1 systemd System and Service Manager
5 * Configuration of unit files may now be extended via drop-in
6 files without having to edit/override the unit files
7 themselves. More specifically, if the administrator wants to
8 change one value for a service file foobar.service he can
9 now do so by dropping in a configuration snippet into
10 /etc/systemd/system/foobar.service.d/*.conf. The unit logic
11 will load all these snippets and apply them on top of the
12 main unit configuration file, possibly extending or
13 overriding its settings. Using these drop-in snippets is
14 generally nicer than the two earlier options for changing
15 unit files locally: copying the files from
16 /usr/lib/systemd/system/ to /etc/systemd/system/ and editing
17 them there; or creating a new file in /etc/systemd/system/
18 that incorporates the original one via ".include". Drop-in
19 snippets into these .d/ directories can be placed in any
20 directory systemd looks for units in, and the usual
21 overriding semantics between /usr/lib, /etc and /run apply
24 * Most unit file settings which take lists of items can now be
25 reset by assigning the empty string to them. For example,
26 normally, settings such as Environment=FOO=BAR append a new
27 environment variable assignment to the environment block,
28 each time they are used. By assigning Environment= the empty
29 string the environment block can be reset to empty. This is
30 particularly useful with the .d/*.conf drop-in snippets
31 mentioned above, since this adds the ability to reset list
32 settings from vendor unit files via these drop-ins.
34 * systemctl gained a new "list-dependencies" command for
35 listing the dependencies of a unit recursively.
37 * Inhibitors are now honored and listed by "systemctl
38 suspend", "systemctl poweroff" (and similar) too, not only
39 GNOME. These commands will also list active sessions by
42 * Resource limits (as exposed by the various control group
43 controllers) can now be controlled dynamically at runtime
44 for all units. More specifically, you can now use a command
45 like "systemctl set-cgroup-attr foobar.service cpu.shares
46 2000" to alter the CPU shares a specific service gets. These
47 settings are stored persistently on disk, and thus allow the
48 administrator to easily adjust the resource usage of
49 services with a few simple commands. This dynamic resource
50 management logic is also available to other programs via the
51 bus. Almost any kernel cgroup attribute and controller is
54 * systemd-vconsole-setup will now copy all font settings to
55 all allocated VTs, where it previously applied them only to
58 * libsystemd-login gained the new sd_session_get_tty() API
61 * This release drops support for a few legacy or
62 distribution-specific LSB facility names when parsing init
63 scripts: $x-display-manager, $mail-transfer-agent,
64 $mail-transport-agent, $mail-transfer-agent, $smtp,
65 $null. Also, the mail-transfer-agent.target unit backing
66 this has been removed. Distributions which want to retain
67 compatibility with this should carry the burden for
68 supporting this themselves and patch support for these back
69 in, if they really need to. Also, the facilities $syslog and
70 $local_fs are now ignored, since systemd does not support
71 early-boot LSB init scripts anymore, and these facilities
72 are implied anyway for normal services. syslog.target has
75 * There are new bus calls on PID1's Manager object for
76 cancelling jobs, and removing snapshot units. Previously,
77 both calls were only available on the Job and Snapshot
80 * systemd-journal-gatewayd gained SSL support.
82 * The various "environment" files, such as /etc/locale.conf
83 now support continuation lines with a backslash ("\") as
84 last character in the line, similar in style (but different)
85 to how this is supported in shells.
87 * For normal user processes the _SYSTEMD_USER_UNIT= field is
88 now implicitly appended to every log entry logged. systemctl
89 has been updated to filter by this field when operating on a
90 user systemd instance.
92 * nspawn will now implicitly add the CAP_AUDIT_WRITE and
93 CAP_AUDIT_CONTROL capabilities to the capabilities set for
94 the container. This makes it easier to boot unmodified
95 Fedora systems in a container, which however still requires
96 audit=0 to be passed on the kernel command line. Auditing in
97 kernel and userspace is unfortunately still too broken in
98 context of containers, hence we recommend compiling it out
99 of the kernel or using audit=0. Hopefully this will be fixed
100 one day for good in the kernel.
