Data centre migration, the last several data center blackouts, caused the most severe blackouts. For example, the power outages at the Delta Airlines Inc data center caused the airline's $150 million loss and the United States' Super Bowl 'power outage caused the organizers to be embarrassed. The power cuts may affect anyone at any moment.
However, data center operators are facing difficulties is to determine the root cause of blackouts because of power outages are very complex, for example, the server is overloaded, causing the system to crash; supply local power supplier, the staff misuse of human factors etc..
Here are five of the most important issues that data center operators should focus on to ensure that data center blackouts do not occur.
(1) can it move with the rapid development of the power system?
At different stages of the data center, the demand for electricity is changing. The addition of a server or switch may have a major impact on the demand for electricity. Therefore, it is important to analyze the power requirements of data centers over a period of time in order to make long-term forecasts easier.
(2) whether the power chain is threatened?
More and more data centers are connected to the network, and besides the terminals and access points contained in the rack, many channels may become a destructive path to the destruction of cybercrime.
In addition, network violations do not even have to be implemented by wires and cables. Cunning criminals may enter a data center and destroy internal power supplies. However, not only external malicious personnel caused harm, but also to alert the data center staff behavior. Some knowledge and inexperienced staff make mistakes when interacting with interfaces and can cause immeasurable harm.
To prevent this, maintenance documentation and process control are critical. Adopting more hardware is not the best choice for preventing catastrophic outages, but in fact, adding additional hardware can actually make control worse.
(3) what is the disaster recovery plan? Did you record the document?
Power outage testing without affecting the business environment is actually the use of virtual switching devices, which will allow data center operators to respond to the worst cases and implement recovery services.
Data center staff always assume that their power supply chains and power backup systems are foolproof, but what would be the result if there were no fail safe tests?
Power failure simulation enables data center operators to find facilities that lack redundancy and detect single point of failure. However, this requires documentation.
As a result, data center operators set up a power fail test mechanism and record their recovery process before catastrophic power failures.
(4) can we monitor the operation in real time?
Data center operators must know where the devices are and how much power they use. This is often difficult for data centers with increasing infrastructure, as this can have a huge impact on power capacity and power distribution.
Note that the only valid method for all moving components is a single view. This overall view has the function of real-time monitoring and alerting, enabling data center operators to mitigate risks and make changes to avoid disaster.
沒有留言:
張貼留言