Browse Prior Art Database

Method to provide dynamic partitioning to EXA blades Disclosure Number: IPCOM000099021D
Original Publication Date: 2005-Mar-09
Included in the Prior Art Database: 2005-Mar-09
Document File: 2 page(s) / 102K

Publishing Venue



This invention allows multiple blade servers that are part of a scalable system to be installed and removed from the BladeCenter independent of each other. This allows a scalable server to span any number of blade bays and allows for the blade servers to be reconfigured dynamically according to required needs. A workload manager could be installed in the Management Module or other device attached to the Management Module. This workload manager could monitor the BladeCenter's current workload and reconfigure the scalable systems to meed workload needs(scale up vs. scale out).

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 75% of the total text.

Page 1 of 2

Method to provide dynamic partitioning to EXA blades

     This invention consists of a BladeCenter, a blade server containing a scalable chipset and BMC, a scaling switch containing a BMC, and a management module that coordinates all the events. The components are connected as shown in figure 1.

The invention operates as follows:
1) The components are inserted into the BladeCenter and powered on.
2) The management module queries the network switches and blade servers via a sideband bus to determine their capabilities.
3) If the management module determines that a scaling switch and scalable blades are installed in the system, it groups the scalable blades and scaling switches into compatible groups. More than one type of scaling mechanism can be supported in the same BladeCenter at the same time.
4) The user, via the management module, configures the blade servers in a group into a single or multiple scalable systems. The user can also identify redundant blades or failover scenarios that would define how the group behaves in the event of a failure.
5) The management module connects to the BMC on each scaling switch via a


[This page contains 1 picture or other non-text object]

Page 2 of 2

sideband bus and transmits configuration information.
6) The scaling switch BMC instructs the scaling switch chip to configure its ports to allow communication between the defined scalable systems.
5) The management module connects to the BMC on each of the scalable blade servers via a sid...