Description¶

sd_journal_open() opens the log journal for reading. It will find all journal files automatically and interleave them automatically when reading. As first argument it takes a pointer to a sd_journal pointer, which, on success, will contain a journal context object. The second argument is a flags field, which may consist of the following flags ORed together: SD_JOURNAL_LOCAL_ONLY makes sure only journal files generated on the local machine will be opened. SD_JOURNAL_RUNTIME_ONLY makes sure only volatile journal files will be opened, excluding those which are stored on persistent storage. SD_JOURNAL_SYSTEM will cause journal files of system services and the kernel (in opposition to user session processes) to be opened. SD_JOURNAL_CURRENT_USER will cause journal files of the current user to be opened. If neither SD_JOURNAL_SYSTEM nor SD_JOURNAL_CURRENT_USER are specified, all journal file types will be opened.

sd_journal_open_namespace() is similar to sd_journal_open() but takes an additional namespace parameter that specifies which journal namespace to operate on. If specified as NULL the call is identical to sd_journal_open(). If non-NULL only data from the namespace identified by the specified parameter is accessed. This call understands two additional flags: if SD_JOURNAL_ALL_NAMESPACES is specified the namespace parameter is ignored and all defined namespaces are accessed simultaneously; if SD_JOURNAL_INCLUDE_DEFAULT_NAMESPACE the specified namespace and the default namespace are accessed but no others (this flag has no effect when namespace is passed as NULL). For details about journal namespaces see systemd-journald.service(8).

sd_journal_open_directory() is similar to sd_journal_open() but takes an absolute directory path as argument. All journal files in this directory will be opened and interleaved automatically. This call also takes a flags argument. The flags parameters accepted by this call are SD_JOURNAL_OS_ROOT, SD_JOURNAL_SYSTEM, and SD_JOURNAL_CURRENT_USER. If SD_JOURNAL_OS_ROOT is specified, journal files are searched for below the usual /var/log/journal and /run/log/journal relative to the specified path, instead of directly beneath it. The other two flags limit which files are opened, the same as for sd_journal_open().

sd_journal_open_directory_fd() is similar to sd_journal_open_directory(), but takes a file descriptor referencing a directory in the file system instead of an absolute file system path. In addition to the flags accepted by sd_journal_open_directory(), this function also accepts SD_JOURNAL_TAKE_DIRECTORY_FD. If SD_JOURNAL_TAKE_DIRECTORY_FD is specified, the function will take the ownership of the specified file descriptor on success, and it will be closed by sd_journal_close(), hence the caller of the function must not close the file descriptor. When the flag is not specified, sd_journal_close() will not close the file descriptor, so the caller should close it after sd_journal_close().

sd_journal_open_files() is similar to sd_journal_open() but takes a NULL-terminated list of file paths to open. All files will be opened and interleaved automatically. This call also takes a flags argument, but it must be passed as 0 as no flags are currently understood for this call. Please note that in the case of a live journal, this function is only useful for debugging, because individual journal files can be rotated at any moment, and the opening of specific files is inherently racy.

sd_journal_open_files_fd() is similar to sd_journal_open_files() but takes an array of open file descriptors that must reference journal files, instead of an array of file system paths. Pass the array of file descriptors as second argument, and the number of array entries in the third. The flags parameter must be passed as 0.

sd_journal objects cannot be used in the child after a fork. Functions which take a journal object as an argument (sd_journal_next() and others) will return -ECHILD after a fork.

sd_journal_close() will close the journal context allocated with sd_journal_open() or sd_journal_open_directory() and free its resources.

When opening the journal only journal files accessible to the calling user will be opened. If journal files are not accessible to the caller, this will be silently ignored.

See sd_journal_next(3) for an example of how to iterate through the journal after opening it with sd_journal_open().

A journal context object returned by sd_journal_open() references a specific journal entry as current entry, similar to a file seek index in a classic file system file, but without absolute positions. It may be altered with sd_journal_next(3) and sd_journal_seek_head(3) and related calls. The current entry position may be exported in cursor strings, as accessible via sd_journal_get_cursor(3). Cursor strings may be used to globally identify a specific journal entry in a stable way and then later to seek to it (or if the specific entry is not available locally, to its closest entry in time) sd_journal_seek_cursor(3).

Notification of journal changes is available via sd_journal_get_fd() and related calls.

Return Value¶

The sd_journal_open(), sd_journal_open_directory(), and sd_journal_open_files() calls return 0 on success or a negative errno-style error code. sd_journal_close() returns nothing.

Notes¶

All functions listed here are thread-agnostic and only a single specific thread may operate on a given object during its entire lifetime. It's safe to allocate multiple independent objects and use each from a specific thread in parallel. However, it's not safe to allocate such an object in one thread, and operate or free it from any other, even if locking is used to ensure these threads don't operate on it at the very same time.

Functions described here are available as a shared library, which can be compiled against and linked to with the libsystemd pkg-config(1) file.

See Also¶

systemd(1), sd-journal(3), systemd-journald.service(8), sd_journal_next(3), sd_journal_get_data(3)