Abstract

In a critical server losing its input voltage/power, the server goes down after 10/20ms of hold-up time. In these hold-up time the data need to be transferred from volatile memory to non-volatile memory (NVDIMM). For years this process is being done successfully, by transferring the data during this 10/20ms hold-up time. However, this process is having the disadvantage of the server going down for each and every event of input power loss using NVDIMMs platforms.

To overcome this disadvantage, a new method or technique is proposed in this publication/disclosure (Fig 1). The proposed technique utilizes the signal (called bulk voltage). Whenever, the energy levels (called bulk voltage) in a power supply becomes low, the server starts transferring the data (compared to each loss of input power earlier) to non-volatile memory reducing the number of shut downs of the server. If the input power loss duration is from 10ms to 80ms depending upon the load on the server, the shutdown of the server can be avoided completely.

The cost of this feature is miniscule compared to the cost of adding UPS/mega cell/Super cap in each and every server/platform.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS