Collecting Linux Log Files
Scenario
Collect Linux log files of a server.
NOTICE:
Do not directly decompress the InfoCollect_Linux.tar.gz tool package of the Linux version
on a Windows OS. Otherwise, the tool cannot collect logs because the lack of execution
permission.
Prerequisites
Conditions
A Huawei V2, V3 or V5 server (excluding KunLun server) is available.
The InfoCollect_Linux.tar.gz file has been uploaded to the target server.
You have logged in to Linux as a user with an ID of 0, for example, root.
You have backed up key data, and checked the CPU usage, memory usage, and
available disk space to ensure that InfoCollect is operating properly.
The IP address of the service network port on the target server must be pinged
NOTE:
Run the whoami and id commands on the OS. In Figure 4-10, the current user is root and
its user ID is 0.
Figure 4-10 Querying user information
Procedure
1. Confirm that the InfoCollect_Linux.tar.gz file has been uploaded to the server Linux
OS. For details, see Uploading the Software Package.
2. Run the following command to go to the directory where
the InfoCollect_Linux.tar.gz file is located, as shown in Figure 4-11.
Figure 4-11 Going to the directory where the file is located
3. Run the following command to decompress the collection tool package, as shown
in Figure 4-12 and Figure 4-13.
tar -zxvf InfoCollect_Linux.tar.gz
Figure 4-12 Decompressing the collection tool package
Figure 4-13 Decompressing the collection tool package
4. Run the following command to go to the InfoCollect_Linux directory, as shown
in Figure 4-14.
cd InfoCollect_Linux
Figure 4-14 Going to the InfoCollect_Linux directory
5. Run the following command to grant the operation permission on all files in
the InfoCollect_Linux directory to the user:
chmod +x *
6. (Optional) Run the following command to modify the config.ini file to specify the log
files to be collected, as shown in Figure 4-15.
Figure 4-15 Modifying the config.ini file
NOTE:
You are advised to check the content of the log files to be collected. You can edit
the flag column to determine whether a log file is to be collected. yes indicates that the
log file is to be collected. noindicates that the log file is not to be collected. By default,
all log files are to be collected.
7. Run the following command to collect log files, as shown in Figure 4-16.
./infoCollect.sh
Figure 4-16 Executing the Linux collection script
NOTE:
The collection process will last several minutes.
The log collection is successful if the message shown in Figure 4-17 is displayed.
Figure 4-17 Message indicating that the log collection is successful
NOTE:
If the information shown in Figure 4-18 is displayed after log file collection is
complete, manually collect crash files to your local computer.
Crash files are large. Manually collect crash files for fault diagnosis if necessary.
Figure 4-18 Crash file information
8. Confirm that all log files are collected based on the content of the filelist.txt file
in Figure 4-19.
You can filter the collected log files and then compress them into a package.
Figure 4-19 Content of the filelist.txt file
NOTE:
You are advised to compress log files before sending them to Support for analysis.
The maximum log file collection duration of InfoCollect is 20 minutes by default.
