive power outage.
For these reasons, the data center has had 100% power availability without outages throughout its operation since opening in 2015—that is, for 10 years.
For organizations operating critical IT infrastructure, the incident serves as a case study of the importance of investing in redundant power systems and thorough preparation in the form of disaster recovery planning.