Browse Prior Art Database

Method for Optimal Battery Charging for a Multi-Battery Computer System under Power-Loss Hazard

IP.com Disclosure Number: IPCOM000246107D
Publication Date: 2016-May-08
Document File: 7 page(s) / 122K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for the optimal scheduling of battery charge-up in a multi-battery computer system under electric power budget constraint. The algorithm determines which batteries to charge at any one point, keeping the system within the total power budget at all times.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 43% of the total text.

Page 01 of 7

Method for Optimal Battery Charging for a Multi-Battery Computer System under Power-Loss Hazard
Critical computer systems must be protected against unexpected power loss, low power, or intermittent power (i.e., the main power is expected to return soon). Usually, these systems store important data in the system cache and write it to non-volatile storage in case of unexpected power loss (i.e., cache vaulting). Utilizing an uninterruptible power supply (UPS) is one way to supply electric power to facilitate cache vaulting. UPS systems provide electric power to a computer system's enclosures (a multitude of enclosures per system) in case of power loss. In addition, in order to avoid shutting down the system and a loss- of-access event, an alternative power source can keep the system running for a relatively short time (i.e., a few minutes). This usage is sometimes referred to as "riding through".

More recently, with the advancement of Lithium Ion battery technology, systems are increasingly using battery backup units (BBUs) for cache vaulting. Each enclosure includes a single or a pair (for redundancy) of BBUs. In case of power loss, the BBUs provide electric power to the enclosure perform cache vaulting to non-volatile storage (e.g., solid state drive (SSD), hard disk drive).

The new challenge becomes how to charge back this multitude of BBUs in a system that went through a power loss event. All BBUs are now discharged to different levels, and need to be charged back when power returns. During this charge-up period, because the computer system is disconnected from its host computer it cannot provide the services it is supposed to provide. The host computer is experiencing a loss of access period. The objective of the new methodology is to shorten the loss of access period as much as possible.

A system may include tens of enclosures and as twice as many BBUs. Ideally, all BBUs would simultaneously charge back at the fastest rate possible (to protect the system as soon as possible against further power loss events), but this is usually not the case. A total power budget constraint prevents the system from providing enough current to charge all

BBUs at the same time. This total power constraint value is determined by different factors such as heat dissipation constraints, maximum temperature constraints, and power grid outlet constraints.

There are no known algorithms for battery charge-up in constrained systems.

The solution is the most optimal method to charge back the BBUs under the constraint of a total power budget. Most optimal here is defined by the shortest time possible to bring all batteries to a pre-defined threshold charge level.

The current disclosure does not cover the algorithm of determining the total power budget from other constraints

1


Page 02 of 7

mentioned above.

Herein, the terms BBU and battery are interchangeable.

The novel contribution is a method for the optimal scheduling of battery charge-up in a multi-batte...