chiark / gitweb /
man: explain that the journal field SYSLOG_IDENTIFIER is usually derived from program...
[elogind.git] / man / systemd.journal-fields.xml
index 8a15598e631ed4000a6e25ef64c9dfd96ab53ac4..6eeb6e395d5b86b706daf662c2ed075b3f3bda42 100644 (file)
                 <title>Description</title>
 
                 <para>Entries in the journal resemble an environment
-                block in their syntax, however with fields that can
+                block in their syntax but with fields that can
                 include binary data. Primarily, fields are formatted
                 UTF-8 text strings, and binary formatting is used only
                 where formatting as UTF-8 text strings makes little
                 sense. New fields may freely be defined by
                 applications, but a few fields have special
                 meaning. All fields with special meanings are
-                optional. In some cases fields may appear more than
+                optional. In some cases, fields may appear more than
                 once per entry.</para>
         </refsect1>
 
                                         (formatted as decimal string),
                                         the identifier string
                                         (i.e. "tag"), and the client
-                                        PID.</para>
+                                        PID. (Note that the tag is
+                                        usually derived from glibc's
+                                        <varname>program_invocation_short_name</varname>
+                                        variable, see <citerefentry><refentrytitle>program_invocation_short_name</refentrytitle><manvolnum>3</manvolnum></citerefentry>.)</para>
                                 </listitem>
 
                         </varlistentry>
                                 <term><varname>_UID=</varname></term>
                                 <term><varname>_GID=</varname></term>
                                 <listitem>
-                                        <para>The process, user and
+                                        <para>The process, user, and
                                         group ID of the process the
                                         journal entry originates from
                                         formatted as a decimal
                                 <term><varname>_CMDLINE=</varname></term>
                                 <listitem>
                                         <para>The name, the executable
-                                        path and the command line of
+                                        path, and the command line of
                                         the process the journal entry
                                         originates from.</para>
                                 </listitem>
                                 <term><varname>_SELINUX_CONTEXT=</varname></term>
                                 <listitem>
                                         <para>The SELinux security
-                                        context of the process the
-                                        journal entry originates
+                                        context (label) of the process
+                                        the journal entry originates
                                         from.</para>
                                 </listitem>
                         </varlistentry>
                                         the major and minor of the
                                         device node, separated by <literal>:</literal>
                                         and prefixed by <literal>b</literal>. Similar
-                                        for character devices, but
+                                        for character devices but
                                         prefixed by <literal>c</literal>. For network
-                                        devices the interface index,
+                                        devices, this is the interface index
                                         prefixed by <literal>n</literal>. For all other
-                                        devices <literal>+</literal> followed by the
-                                        subsystem name, followed by
+                                        devices, this is the subsystem name
+                                        prefixed by <literal>+</literal>, followed by
                                         <literal>:</literal>, followed by the kernel
                                         device name.</para>
                                 </listitem>
                 JSON Format</ulink>, the addresses of journal entries
                 are serialized into fields prefixed with double
                 underscores. Note that these are not proper fields when
-                stored in the journal but for addressing meta data of
+                stored in the journal but for addressing metadata of
                 entries. They cannot be written as part of structured
                 log entries via calls such as
                 <citerefentry><refentrytitle>sd_journal_send</refentrytitle><manvolnum>3</manvolnum></citerefentry>. They