Service Procedures for Linux on IBM Power Systems Servers
An IBM Redpaper publication
Published 19 October 2020
ISBN-10: 0738459062
ISBN-13: 9780738459066
IBM Form #: REDP-5611-00
(28 pages)
View online
Authors: Scott Vetter, Luciano Chavez, Janani Janakiraman, Brahadambal Srinivasan
Abstract
Collecting data on first occurance of the problem can id in problem determination and timely resolution of defects. At IBM®, this process of collecting data on first occurance if often referred to as First Failure Data Capture (FFDC). Gathering this data before reporting a defect helps to understand the problem more quickly and thoroughly, which saves time analyzing data and reduces the time and mission affects in fixing defects.
Several diagnostic capabilities are built into the Linux operating system that enable you to determine the application level problems and system level problems. Collecting FFDC logs early, even before opening a defect report, helps to quickly determine whether:
- Symptoms match known problems (rediscovery)
- A report can be identified and resolved as a not-a-defect problem
- A workaround to reduce severity exists
Table of contents
Introduction
Configuring systems to collect logs
Firmware Assisted Dump
Defect submission guidelines
System crash and hang issues
Linux installation issues
DLPAR and LPM issues
Storage and I/O Issues
Networking issues
kdump issues
Virtualization issues
Conclusion
Firmware Assisted Dump