102 * nspawn gained the new --bind= and --bind-ro= parameters to
103 bind mount specific directories from the host into the
106 * nspawn will now mount its own devpts file system instance
107 into the container, in order not to leak pty devices from
108 the host into the container.
110 * systemd will now read the firmware boot time performance
111 information from the EFI variables, if the used boot loader
112 supports this, and takes it into account for boot performance
113 analysis via "systemd-analyze". This is currently supported
114 only in conjunction with Gummiboot, but could be supported
115 by other boot loaders too. For details see:
117 http://www.freedesktop.org/wiki/Software/systemd/BootLoaderInterface
119 * A new generator has been added that automatically mounts the
120 EFI System Partition (ESP) to /boot, if that directory
121 exists, is empty, and no other file system has been
122 configured to be mounted there.
124 * logind will now send out PrepareForSleep(false) out
125 unconditionally, after coming back from suspend. This may be
126 used by applications as asynchronous notification for
127 system resume events.
129 * "systemctl unlock-sessions" has been added, that allows
130 unlocking the screens of all user sessions at once, similar
131 how "systemctl lock-sessions" already locked all users
132 sessions. This is backed by a new D-Bus call UnlockSessions().
134 * "loginctl seat-status" will now show the master device of a
135 seat. (i.e. the device of a seat that needs to be around for
136 the seat to be considered available, usually the graphics
139 * tmpfiles gained a new "X" line type, that allows
140 configuration of files and directories (with wildcards) that
141 shall be excluded from automatic cleanup ("aging").
143 * udev default rules set the device node permissions now only
144 at "add" events, and do not change them any longer with a
145 later "change" event.
147 * The log messages for lid events and power/sleep keypresses
148 now carry a message ID.
150 * We now have a substantially larger unit test suite, but this
151 continues to be work in progress.
153 * udevadm hwdb gained a new --root= parameter to change the
154 root directory to operate relative to.
156 * logind will now issue a background sync() request to the kernel
157 early at shutdown, so that dirty buffers are flushed to disk early
158 instead of at the last moment, in order to optimize shutdown
161 * A new bootctl tool has been added that is an interface for
162 certain boot loader operations. This is currently a preview
163 and is likely to be extended into a small mechanism daemon
164 like timedated, localed, hostnamed, and can be used by
165 graphical UIs to enumerate available boot options, and
166 request boot into firmware operations.
168 * systemd-bootchart has been relicensed to LGPLv2.1+ to match
169 the rest of the package. It also has been updated to work
170 correctly in initrds.
172 * Policykit previously has been runtime optional, and is now
173 also compile time optional via a configure switch.
175 * systemd-analyze has been reimplemented in C. Also "systemctl
176 dot" has moved into systemd-analyze.
178 * "systemctl status" with no further parameters will now print
179 the status of all active or failed units.
181 * Operations such as "systemctl start" can now be executed
182 with a new mode "--irreversible" which may be used to queue
183 operations that cannot accidentally be reversed by a later
184 job queuing. This is by default used to make shutdown
185 requests more robust.
187 * The Python API of systemd now gained a new module for
188 reading journal files.
190 * A new tool kernel-install has been added that can install
191 kernel images according to the Boot Loader Specification:
193 http://www.freedesktop.org/wiki/Specifications/BootLoaderSpec
195 * Boot time console output has been improved to provide
196 animated boot time output for hanging jobs.
198 * A new tool systemd-activate has been added which can be used
199 to test socket activation with, directly from the command
200 line. This should make it much easier to test and debug
201 socket activation in daemons.
203 * journalctl gained a new "--reverse" (or -r) option to show
204 journal output in reverse order (i.e. newest line first).
206 * journalctl gained a new "--pager-end" (or -e) option to jump
207 to immediately jump to the end of the journal in the
208 pager. This is only supported in conjunction with "less".
210 * journalctl gained a new "--user-unit=" option, that works
211 similar to "--unit=" but filters for user units rather than
214 * A number of unit files to ease adoption of systemd in
215 initrds has been added. This moves some minimal logic from
216 the various initrd implementations into systemd proper.
