journalctl-系统日志管理.md

journalctl command

The journalctl command is a command of the Systemd logging system. Its main purpose is to view the logs recorded by the Systemd logging system. Before the advent of Systemd, logs of the Linux system and various applications were managed separately. After Systemd replaced initd, it began to uniformly manage the boot logs of all Unit. With just one journalctl command, you can view the logs of all kernels and applications.

Syntax

journalctl [OPTIONS...] [MATCHES...]

Parameters

• --no-full, --full, -l: When a field matches an available column, it is omitted by default. The default setting is to display complete fields, allowing them to wrap or be truncated. The old option -l/--full is no longer useful except for canceling --no-full.
• -a, --all: Display all fields in full, even if they contain non-printable characters or are very long.
• -f, --follow: Display only the most recent log entries and continue printing as new entries are appended to the log.
• -e, --pager-end: Immediately jumps to the end of the log, implying -n 1000 to ensure that the pager does not buffer logs of unlimited size. You can use explicit -n and some other numerical values in the command line to override it. Note that this option only supports less page.
• -n, --lines=: Display the most recent log events and limit the number of events displayed. If used with --follow, this option is implicit. The parameter is a positive integer and is optional, with a default value of 10.
• --no-tail: Display all stored output lines, even in follow mode, thereby undoing the effect of --line=.
• -r, --reverse: Reverse the output to show the latest entry first.
• -o, --output=: Control the format of the displayed log entries, accepts one of the following options:
  • short: Default value, and generates output similar to the traditional syslog file format, with each log entry displayed on one line.
  • short-iso: Very similar to short, but displays an ISO 8601 wallclock timestamp.
  • short-precise: Very similar to short, but displays timestamps with microsecond precision.
  • short-monotonic: Similar, but displays monotonic timestamps instead of wallclock timestamps.
  • verbose: Display entries with the complete structure of all fields.
  • export: Serialize logs into a binary stream suitable for backup and network transmission, primarily text-based.
  • json: Format entries as a JSON data structure, one per line.
  • json-pretty: Format entries as a JSON data structure, but format them across multiple lines to make them easier to read.
  • json-sse: Format entries as a JSON data structure, but wrap them in a format suitable for server-sent Events.
  • cat: Generate a very concise output, only showing the actual message of each log entry, without metadata or even timestamps.
• -x, --catalog: Augment log lines with explanatory text from the message catalog, adding explanatory help text from available log messages to the output, which explains the context of errors or log events, possible solutions, and points to support forums, developer documentation, and any other relevant manuals. Note that help text is not available for all messages, but only for selected ones. Also, do not use -x when attaching journalctl output to error reports.
• -q, --quiet: When running as a regular user, suppress any warning messages about inaccessible system logs.
• -m, --merge: Display all available logs, including cross-entries from remote logs.
• -b [ID][±offset], --boot=[ID][±offset]: Display messages from a specific boot, matching for _BOOT_ID=, the parameter may be empty, in which case the logs from the current boot will be shown. If the boot ID is omitted, a positive offset will look backwards from the start of the logs, while a value equal to or less than zero will look back from the end of the logs. Thus, 1 denotes the first boot found in chronological order in the logs, 2 the second, and so on, while -0 denotes the last boot, -1 the boot before the last, and so on, with an empty offset equivalent to specifying -0, unless the current boot is not the last boot, for instance, because a --directory was specified to view logs from a different machine. If a 32-character ID is specified, an optional offset may follow, identifying a boot relative to the boot ID, with negative values representing earlier boots, and positive values representing later boots. If no offset is specified, the value is assumed to be zero, and the logs for the boot given by ID are shown.
• --list-boots: Display a list of boot numbers (relative to the current boot), their IDs, and the timestamps of the first and last message associated with the boot.
• -k, --dmesg: Only display kernel messages, meaning -b and adding a match for _TRANSPORT=kernel.
• -u, --unit=UNIT|PATTERN: Display messages for the specified systemd unit unit, or for any unit that matches PATTERN. If a pattern is specified, the list of unit names found in the logs is compared to the specified pattern, and all matches are picked. For each unit name, an additional match for messages originating from that unit will be added for _SYSTEMD_UNIT= unit, as well as extra matches for messages from systemd and coredumps about the specified unit. This parameter can be specified multiple times.
• --user-unit=: Display messages for the specified user session unit, adding matches for the messages from the unit with _SYSTEMD_USER_UNIT= and _UID=, and extra matches for messages from the systemd session and coredumps about the specified unit. This parameter can be specified multiple times.
• -p, --priority=: Filter the output according to the message priority or priority range, accepting a single number or a text log level (i.e., between 0 emerg and 7 debug), or a range of numeric/text log levels represented as ... Log levels are the usual syslog log levels found in syslog, namely emerg 0, alert 1, crit 2, err 3, warning 4, notice 5, info 6, debug 7. If a single log level is specified, all messages at or lower (values lower have higher priority) than that level will be shown. If a range is specified, all messages within that range, including the start and end values of the range, will be shown, adding a PRIORITY= match for the specified priority.
• -c, --cursor=: Display entries starting from the log position specified by the passed cursor.
