Search

We present an architecture and algorithms for performing automated software problem determination using call-stack matching. In an environment where software is used by a large user community, the same problem may re-occur many times. We show that this can be detected by matching the program call-stack against a historical database of call-stacks, so that as soon as the problem has been resolved once, future cases of the same or similar problems can be automatically resolved. This would greatly reduce the number of cases that need to be dealt with by human support analysts. We also show how a call-stack matching algorithm can be automatically learned from a small sample of call-stacks labeled by human analysts, and examine the performance of this learning algorithm on two different data sets.

Defect causal analysis offers a simple, low-cost method for systematically improving the quality of software produced by a team, project, or organization. DCA takes advantage of one of the most widely available types of quality information, the software problem report. This information drives a team-based technique for defect causal analysis. The analysis leads to process changes that help prevent defects and ensure their early detection. The paper discusses the three principles that drive the DCA approach. It considers the impact of DCA on an organization.

A logic error in the software of Cosmos II results in apparent malsensing of atrial signals at high rather than low sensitivity settings. The paradoxical function was found in 54 out of 87 units tested (= 62%) if the atrial sensitivity was increased and in 54 out of 87 units tested (= 62%) if the atrial sensitivity was increased and the ventricular refractory period was lengthened beyond critical limits that could be defined individually. The problem inherent in both unipolar (66%) and bipolar systems (54%) is attributed to a summation of a physiological signal and internal noise, which has to be detected at the atrial level within a specific timing window to disable atrial sensing for the next pacing cycle. The malfunction virtually disappears if the ventricular refractory period does not exceed 210 msec. It is recommended, therefore, that Cosmos II pacemakers be left at their nominal ventricular refractory value (200 msec).

We describe the case of a dual chamber rate responsive pacemaker (Relay, model 294-03, Intermedics, Angleton, TX, USA) implanted in a 68-year-old male for sick sinus syndrome, which was not working properly when programmed in the DDIR mode, thus determining occasionally a sort of "VVI" pacing. However, the pacemaker performed well when programmed in the DDDR mode. We discovered that this was not a malfunction of a single device but rather a general behavior of this family of Intermedics dual chamber pacemakers (also not rate responsive), caused by a software problem.

ICD Software Error. Loss of ventricular output resulting from an unexpected software error in a dual chamber implantable cardioverter defibrillator (ICD) is reported. A 70-year-old man with a dual chamber ICD implanted for a history of cardiac arrest and infra-Hisian block presented with acute onset of dizziness. He was found to have loss of ventricular output due to an internal software problem. The problem was corrected by software reprogramming via the programmer. This malfunction exemplifies the potential ability to correct current-generation ICD software problems noninvasively, thus avoiding the need for replacement.

Due to an error caused by a homemade data-analysis program, on page 1875, Geoffrey Chang and his colleagues retract three Science papers and report that two papers in other journals also contain erroneous structures. (Read more.)

Network coverage is an important issue for wireless sensor network. It depends on several factors including sensing model that has been used to design the network model. The sensor nodes are prone to be non-functional because of noise, battery energy depletion, software and hardware problems. Hence, it is relevant to assess network coverage considering node failure as a parameter. In the literature, coverage has been extensively studied for Boolean and shadow-fading sensing models. However, Elfes sensing model has not been considered. In this study, the authors investigate the impact of sensing model and node failure on network coverage. Also, temporal variation of coverage has been investigated. A strategy for maintaining constant coverage is proposed.

The title of the article “software glitch threw off mortality estimates” (News of the Week, J. Kaiser, 14 June, p. [1945][1]) is misleading. There was no software glitch involved—the tool worked exactly as specified. The scientists overestimated the risks of fine particles in the air because they specified incorrect parameters for the algorithm. That the scientists misused their tool is not the fault of the tool. This is a blunder on the part of the scientists, not the software. [1]: /lookup/doi/10.1126/science.296.5575.1945

washington A software glitch has meant that the Near-Earth Asteroid Rendezvous (NEAR) spacecraft will have to wait another 13 months before the start of a planned orbit of the asteroid Eros, which was intended to begin next week.