218 * The journal files are now owned by a new group
219 "systemd-journal", which exists specifically to allow access
220 to the journal, and nothing else. Previously, we used the
221 "adm" group for that, which however possibly covers more
222 than just journal/log file access. This new group is now
223 already used by systemd-journal-gatewayd to ensure this
224 daemon gets access to the journal files and as little else
225 as possible. Note that "make install" will also set FS ACLs
226 up for /var/log/journal to give "adm" and "wheel" read
227 access to it, in addition to "systemd-journal" which owns
228 the journal files. We recommend that packaging scripts also
229 add read access to "adm" + "wheel" to /var/log/journal, and
230 all existing/future journal files. To normal users and
231 administrators little changes, however packagers need to
232 ensure to create the "systemd-journal" system group at
233 package installation time.
235 * The systemd-journal-gatewayd now runs as unprivileged user
236 systemd-journal-gateway:systemd-journal-gateway. Packaging
237 scripts need to create these system user/group at
240 * timedated now exposes a new boolean property CanNTP that
241 indicates whether a local NTP service is available or not.
243 * systemd-detect-virt will now also detect xen PVs
245 * The pstore file system is now mounted by default, if it is
248 * In addition to the SELinux and IMA policies we will now also
249 load SMACK policies at early boot.
251 Contributions from: Adel Gadllah, Aleksander Morgado, Auke
252 Kok, Ayan George, Bastien Nocera, Colin Walters, Daniel Buch,
253 Daniel Wallace, Dave Reisner, David Herrmann, David Strauss,
254 Eelco Dolstra, Enrico Scholz, Frederic Crozat, Harald Hoyer,
255 Jan Janssen, Jonathan Callen, Kay Sievers, Lennart Poettering,
256 Lukas Nykryn, Mantas Mikulėnas, Marc-Antoine Perennou, Martin
257 Pitt, Mauro Dreissig, Max F. Albrecht, Michael Biebl, Michael
258 Olbrich, Michal Schmidt, Michal Sekletar, Michal Vyskocil,
259 Michał Bartoszkiewicz, Mirco Tischler, Nathaniel Chen, Nestor
260 Ovroy, Oleksii Shevchuk, Paul W. Frields, Piotr Drąg, Rob
261 Clark, Ryan Lortie, Simon McVittie, Simon Peeters, Steven
262 Hiscocks, Thomas Hindoe Paaboel Andersen, Tollef Fog Heen, Tom
263 Gundersen, Umut Tezduyar, William Giokas, Zbigniew
264 Jędrzejewski-Szmek, Zeeshan Ali (Khattak)
268 * Timer units now support calendar time events in addition to
269 monotonic time events. That means you can now trigger a unit
270 based on a calendar time specification such as "Thu,Fri
271 2013-*-1,5 11:12:13" which refers to 11:12:13 of the first
272 or fifth day of any month of the year 2013, given that it is
273 a thursday or friday. This brings timer event support
274 considerably closer to cron's capabilities. For details on
275 the supported calendar time specification language see
278 * udev now supports a number of different naming policies for
279 network interfaces for predictable names, and a combination
280 of these policies is now the default. Please see this wiki
281 document for details:
283 http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames
285 * Auke Kok's bootchart implementation has been added to the
286 systemd tree. It's an optional component that can graph the
287 boot in quite some detail. It's one of the best bootchart
288 implementations around and minimal in its code and
291 * nss-myhostname has been integrated into the systemd source
292 tree. nss-myhostname guarantees that the local hostname
293 always stays resolvable via NSS. It has been a weak
294 requirement of systemd-hostnamed since a long time, and
295 since its code is actually trivial we decided to just
296 include it in systemd's source tree. It can be turned off
297 with a configure switch.
299 * The read-ahead logic is now capable of properly detecting
300 whether a btrfs file system is on SSD or rotating media, in
301 order to optimize the read-ahead scheme. Previously, it was
302 only capable of detecting this on traditional file systems
305 * In udev, additional device properties are now read from the
306 IAB in addition to the OUI database. Also, Bluetooth company
307 identities are attached to the devices as well.
