Covering the latest industry trends and an excellent source of thought leadership.
No one likes to admit it, but data center outages or disruptions do happen sometimes. If you host your data with a provider, you need to know how it is being affected at all hours of the day, every day of the year. Do you know what goes on behind closed doors when something happens? Today, we’ll discuss some processes data centers follow in a power, network or equipment failure.
Power outage: When there is a power outage that impacts a data center, someone will be alerted to the problem, and whoever is onsite will conduct a walk-through of the data center floor, call the building manager to inform of incident, and if the center is on redundant power, wait for the generators to kick on. Many providers invest in redundant power to prevent exactly these disruptions during a power outage, and most clients may not even be aware that there is an outage unless their provider informs them their systems are running on backup power.
Network outage: A network outage, such as an Internet disruption or a DDOS attack, is treated differently. Depending on the nature and severity of the incident, an alarm is triggered, and a conference call about outage and details of incident(s) is had with the network and operations teams. Roles are then delegated accordingly. One person might call the internet service provider to report or monitor the incident, another might check the power systems to make sure everything is running smoothly. During this time, equipment will also be checked to determine if it is the root of the problem, if that hasn’t already been determined.
If the disruption is due to external factors such as an affected fiber that runs to the center, sometimes the data center can only wait for the ISP to fix whatever issue is on their end. In that case, the data center should provide a clear communication plan to clients to explain what is happening and keep them abreast of any updates.
However, no matter whether the cause is internal or external, clients expect to be online at all times and hold their data center responsible. Clear communication should be provided to clients so they are aware of any situations and kept informed of any updates.
Equipment failure can also affect a data center, and by extension, the client. If a disruption is caused by failure of a firewall or other network equipment, the process to fix the issue is two-fold. One engineer will document the problem and outline the changes that need to be made. A second engineer will then review the document and proposed changes, double check to make sure the proper equipment is being fixed, and then apply the changes. If a provider has already invested in redundancy, many of these issues will only have a minimum impact on clients rather than being potential disasters.
It’s never good, but data centers do experience downtime. Do you know what goes on behind the scenes with your data? Make sure your provider clearly communicates their process during an outage. If they don’t, perhaps it’s time to invest in someone new.