• --after-cursor=: Display entries starting from the log position specified by this cursor and show the cursor when using the --show-cursor option.
• --show-cursor: Show the cursor after the two dashes of the last entry, similar to -- cursor: s=0639.... The format of the cursor is private and may change.
• --since=, --until=: Display entries starting from the specified date or up to the specified date, or within the specified date or up to the specified date. The format of the date specification should be 2012-10-30 18:17:16. If the time part is omitted, it is assumed to be 00:00:00. If only the seconds component is omitted, it is assumed to be :00. If the date part is omitted, it is assumed to be the current date, or the strings yesterday, today, tomorrow, representing the day before 00:00:00 of the current date, 00:00:00 of the current date, and the day after the current date, respectively. now denotes the current time. Lastly, a relative count may be specified with a - or + prefix, representing the number of times before or after the current time.
• -F, --field=: Print all possible data values accepted for the specified field within all log entries.
• --system, --user: Display messages from system services and the kernel (using --system), or from the current user services (using --user). If neither is specified, it displays all messages visible to the user.
• -M, --machine=: Display messages from a running local container, specifying the name of the container to connect to.
• -D DIR, --directory=DIR: Operate on the specified log directory DIR if provided, instead of default run-time and system log paths.
• --file=GLOB: Operate on the specified log files matching the file glob if provided, instead of default run-time and system log paths, and multiple times can be specified in which case files are properly interleaved.
• --root=ROOT: Operate on the directory tree under the directory path given as argument, if provided, instead of the root directory, e.g., --update catalog would create root/var/lib/systemd/catalog/database.
• --new-id128: Generate a new 128-bit ID suitable for identifying messages instead of showing log content, prepared for developers who need to use new identifiers for their introduced messages and wish to make them identifiable. This will print the new ID in three different formats that can be copied into source code or similar files.
• --header: Instead of showing log content, display internal header information of the accessed log fields.
• --disk-usage: Show the current disk usage of all log files.
• --list-catalog [128-bit-ID...]: List the contents of the message catalog in the form of a table of message ids and their short descriptive strings, and only for those entries specified if any 128-bit id is specified.
• --dump-catalog [128-bit-ID...]: Display the contents of the message catalog, with entries separated by lines comprising of two dashes and the ID, in a format identical to a .catalog file, and only for those entries specified if any 128-bit id is specified.
• --update-catalog: Update the message catalog index needed whenever new catalog files are installed, removed, or updated to regenerate the binary catalog index.
• --setup-keys: Generate a new key pair for sealing the forward-secure seals FSS, instead of showing log content. This will generate a sealed key and a verification key, with the sealed key stored in a log data directory and kept on the host, while the verification key should be stored externally.
• --force: Recreate the FSS keys when passed with --setup keys and forward-secure seals FSS are already set up.
• --interval=: Specify the change interval for the sealed key when generating an FSS key pair with --setup-keys, with shorter intervals increasing CPU consumption but narrowing the time range for undetected log changes, with a default of 15 minutes.
• --verify: Check the internal consistency of the log files, verifying the authenticity of the log files if they were generated with FSS enabled and the FSS verification key is specified with --verify key=.
• --verify-key=: Specify the FSS verification key for the --verify operation.
• --no-pager: Do not pipe the program's output through a pager.
• --vacuum-size=BYTES: Reduce disk usage to below the specified size.
• --vacuum-files=INT: Retain only the specified number of log files.
• --vacuum-time=TIME: Delete any entries earlier than the specified time point.
• --rotate: Request the log daemon to rotate the log files, which will block until the rotation operation is complete. Log rotation can ensure that all active log files are closed, renamed for archiving, and new blank log files are created as the new active log files. It is usually used together with --vacuum-size=, --vacuum-time=, and --vacuum-file= to improve the efficiency of log cleaning.
• -h, --help: Output help information.
• --version: Output version information.

Example

Display all the logs since the last boot.

journalctl

Display kernel logs.

journalctl -k

Using the -n parameter, you can display the last n lines of logs. If no line number is specified, the default is 10 lines.

journalctl -n 20

View the logs for a specific process.

journalctl _PID=1

Display logs from the last 30 minutes.

journalctl --since=-30m

Display logs since the year 2021.

journalctl --since="2021-01-01"

Display today's logs.

journalctl --since=today

Using the -f parameter provides the functionality similar to tail -f, continuously monitoring the newest logs.

journalctl -f

View the logs for a specific Unit.

journalctl -u nginx.service

Check the disk space used by logs.

journalctl --disk-usage

Require the log daemon to rotate log files. Log rotation ensures that all active log files are closed, renamed for archiving, and new blank log files are created as new active log files.

journalctl --rotate

Clean up logs older than a week.

journalctl --vacuum-time=1week

Reduce disk usage to below the specified size.

journalctl --vacuum-size=10M

Daily Question

https://github.com/WindrunnerMax/EveryDay

References

http://www.jinbuguo.com/systemd/journalctl.html#
https://www.commandlinux.com/man-page/man1/journalctl.1.html
https://blog.orchidflower.cn/2020/04/20/linux-command-introduction-04-journalctl/