[ Pobierz całość w formacie PDF ]
.Another problemthat can prevent your machine from starting is corrupt files.A file system can becorrupted if it is not shut down properly; for example, if an administrator acciden-tally turns a server off without properly unmounting the file systems, they caneasily become corrupted.Files can also become corrupt in the event of a powerfailure because the system has no time to close opened files and shut down prop-erly.Another problem that can hinder your system boot is misconfigured softwareand hardware.For example, you may have made a mistake in the name of a filesystem in one of your boot configuration files, and the file system will not load.A wide variety of tools and utilities are available to help you troubleshoot and fixa boot problem.This chapter deals with several of these tools, including systemrescue boot disks, hard drive utilities, and system hardware and software statuscommands.Examining the Startup ProcessBetween the time you turn the machine on and the time you get to login, the bootprocess goes through many different steps and stages to get to that point.Eachstep along the way can break down because of a hardware or software problem,which will prevent the system from booting at all.A good theoretical knowledgeof the Linux boot process is essential when troubleshooting boot problems.Boot process stepsThe following is a list of events that transpire from the moment you start to bootthe Linux system.1.BIOS: The system s BIOS (Basic Input Output System) initializes and performsinternal checks on the hardware components of the system, including the CPU,memory, hard drives, and video system.If an error occurs with a component,the system will emit audible beeps and halt the process during initialization.2.MBR: The system checks the MBR (Master Boot Record) of the boot harddrive, which loads into memory the boot manager for the system.In thiscase, the boot manager is LILO.3.LILO: LILO stands for LInux LOader.LILO is a Linux boot manager, which isresponsible for loading the Linux kernel and system into memory.LILO canalso boot the computer into different operating systems if they are running 244881-6 ch17.F 11/12/01 8:31 AM Page 510Part VI &' Troubleshooting and Maintaining System Hardware510on separate partitions from the Linux system, such as a dual boot system.Forexample, you may have Linux and Windows 98 operating on the same system.LILO tells the system the location of the kernel image and begins to load it.4.Boot prompt: After LILO has finished loading, you are given the followingprompt:boot:Press the tab key to see a list of operating systems that have been configuredwith LILO.Choose Linux (it may be the default) to enable the boot loader tostart loading the kernel image file specified in the /etc/lilo.conffile.5.Kernel initializes: At this point, the kernel is uncompressed and loaded intomemory.The kernel then initializes the processor, memory, and console andconsole settings.From there, the kernel starts the disk bus subsystems, net-working is initialized, and hardware device drivers are loaded.6.init process: After hardware and device initialization has been completed,the kernel spawns the first Linux process, which is init.The initprocess isthe father of all processes in a Linux system.It reads your startup configura-tion file in /etc/inittab, and brings the system up to the appropriaterunlevel.7.Startup scripts execute: As part of the init process initialization, the appropri-ate scripts for the designated runlevel are run.These scripts start the neces-sary processes and services for the runlevel, mount your file systems, andinitiate the console sessions.8.Login: At this point, your system is completely loaded.To begin working onthe system, enter your username and password to gain access.For the exam, know the general boot-up process and what happens at each step.Analyzing Boot Process Errors6.11 Take appropriate action on boot errors (e.g., LILO, bootstrap)In order to fully understand problems with the boot process, you must be able toanalyze the existing data up to the point where the boot process halted.By care-fully analyzing error messages and log files of your boot process, you can quicklynarrow down the point of failure.Interpreting boot errors, warnings, and informa-tional messages is an extremely important aspect of troubleshooting, that allowsyou to reduce the time it takes to solve a problem.Common Boot ProblemsAfter you know the steps in the boot-up process, you should examine some of thecommon problems that are related to each step. 244881-6 ch17.F 11/12/01 8:31 AM Page 511Chapter 17 &' Troubleshooting the Boot Process511HardwareAny sort of hardware problem will most likely occur during the BIOS initializationprocess because the BIOS must examine these systems before loading the operat-ing system.This is called the POST (Power-On Self Test).The POST routing exam-ines all internal and external hardware for faults and proper configuration.Thefollowing is a list of common components that may fail, along with the error mes-sages or warnings that you may encounter.&' Motherboard: Typically, if you boot up your system and get no responseat all except for a number of beeps from the systemboard, then you have ahardware problem with the systemboard circuitry.The beeps from the sys-temboard are in a sequence, which you can look up in the manual to deter-mine which component including the video subsystem, cache, memory,or CPU may be faulty.&' Memory: The POST routine performs checksum tests on the memory, and ifany part of the memory addresses fail, it may halt the boot process altogether.&' Peripherals: The system checks for the presence of a keyboard, mouse, andany other input devices.These devices aren t needed for full operation especially if you can use the network to connect to the machine or serialport but sometimes a faulty keyboard or mouse can lock up a system.&' Hard Drive subsystems: The system examines the hard drives and disk drivecontrollers, which may be SCSI or IDE.The system checks that they are usingproper addresses that don t conflict with each other.&' Other Cards and Peripherals: At this time, any other cards and peripheralsyou have may be tested, such as SCSI cards, or network cards.Any type of IRQor IO address conflict may render the device inoperable, and these can bedetected in the POST processSoftwareSoftware processes take over from the hardware checking routines after the BIOShas examined the master boot record (MBR) of the machine for a boot loader.Thefollowing sections explore some of the most common problems that can occur atboot time.MBRThe BIOS loads the master boot record of the first boot hard drive.If it can t find ahard drive, an Operating system not present error results.If there is a drive thatcontains Linux, BIOS loads LILO (the Linux Loader boot manager).If your MBR hasbecome corrupt as is often the case if you happen to get a virus from a disketteor Windows partition you can reset it by booting from a DOS disk and using thefdisk/mbr to wipe the master boot record.Then, you must reinstall LILO by bootinginto your system with a boot disk, mounting your root partition, and running LILOagain. 244881-6 ch17
[ Pobierz całość w formacie PDF ]