Browse Prior Art Database

Rack Level Automated Maintenance Machine Disclosure Number: IPCOM000240653D
Publication Date: 2015-Feb-16
Document File: 4 page(s) / 265K

Publishing Venue

The Prior Art Database


This disclosure describes a system for performing automated maintenance of a rack containing IT equipment and leverages technology from other industries to make the solution a cost optimized and feasbile system. The system incurs a limited development expense and up-front cost, but reduces labor expense, increases system reliability, and maximizes performance by replacing the need for more operators in the data center and maximizing IT equipment performance.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 21% of the total text.

Page 01 of 4

Rack Level Automated Maintenance Machine

Disclosed is a novel system and method for maintaining a rack of servers through the use of automation. A manipulator is constructed to operate in a 2 dimensional system embedded in the front or rear door of a server rack with the ability to traverse the entire frontal area of the rack. Examples of this system could be a gantry or a mechanical arm. When a serviceable component on the front or rear of the server requires service, the manipulator removes the defective part, places the defective part into a portion of the rack designated for waste, retrieves a replacement part from a portion of the rack stocked with new parts, and places the new component into the proper place in the server. In this way, server maintenance times are drastically reduced, the need for constant human presence in the data center is minimized, errors associated with humans are removed, and lower reliability (and cost) parts may be used with decreased harm.

Currently, there are only three real strategies employed for maintaining a rack of servers. The first is manual maintenance with human operators. The server needing a replacement part sends an error code to a central network control center, where the error code is read by an operator. Eventually, the operator will acquire a replacement part from a storage area, transport himself to the defective server and manually replace the defective component. The operator will then manually discard the defective component. The second strategy is to have a technician remove the server containing the defective part from the rack, insert a properly working server in its place, take the defective server to a bench or other area for maintenance by hand, then use the now operational server to replace the next server containing defective parts. The final key service strategy is called "fail in place", where a server is designed and built with enough redundancy to still be operational despite numerous component failures internal to the system. When/if enough failures occur, however, service is performed using one of the first two service strategies.

The inefficiency associated with replacing parts in servers is being exacerbated by several key issues. Datacenters are increasingly becoming more secure; companies are very worried about who has access to the data center. This is driving increased cost to hire only the best service technicians with the best records, longer service times due to decreased technicians allowed into the datacenter, and cost of clearing/overseeing warranty service technicians from the hardware vendor. Datacenters are also becoming exponentially larger; of the largest 15 data centers in the world almost all were built in the last 15 years, and the largest data centers are still under construction. An army of technicians is required to keep up with server maintenance in these facilities.

Most of the work that has been done to improve maintenance and part repla...