309 * In service files %U may be used as specifier that is
310 replaced by the configured user name of the service.
312 * nspawn may now be invoked without a controlling TTY. This
313 makes it suitable for invocation as its own service. This
314 may be used to set up a simple containerized server system
315 using only core OS tools.
317 * systemd and nspawn can now accept socket file descriptors
318 when they are started for socket activation. This enables
319 implementation of socket activated nspawn
320 containers. i.e. think about autospawning an entire OS image
321 when the first SSH or HTTP connection is received. We expect
322 that similar functionality will also be added to libvirt-lxc
325 * journalctl will now suppress ANSI color codes when
328 * systemctl will no longer show control group information for
329 a unit if a the control group is empty anyway.
331 * logind can now automatically suspend/hibernate/shutdown the
334 * /etc/machine-info and hostnamed now also expose the chassis
335 type of the system. This can be used to determine whether
336 the local system is a laptop, desktop, handset or
337 tablet. This information may either be configured by the
338 user/vendor or is automatically determined from ACPI and DMI
339 information if possible.
341 * A number of PolicyKit actions are now bound together with
342 "imply" rules. This should simplify creating UIs because
343 many actions will now authenticate similar ones as well.
345 * Unit files learnt a new condition ConditionACPower= which
346 may be used to conditionalize a unit depending on whether an
347 AC power source is connected or not, of whether the system
348 is running on battery power.
350 * systemctl gained a new "is-failed" verb that may be used in
351 shell scripts and suchlike to check whether a specific unit
352 is in the "failed" state.
354 * The EnvironmentFile= setting in unit files now supports file
355 globbing, and can hence be used to easily read a number of
356 environment files at once.
358 * systemd will no longer detect and recognize specific
359 distributions. All distribution-specific #ifdeffery has been
360 removed, systemd is now fully generic and
361 distribution-agnostic. Effectively, not too much is lost as
362 a lot of the code is still accessible via explicit configure
363 switches. However, support for some distribution specific
364 legacy configuration file formats has been dropped. We
365 recommend distributions to simply adopt the configuration
366 files everybody else uses now and convert the old
367 configuration from packaging scripts. Most distributions
368 already did that. If that's not possible or desirable,
369 distributions are welcome to forward port the specific
370 pieces of code locally from the git history.
372 * When logging a message about a unit systemd will now always
373 log the unit name in the message meta data.
375 * localectl will now also discover system locale data that is
376 not stored in locale archives, but directly unpacked.
378 * logind will no longer unconditionally use framebuffer
379 devices as seat masters, i.e. as devices that are required
380 to be existing before a seat is considered preset. Instead,
381 it will now look for all devices that are tagged as
382 "seat-master" in udev. By default framebuffer devices will
383 be marked as such, but depending on local systems other
384 devices might be marked as well. This may be used to
385 integrate graphics cards using closed source drivers (such
386 as NVidia ones) more nicely into logind. Note however, that
387 we recommend using the open source NVidia drivers instead,
388 and no udev rules for the closed-source drivers will be
389 shipped from us upstream.
