blob: 4b0dd90cbbdb7893ff4fb8fbf85b256c28929290 [file] [log] [blame]
<?xml version='1.0'?> <!--*-nxml-*-->
<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
<!-- SPDX-License-Identifier: LGPL-2.1-or-later -->
<refentry id="journald.conf"
<refpurpose>Journal service configuration files</refpurpose>
<para>These files configure various parameters of the systemd journal service,
for a general description of the syntax.</para>
<para>The <command>systemd-journald</command> instance managing the default namespace is configured by
<filename>/etc/systemd/journald.conf</filename> and associated drop-ins. Instances managing other
namespaces read <filename>/etc/systemd/journald@<replaceable>NAMESPACE</replaceable>.conf</filename>
and associated drop-ins with the namespace identifier filled in. This allows each namespace to carry
a distinct configuration. See
for details about journal namespaces.</para>
<xi:include href="standard-conf.xml" xpointer="main-conf" />
<para>All options are configured in the
[Journal] section:</para>
<variablelist class='config-directives'>
<listitem><para>Controls where to store journal data. One of <literal>volatile</literal>,
<literal>persistent</literal>, <literal>auto</literal> and <literal>none</literal>. If
<literal>volatile</literal>, journal log data will be stored only in memory, i.e. below the
<filename>/run/log/journal</filename> hierarchy (which is created if needed). If
<literal>persistent</literal>, data will be stored preferably on disk, i.e. below the
<filename>/var/log/journal</filename> hierarchy (which is created if needed), with a fallback to
<filename>/run/log/journal</filename> (which is created if needed), during early boot and if the disk
is not writable. <literal>auto</literal> behaves like <literal>persistent</literal> if the
<filename>/var/log/journal</filename> directory exists, and <literal>volatile</literal> otherwise
(the existence of the directory controls the storage mode). <literal>none</literal> turns off all
storage, all log data received will be dropped (but forwarding to other targets, such as the console,
the kernel log buffer, or a syslog socket will still work). Defaults to <literal>auto</literal> in
the default journal namespace, and <literal>persistent</literal> in all others.</para>
<para>Note that journald will initially use volatile storage, until a call to
<command>journalctl --flush</command> (or sending <constant>SIGUSR1</constant> to journald) will cause
it to switch to persistent logging (under the conditions mentioned above). This is done automatically
on boot via <literal>systemd-journal-flush.service</literal>.</para>
<para>Note that when this option is changed to <literal>volatile</literal>, existing persistent data
is not removed. In the other direction,
<citerefentry><refentrytitle>journalctl</refentrytitle><manvolnum>1</manvolnum></citerefentry> with
the <option>--flush</option> option may be used to move volatile data to persistent storage.</para>
<listitem><para>Can take a boolean value. If enabled (the
default), data objects that shall be stored in the journal
and are larger than the default threshold of 512 bytes are
compressed before they are written to the file system. It
can also be set to a number of bytes to specify the
compression threshold directly. Suffixes like K, M, and G
can be used to specify larger units.</para></listitem>
<listitem><para>Takes a boolean value. If enabled (the
default), and a sealing key is available (as created by
<option>--setup-keys</option> command), Forward Secure Sealing
(FSS) for all persistent journal files is enabled. FSS is
based on <ulink
url="">Seekable Sequential Key
Generators</ulink> by G. A. Marson and B. Poettering
(doi:10.1007/978-3-642-40203-6_7) and may be used to protect
journal files from unnoticed alteration.</para></listitem>
<listitem><para>Controls whether to split up journal files per user, either <literal>uid</literal> or
<literal>none</literal>. Split journal files are primarily useful for access control: on UNIX/Linux access
control is managed per file, and the journal daemon will assign users read access to their journal files. If
<literal>uid</literal>, all regular users (with UID outside the range of system users, dynamic service users,
and the nobody user) will each get their own journal files, and system users will log to the system journal.
See <ulink url="">Users, Groups, UIDs and GIDs on systemd systems</ulink>
for more details about UID ranges.
