Browse Prior Art Database

Verification of device connections for redundant system controllers

IP.com Disclosure Number: IPCOM000166229D
Original Publication Date: 2008-Jan-05
Included in the Prior Art Database: 2008-Jan-05
Document File: 4 page(s) / 344K

Publishing Venue

IBM

Abstract

The use of a redundant service processor within a computer system provides the additional challenge of verifying the redundant service processor's connections to end devices within the system without adding significant time to a manufacturing or repair process. One method of performing this verification is to do a full initialization of the system using both service processors in the primary role, in effect testing both service processor's connections to the end devices. However, using both service processors to perform a full system initialization doubles the time required to complete the test. An alternative method is to provide the capability for both service processors to test their connections to an end device without having to perform initialization of the end device.

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 52% of the total text.

Page 1 of 4

Verification of device connections for redundant system controllers

The method used to select which service processor's signals reach or control the end device impact how the connection between a service processor and the end device can be verified. Described below are two methods of multiplexing the service processors' signals to the end device, and how connection verification can work within both methods. The verification methods associated with the second multiplexing method are preferred as they require the least amount of time to perform.

One method, as shown in figure 1, is to use a multiplexor (110) which takes in signals (150, 160) from each service processor (120, 130). The signal output (170) from the multiplexor (110) to the end device (140) is determined by a switch on the multiplexor. This way, either service processor (120, 130) can communicate with the end device (140), but never both at the same time. Going forward, this will be referred to as a transparent multiplexor because a service processor cannot verify its connection with the multiplexor, only with the end device. The end device in this example is connected via JTAG, but could be connect through some other interface such as RS485 or I2C. Because the multiplexor (110) is transparent to the service processors' (120, 130) signals (150, 160), verification of the connections between the service processors (120, 130) and the end device (140) requires that each service processor (120, 130) assume the primary role so that it may communicate with the end device (140). This switching of the primary role requires extra time, and depending on the structure of the firmware and capabilities of the end devices, may even require that the system be fully initialized with each service processor in the primary role.

To solve the problem of having to access the end device with both service processors in the primary role, one service processor is chosen as the primary, and it assesses the sibling service processor's ability to communicate with an end device by determining its ability to communicate with the intelligent multiplexor or to control the sibling's internal intelligent multiplexor to access the end device. Examples of this are demonstrated by figures 2, 3 and 4.

Figure 2 shows an alternative connection scheme with a support interface (270, 280) being used between the service processors (210, 220) and an intelligent multiplexor (230) which generates the appropriate signals (290) to the end device (240). The support interface (270, 280) can be used to send high level commands to the intelligent multiplexor...