391 Contributions from: Adam Williamson, Alessandro Crismani, Auke
392 Kok, Colin Walters, Daniel Wallace, Dave Reisner, David
393 Herrmann, David Strauss, Dimitrios Apostolou, Eelco Dolstra,
394 Eric Benoit, Giovanni Campagna, Hannes Reinecke, Henrik
395 Grindal Bakken, Hermann Gausterer, Kay Sievers, Lennart
396 Poettering, Lukas Nykryn, Mantas Mikulėnas, Marcel Holtmann,
397 Martin Pitt, Matthew Monaco, Michael Biebl, Michael Terry,
398 Michal Schmidt, Michal Sekletar, Michał Bartoszkiewicz, Oleg
399 Samarin, Pekka Lundstrom, Philip Nilsson, Ramkumar
400 Ramachandra, Richard Yao, Robert Millan, Sami Kerola, Shawn
401 Landden, Thomas Hindoe Paaboel Andersen, Thomas Jarosch,
402 Tollef Fog Heen, Tom Gundersen, Umut Tezduyar, Zbigniew
407 * udev gained support for loading additional device properties
408 from an indexed database that is keyed by vendor/product IDs
409 and similar device identifiers. For the beginning this
410 "hwdb" is populated with data from the well-known PCI and
411 USB database, but also includes PNP, ACPI and OID data. In
412 the longer run this indexed database shall grow into
413 becoming the one central database for non-essential
414 userspace device metadata. Previously, data from the PCI/USB
415 database was only attached to select devices, since the
416 lookup was a relatively expensive operation due to O(n) time
417 complexity (with n being the number of entries in the
418 database). Since this is now O(1), we decided to add in this
419 data for all devices where this is available, by
420 default. Note that the indexed database needs to be rebuilt
421 when new data files are installed. To achieve this you need
422 to update your packaging scripts to invoke "udevadm hwdb
423 --update" after installation of hwdb data files. For
424 RPM-based distributions we introduced the new
425 %udev_hwdb_update macro for this purpose.
427 * The Journal gained support for the "Message Catalog", an
428 indexed database to link up additional information with
429 journal entries. For further details please check:
431 http://www.freedesktop.org/wiki/Software/systemd/catalog
433 The indexed message catalog database also needs to be
434 rebuilt after installation of message catalog files. Use
435 "journalctl --update-catalog" for this. For RPM-based
436 distributions we introduced the %journal_catalog_update
437 macro for this purpose.
439 * The Python Journal bindings gained support for the standard
440 Python logging framework.
442 * The Journal API gained new functions for checking whether
443 the underlying file system of a journal file is capable of
444 properly reporting file change notifications, or whether
445 applications that want to reflect journal changes "live"
446 need to recheck journal files continously in appropriate
449 * It is now possible to set the "age" field for tmpfiles
450 entries to 0, indicating that files matching this entry
451 shall always be removed when the directories are cleaned up.
453 * coredumpctl gained a new "gdb" verb which invokes gdb
454 right-away on the selected coredump.
456 * There's now support for "hybrid sleep" on kernels that
457 support this, in addition to "suspend" and "hibernate". Use
458 "systemctl hybrid-sleep" to make use of this.
460 * logind's HandleSuspendKey= setting (and related settings)
461 now gained support for a new "lock" setting to simply
462 request the screen lock on all local sessions, instead of
463 actually executing a suspend or hibernation.
465 * systemd will now mount the EFI variables file system by
468 * Socket units now gained support for configuration of the
469 SMACK security label.
471 * timedatectl will now output the time of the last and next
472 daylight saving change.
474 * We dropped support for various legacy and distro-specific
475 concepts, such as insserv, early-boot SysV services
476 (i.e. those for non-standard runlevels such as 'b' or 'S')
477 or ArchLinux /etc/rc.conf support. We recommend the
478 distributions who still need support this to either continue
479 to maintain the necessary patches downstream, or find a
480 different solution. (Talk to us if you have questions!)
482 * Various systemd components will now bypass PolicyKit checks
483 for root and otherwise handle properly if PolicyKit is not
484 found to be around. This should fix most issues for
485 PolicyKit-less systems. Quite frankly this should have been
486 this way since day one. It is absolutely our intention to
487 make systemd work fine on PolicyKit-less systems, and we
488 consider it a bug if something doesn't work as it should if
489 PolicyKit is not around.
491 * For embedded systems it is now possible to build udev and
492 systemd without blkid and/or kmod support.
494 * "systemctl switch-root" is now capable of switching root
495 more than once. I.e. in addition to transitions from the
496 initrd to the host OS it is now possible to transition to
497 further OS images from the host. This is useful to implement
498 offline updating tools.
500 * Various other additions have been made to the RPM macros
501 shipped with systemd. Use %udev_rules_update() after
502 installing new udev rules files. %_udevhwdbdir,
503 %_udevrulesdir, %_journalcatalogdir, %_tmpfilesdir,
504 %_sysctldir are now available which resolve to the right
505 directories for packages to place various data files in.