If <literal>none</literal>, journal files are not split up by user and all messages are
instead stored in the single system journal. In this mode unprivileged users generally do not have access to
their own log data. Note that splitting up journal files by user is only available for journals stored
persistently. If journals are stored on volatile storage (see <varname>Storage=</varname> above), only a single
journal file is used. Defaults to <literal>uid</literal>.</para></listitem>
<listitem><para>Configures the rate limiting that is applied
to all messages generated on the system. If, in the time
interval defined by <varname>RateLimitIntervalSec=</varname>,
more messages than specified in
<varname>RateLimitBurst=</varname> are logged by a service,
all further messages within the interval are dropped until the
interval is over. A message about the number of dropped
messages is generated. This rate limiting is applied
per-service, so that two services which log do not interfere
with each other's limits. Defaults to 10000 messages in 30s.
The time specification for
<varname>RateLimitIntervalSec=</varname> may be specified in the
following units: <literal>s</literal>, <literal>min</literal>,
<literal>h</literal>, <literal>ms</literal>,
<literal>us</literal>. To turn off any kind of rate limiting,
set either value to 0.</para>
<para>Note that the effective rate limit is multiplied by a
factor derived from the available free disk space for the journal.
Currently, this factor is calculated using the base 2 logarithm.</para>
<title>Example <varname>RateLimitBurst=</varname> rate
modifications by the available disk space</title>
<tgroup cols='2'>
<colspec colname='freespace' />
<colspec colname='multiplier' />
<entry>Available Disk Space</entry>
<entry>Burst Multiplier</entry>
<entry>&lt;= 1MB</entry>
<entry>&lt;= 16MB</entry>
<entry>&lt;= 256MB</entry>
<entry>&lt;= 4GB</entry>
<entry>&lt;= 64GB</entry>
<entry>&lt;= 1TB</entry>
<para>If a service provides rate limits for itself through
<varname>LogRateLimitIntervalSec=</varname> and/or <varname>LogRateLimitBurst=</varname>
in <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>,
those values will override the settings specified here.</para>
<listitem><para>Enforce size limits on the journal files
stored. The options prefixed with <literal>System</literal>
apply to the journal files when stored on a persistent file
system, more specifically
<filename>/var/log/journal</filename>. The options prefixed
with <literal>Runtime</literal> apply to the journal files
when stored on a volatile in-memory file system, more
specifically <filename>/run/log/journal</filename>. The former
is used only when <filename>/var/</filename> is mounted,
writable, and the directory
<filename>/var/log/journal</filename> exists. Otherwise, only
the latter applies. Note that this means that during early
boot and if the administrator disabled persistent logging,
only the latter options apply, while the former apply if
persistent logging is enabled and the system is fully booted
up. <command>journalctl</command> and
<command>systemd-journald</command> ignore all files with
names not ending with <literal>.journal</literal> or
<literal>.journal~</literal>, so only such files, located in
the appropriate directories, are taken into account when
calculating current disk usage.</para>
<para><varname>SystemMaxUse=</varname> and
<varname>RuntimeMaxUse=</varname> control how much disk space
the journal may use up at most.
<varname>SystemKeepFree=</varname> and
<varname>RuntimeKeepFree=</varname> control how much disk
space systemd-journald shall leave free for other uses.
