Troubleshooting of router crashes (1)

IntroductionGenerally, "system crash" refers to the situation where the system detects unrecoverable errors and restarts.

Errors that cause crashes are usually detected by the processor hardware, which automatically generates special error processing code in the ROM monitor. ROM monitor can identify errors, print messages, save fault information, and restart the system.

Start ruleFor more information about file rules, seeCisco Technical Tips Conventions.

PrerequisitesThis article does not have any specific prerequisites.

Components usedThis document is not limited to any specific software or hardware version.

Obtain crash information

In the case of a router crash, the most important task is to collect as much information about the crash as possible before the router is manually reloaded or restarted. After manual reload or power-on is started, all other information about the crash will be lost except the information successfully saved in the crashinfo file. The following output shows information about the crash.

If you can obtainShow version, show stacks, and show contextOrShow tech supportCommand output to display potential problems and repair methods. To use this information, log on and enable it, but you must be a registered user.

Command	Description
Show version	The earliest command in Cisco IOS (r) software version 10.0.Show versionThe EXEC command displays the system hardware configuration, software version, name and source of the configuration file and software image, the running time of the router, and information about the system restart process. Important:If the router is reloaded after the crash (for example, power-on or restart or use the reload command), this information will be lost. Therefore, try to collect related information before reloading !!!
Show stacks	The first command that appears in Cisco IOS software version 10.0.Show stacksThe EXEC command is used to monitor the stack usage of programs and interrupt programs. When a router crashes, the show stacks command output is one of the most important sources of information. Important:If the router is reloaded after the crash (such as power-on restart or use)Reload command. Therefore, try to collect related information before reloading!
Show context	The first command that appears in Cisco IOS software version 10.3.Show contextThe EXEC command is used to display information stored in non-volatile RAM (NVRAM) in the event of an accident. The context information is specific to the processor and infrastructure, while the software version and running time information is not. The context information of different vro types varies accordingly.Show contextThe command displays the following information: Cause of system restart Stack record Software Version Signal Number, code, and route running time All registration content during crash
Show tech-support	It was first found in Cisco IOS software version 11.2. This command helps collect basic information about a vro when reporting issues. This command includes: Show version Show running-config Show stacks Show interface Show controller Show process cpu Show process memory Show buffers
Console log	If you are connected to the vro Control Panel during a crash, you will see the following information during the crash: *** System received a Software forced crash *** signal= 0x17, code= 0x24, context= 0x619978a0 PC = 0x602e59dc, Cause = 0x4020, Status Reg = 0x34008002 DCL Masked Interrupt Register = 0x000000f7 DCL Interrupt Value Register = 0x00000010 MEMD Int 6 Status Register = 0x00000000 Save the information and previous logs. Make sure to obtainShow stacksCommand output.
Syslog	If the router is set to send logs to the syslog server, you can display the events that occurred before the crash on the syslog server. However, when a router crashes, the most useful information may not be sent to the syslog server. Therefore, syslog output is usually not very useful for troubleshooting crash.
Crashinfo	The crashinfo file contains many useful information about the current crash, which is stored in bootflash or flash memory. When a router crashes due to data or stack corruptionShow stacksIn addition to command output, more overload information is required to eliminate such crash faults. Crashinfo is written toBootflash: crashinfo. For Cisco 7500 universal interface Processor 2 (VIP2), this file is savedBootflash:Vip2_slot_no _Crashinfo, Where slot_no is the VIP 2 slot number. For the Cisco 7000 Route Processor (RP), this file is savedFlash: crashinfo. For more information, seeObtain information from the Crashinfo File.
Core dump	Core dump is a full copy of the vro memory image. This information is not necessary to solve most types of crashes, but it is best to use this information when recording new failures. To add more information such as debug sanity, scheduler heapcheck process, and memory check-interval 1 to core dump, you may need to enable some debugging tools. For more information, seeCreate Core Dumps.
Rom monitor	If the config-register setting of the vro。 ends with 0, the vro may terminate in the ROM monitor after the crash. If the vro is 68 k, the prompt is "> ". You canKCommand to obtain the stack record. If the processor is a simplified instruction set computing Organization (PROTEUS), the prompt is "rommon 1> ". Obtain the output of stack 50 and show context.