In last summers case, the first failure was a potential major component breakdown, which triggered what they call a Buchholz Alarm. When this happens, the system is left without backup circuits for around 10 minutes, which is considered an acceptable risk. It was during this period that a second fault occurred to another piece of equipment and the entire system went down.
You might say it was a case of bad luck, until you learn that there are 43,000 individual pieces of equipment such as the one in the second failure and that there are on average around 13 Buchholz Alarms each year!
For businesses across the globe, server downtime means loss of transactions, loss of revenue, potential legal implications (e.g., online trading), data corruption, and, ultimately, loss of customers.
If a customer goes to your Website and it is down, he will usually go to the next one that is working. The other problem is that even if the power drops for a second, it could take your server hours to reboot and successfully restore the data to the point at which the interruption occurred.
So what can be done? Well, businesses need to take this issue seriously. If it is just a single device, a UPS (Uninterruptible Power Supply) will give some autonomy for a limited period. But if more equipment, or large data systems, are being used, then diverse power supplies from opposite sides of the city would help, and local generators would be a major benefit.
Bob Friday, Chief AI Officer for Juniper Networks, explains how the advanced technology is transforming operations.
Contact center leaders from 8x8, Awaken Intelligence, and 360insight discuss the importance of agent experience.
AI may force enterprises to rewire parts of their data centers so they are fully optimized to run such workloads. The question is do you use Ethernet or InfiniBand?