<command>systemd-journald</command> will respect both limits
and use the smaller of the two values.</para>
<para>The first pair defaults to 10% and the second to 15% of
the size of the respective file system, but each value is
capped to 4G. If the file system is nearly full and either
<varname>SystemKeepFree=</varname> or
<varname>RuntimeKeepFree=</varname> are violated when
systemd-journald is started, the limit will be raised to the
percentage that is actually free. This means that if there was
enough free space before and journal files were created, and
subsequently something else causes the file system to fill up,
journald will stop using more space, but it will not be
removing existing files to reduce the footprint again,
either. Also note that only archived files are deleted to reduce the
space occupied by journal files. This means that, in effect, there might
still be more space used than <varname>SystemMaxUse=</varname> or
<varname>RuntimeMaxUse=</varname> limit after a vacuuming operation is
<para><varname>SystemMaxFileSize=</varname> and
<varname>RuntimeMaxFileSize=</varname> control how large
individual journal files may grow at most. This influences
the granularity in which disk space is made available through
rotation, i.e. deletion of historic data. Defaults to one
eighth of the values configured with
<varname>SystemMaxUse=</varname> and
<varname>RuntimeMaxUse=</varname>, so that usually seven
rotated journal files are kept as history.</para>
<para>Specify values in bytes or use K, M, G, T, P, E as
units for the specified sizes (equal to 1024, 1024², … bytes).
Note that size limits are enforced synchronously when journal
files are extended, and no explicit rotation step triggered by
time is needed.</para>
<para><varname>SystemMaxFiles=</varname> and
<varname>RuntimeMaxFiles=</varname> control how many
individual journal files to keep at most. Note that only
archived files are deleted to reduce the number of files until
this limit is reached; active files will stay around. This
means that, in effect, there might still be more journal files
around in total than this limit after a vacuuming operation is
complete. This setting defaults to 100.</para></listitem>
<listitem><para>The maximum time to store entries in a single
journal file before rotating to the next one. Normally,
time-based rotation should not be required as size-based
rotation with options such as
<varname>SystemMaxFileSize=</varname> should be sufficient to
ensure that journal files do not grow without bounds. However,
to ensure that not too much data is lost at once when old
journal files are deleted, it might make sense to change this
value from the default of one month. Set to 0 to turn off this
feature. This setting takes time values which may be suffixed
with the units <literal>year</literal>,
<literal>month</literal>, <literal>week</literal>,
<literal>day</literal>, <literal>h</literal> or
<literal>m</literal> to override the default time unit of
<listitem><para>The maximum time to store journal entries.
This controls whether journal files containing entries older
than the specified time span are deleted. Normally, time-based
deletion of old journal files should not be required as
size-based deletion with options such as
<varname>SystemMaxUse=</varname> should be sufficient to
ensure that journal files do not grow without bounds. However,
to enforce data retention policies, it might make sense to
change this value from the default of 0 (which turns off this
feature). This setting also takes time values which may be
suffixed with the units <literal>year</literal>,
<literal>month</literal>, <literal>week</literal>,
<literal>day</literal>, <literal>h</literal> or <literal>
m</literal> to override the default time unit of
<listitem><para>The timeout before synchronizing journal files
to disk. After syncing, journal files are placed in the
OFFLINE state. Note that syncing is unconditionally done
immediately after a log message of priority CRIT, ALERT or
EMERG has been logged. This setting hence applies only to
messages of the levels ERR, WARNING, NOTICE, INFO, DEBUG. The
default timeout is 5 minutes. </para></listitem>
<listitem><para>Control whether log messages received by the journal daemon shall be forwarded to a
traditional syslog daemon, to the kernel log buffer (kmsg), to the system console, or sent as wall
messages to all logged-in users. These options take boolean arguments. If forwarding to syslog is
enabled but nothing reads messages from the socket, forwarding to syslog has no effect. By default,
only forwarding to wall is enabled. These settings may be overridden at boot time with the kernel
command line options <literal>systemd.journald.forward_to_syslog</literal>,
<literal>systemd.journald.forward_to_console</literal>, and
<literal>systemd.journald.forward_to_wall</literal>. If the option name is specified without
<literal>=</literal> and the following argument, true is assumed. Otherwise, the argument is parsed
as a boolean.</para>
<para>When forwarding to the console, the TTY to log to can be changed with
<varname>TTYPath=</varname>, described below.</para>
<para>When forwarding to the kernel log buffer (kmsg), make sure to select a suitably large size for
the log buffer, for example by adding <literal>log_buf_len=8M</literal> to the kernel command line.
