Tracing is a function that logs internal events. Trace files are automatically created and saved to the tracelogs directory on the harddisk: file system on the chassis, which stores tracing files in bootflash:. Trace files are used to store tracing data.

Note

Starting release Cisco IOS XE Release 3.14 and later, logs are stored in compressed format. The logs in the bootflash are stored in compressed format with .gz file extension. Use the archiving tools such as gunzip, gzip, 7-zip to extract the files. If the sytem reloads unexpectedly, some of the files may not be in compressed format. Extraction of log files may lead to time hogs or CPU logs. We recommend to perform this by copying the files to the PC. Extraction of files cannot be performed at the IOS prompt. Log files not handled by the bootflash trace are not stored in the compressed format (for example, system_shell_R*.log ).

The contents of trace files are useful for the following purposes:

• Troubleshooting—If a chassis is having an issue, the trace file output may provide information that is useful for locating and solving the problem. Trace files can almost always be accessed through diagnostic mode even if other system issues are occurring.
• Debugging—The trace file outputs can help users get a more detailed view of system actions and operations.

The tracing function logs the contents of internal events on the chassis. Trace files with all trace output for a module are periodically created and updated and are stored in the tracelog directory. Trace files can be erased from this directory to recover space on the file system without impacting system performance.

The most recent trace information for a specific module can be viewed using the show platform software trace message privileged EXEC and diagnostic mode command. This command can be entered to gather trace log information even during an IOS failure because it is available in diagnostic mode.

Trace files can be copied to other destinations using most file transfer functions (such as FTP, TFTP, and so on) and opened using a plaintext editor.

Tracing cannot be disabled on the chassis. Trace levels, however, which set the message types that generate trace output, are user-configurable and can be set using the set platform software trace command. If a user wants to modify the trace level to increase or decrease the amount of trace message output, the user should set a new tracing level using the set platform software trace command. Trace levels can be set by process using the all-modules keyword within the set platform software trace command, or by module within a process. See the set platform software trace command reference for more information on this command, and the Tracing Levels of this document for additional information on tracing levels.

Tracing levels determine how much information about a module should be stored in the trace buffer or file.

Table 1 shows all of the trace levels that are available and provides descriptions of what types of messages are displayed with each tracing level.

Trace level settings are leveled, meaning that every setting will contain all messages from the lower setting plus the messages from its own setting. For instance, setting the trace level to 3(error) ensures that the trace file will contain all output for the 0 (emergencies), 1 (alerts), 2 (critical), and 3 (error) settings. Setting the trace level to 4 (warning) will ensure that all trace output for the specific module will be included in that trace file.

The default tracing level for every module on the chassis is notice.

All trace levels are not user-configurable. Specifically, the alert, critical, and notice tracing levels cannot be set by users. If you wish to trace these messages, set the trace level to a higher level that will collect these messages.

When setting trace levels, it is also important to remember that the setting is not done in a configuration mode, so trace level settings are returned to their defaults after every router reload.

Caution

Setting tracing of a module to the debug level or higher can have a negative performance impact. Setting tracing to this level or higher should be done with discretion.

Caution

Setting a large number of modules to high tracing levels can severely degrade performance. If a high level of tracing is needed in a specific context, it is almost always preferable to set a single module on a higher tracing level rather than setting multiple modules to high tracing levels.

By default, all modules on the chassis are set to notice. This setting will be maintained unless changed by a user.

To see the tracing level for any module on the chassis, enter the show platform software trace level command in privileged EXEC or diagnostic mode.

In the following example, the show platform software trace level command is used to view the tracing levels of the Forwarding Manager processes on the active RSP:

Router# show platform software trace level forwarding-manager rp active 
Module Name                     Trace Level      
-----------------------------------------------
acl                             Notice           
binos                           Notice           
binos/brand                     Notice           
bipc                            Notice           
bsignal                         Notice           
btrace                          Notice           
cce                             Notice           
cdllib                          Notice           
cef                             Notice           
chasfs                          Notice           
chasutil                        Notice           
erspan                          Notice           
ess                             Notice           
ether-channel                   Notice           
evlib                           Notice           
evutil                          Notice           
file_alloc                      Notice           
fman_rp                         Notice           
fpm                             Notice           
fw                              Notice           
icmp                            Notice           
interfaces                      Notice           
iosd                            Notice           
ipc                             Notice           
ipclog                          Notice           
iphc                            Notice           
ipsec                           Notice           
mgmte-acl                       Notice           
mlp                             Notice           
mqipc                           Notice           
nat                             Notice           
nbar                            Notice           
netflow                         Notice           
om                              Notice           
peer                            Notice           
qos                             Notice           
route-map                       Notice           
sbc                             Notice           
services                        Notice           
sw_wdog                         Notice           
tdl_acl_config_type             Notice           
tdl_acl_db_type                 Notice           
tdl_cdlcore_message             Notice           
tdl_cef_config_common_type      Notice           
tdl_cef_config_type             Notice           
tdl_dpidb_config_type           Notice           
tdl_fman_rp_comm_type           Notice           
tdl_fman_rp_message             Notice           
tdl_fw_config_type              Notice           
tdl_hapi_tdl_type               Notice           
tdl_icmp_type                   Notice           
tdl_ip_options_type             Notice           
tdl_ipc_ack_type                Notice           
tdl_ipsec_db_type               Notice           
tdl_mcp_comm_type               Notice           
tdl_mlp_config_type             Notice           
tdl_mlp_db_type                 Notice           
tdl_om_type                     Notice           
tdl_ui_message                  Notice           
tdl_ui_type                     Notice           
tdl_urpf_config_type            Notice           
tdllib                          Notice           
trans_avl                       Notice           
uihandler                       Notice           
uipeer                          Notice           
uistatus                        Notice           
urpf                            Notice           
vista                           Notice           
wccp                            Notice           

To set a tracing level for any module on the chassis, or for all modules within a process, enter the set platform software trace privileged EXEC and diagnostic mode command.

In the following example, the trace level for the ACL module in the Forwarding Manager of the ESP processor in slot 0 is set to info.

set platform software trace forwarding-manager F0 acl info

See the set platform software trace command reference for additional information about the options for this command.

To view the trace messages in the trace buffer or file, enter the show platform software trace message privileged EXEC and diagnostic mode command.

In the following example, the trace messages for the Host Manager process in Route Switch Processor slot 0 are viewed using the show platform software trace message command:

Router# show platform software trace message host-manager R0
08/23 12:09:14.408 [uipeer]: (info): Looking for a ui_req msg
08/23 12:09:14.408 [uipeer]: (info): Start of request handling for con 0x100a61c8
08/23 12:09:14.399 [uipeer]: (info): Accepted connection for 14 as 0x100a61c8
08/23 12:09:14.399 [uipeer]: (info): Received new connection 0x100a61c8 on descriptor 14
08/23 12:09:14.398 [uipeer]: (info): Accepting command connection on listen fd 7
08/23 11:53:57.440 [uipeer]: (info): Going to send a status update to the shell manager in slot 0
08/23 11:53:47.417 [uipeer]: (info): Going to send a status update to the shell manager in slot 0