<command>systemd</command> will automatically disable kernel's rate-limiting applied to userspace
processes (equivalent to setting <literal>printk.devkmsg=on</literal>).</para></listitem>
<listitem><para>Controls the maximum log level of messages
that are stored in the journal, forwarded to syslog, kmsg, the
console or wall (if that is enabled, see above). As argument,
takes one of
or integer values in the range of 0–7 (corresponding to the
same levels). Messages equal or below the log level specified
are stored/forwarded, messages above are dropped. Defaults to
<literal>debug</literal> for <varname>MaxLevelStore=</varname>
and <varname>MaxLevelSyslog=</varname>, to ensure that the all
messages are stored in the journal and forwarded to syslog.
Defaults to
<literal>notice</literal> for <varname>MaxLevelKMsg=</varname>,
<literal>info</literal> for <varname>MaxLevelConsole=</varname>,
and <literal>emerg</literal> for
<varname>MaxLevelWall=</varname>. These settings may be
overridden at boot time with the kernel command line options
<listitem><para>Takes a boolean value. If enabled <command>systemd-journal</command> processes
<filename>/dev/kmsg</filename> messages generated by the kernel. In the default journal namespace
this option is enabled by default, it is disabled in all others.</para></listitem>
<listitem><para>Takes a boolean value. If enabled <command>systemd-journal</command> will turn on
kernel auditing on start-up. If disabled it will turn it off. If unset it will neither enable nor
disable it, leaving the previous state unchanged. Note that this option does not control whether
<command>systemd-journald</command> collects generated audit records, it just controls whether it
tells the kernel to generate them. This means if another tool turns on auditing even if
<command>systemd-journald</command> left it off, it will still collect the generated
messages. Defaults to on.</para></listitem>
<listitem><para>Change the console TTY to use if
<varname>ForwardToConsole=yes</varname> is used. Defaults to
<listitem><para>The maximum line length to permit when converting stream logs into record logs. When a systemd
unit's standard output/error are connected to the journal via a stream socket, the data read is split into
individual log records at newline (<literal>\n</literal>, ASCII 10) and <constant>NUL</constant> characters. If no such delimiter is
read for the specified number of bytes a hard log record boundary is artificially inserted, breaking up overly
long lines into multiple log records. Selecting overly large values increases the possible memory usage of the
Journal daemon for each stream client, as in the worst case the journal daemon needs to buffer the specified
number of bytes in memory before it can flush a new log record to disk. Also note that permitting overly large
line maximum line lengths affects compatibility with traditional log protocols as log records might not fit
anymore into a single <constant>AF_UNIX</constant> or <constant>AF_INET</constant> datagram. Takes a size in
bytes. If the value is suffixed with K, M, G or T, the specified size is parsed as Kilobytes, Megabytes,
Gigabytes, or Terabytes (with the base 1024), respectively. Defaults to 48K, which is relatively large but
still small enough so that log records likely fit into network datagrams along with extra room for
metadata. Note that values below 79 are not accepted and will be bumped to 79.</para></listitem>
<title>Forwarding to traditional syslog daemons</title>
Journal events can be transferred to a different logging daemon
in two different ways. With the first method, messages are
immediately forwarded to a socket
(<filename>/run/systemd/journal/syslog</filename>), where the
traditional syslog daemon can read them. This method is
controlled by the <varname>ForwardToSyslog=</varname> option. With a
second method, a syslog daemon behaves like a normal journal
client, and reads messages from the journal files, similarly to
With this, messages do not have to be read immediately,
which allows a logging daemon which is only started late in boot
to access all messages since the start of the system. In
addition, full structured meta-data is available to it. This
method of course is available only if the messages are stored in
a journal file at all. So it will not work if
<varname>Storage=none</varname> is set. It should be noted that
usually the <emphasis>second</emphasis> method is used by syslog
daemons, so the <varname>Storage=</varname> option, and not the
<varname>ForwardToSyslog=</varname> option, is relevant for them.
<title>See Also</title>