507 * journalctl gained the new --full switch (in addition to
508 --all, to disable ellipsation for long messages.
510 Contributions from: Anders Olofsson, Auke Kok, Ben Boeckel,
511 Colin Walters, Cosimo Cecchi, Daniel Wallace, Dave Reisner,
512 Eelco Dolstra, Holger Hans Peter Freyther, Kay Sievers,
513 Chun-Yi Lee, Lekensteyn, Lennart Poettering, Mantas Mikulėnas,
514 Marti Raudsepp, Martin Pitt, Mauro Dreissig, Michael Biebl,
515 Michal Schmidt, Michal Sekletar, Miklos Vajna, Nis Martensen,
516 Oleksii Shevchuk, Olivier Brunel, Ramkumar Ramachandra, Thomas
517 Bächler, Thomas Hindoe Paaboel Andersen, Tom Gundersen, Tony
518 Camuso, Umut Tezduyar, Zbigniew Jędrzejewski-Szmek
522 * journalctl gained new --since= and --until= switches to
523 filter by time. It also now supports nice filtering for
524 units via --unit=/-u.
526 * Type=oneshot services may use ExecReload= and do the
529 * The journal daemon now supports time-based rotation and
530 vacuuming, in addition to the usual disk-space based
533 * The journal will now index the available field values for
534 each field name. This enables clients to show pretty drop
535 downs of available match values when filtering. The bash
536 completion of journalctl has been updated
537 accordingly. journalctl gained a new switch -F to list all
538 values a certain field takes in the journal database.
540 * More service events are now written as structured messages
541 to the journal, and made recognizable via message IDs.
543 * The timedated, localed and hostnamed mini-services which
544 previously only provided support for changing time, locale
545 and hostname settings from graphical DEs such as GNOME now
546 also have a minimal (but very useful) text-based client
547 utility each. This is probably the nicest way to changing
548 these settings from the command line now, especially since
549 it lists available options and is fully integrated with bash
552 * There's now a new tool "systemd-coredumpctl" to list and
553 extract coredumps from the journal.
555 * We now install a README each in /var/log/ and
556 /etc/rc.d/init.d explaining where the system logs and init
557 scripts went. This hopefully should help folks who go to
558 that dirs and look into the otherwise now empty void and
561 * When user-services are invoked (by systemd --user) the
562 $MANAGERPID env var is set to the PID of systemd.
564 * SIGRTMIN+24 when sent to a --user instance will now result
565 in immediate termination of systemd.
567 * gatewayd received numerous feature additions such as a
568 "follow" mode, for live syncing and filtering.
570 * browse.html now allows filtering and showing detailed
571 information on specific entries. Keyboard navigation and
572 mouse screen support has been added.
574 * gatewayd/journalctl now supports HTML5/JSON
575 Server-Sent-Events as output.
577 * The SysV init script compatibility logic will now
578 heuristically determine whether a script supports the
579 "reload" verb, and only then make this available as
582 * "systemctl status --follow" has been removed, use "journalctl
585 * journald.conf's RuntimeMinSize=, PersistentMinSize= settings
586 have been removed since they are hardly useful to be
589 * And I'd like to take the opportunity to specifically mention
590 Zbigniew for his great contributions. Zbigniew, you rock!
592 Contributions from: Andrew Eikum, Christian Hesse, Colin
593 Guthrie, Daniel J Walsh, Dave Reisner, Eelco Dolstra, Ferenc
594 Wágner, Kay Sievers, Lennart Poettering, Lukas Nykryn, Mantas
595 Mikulėnas, Martin Mikkelsen, Martin Pitt, Michael Olbrich,
596 Michael Stapelberg, Michal Schmidt, Sebastian Ott, Thomas
597 Bächler, Umut Tezduyar, Will Woods, Wulf C. Krueger, Zbigniew
598 Jędrzejewski-Szmek, Сковорода Никита Андреевич
602 * If /etc/vconsole.conf is non-existent or empty we will no
603 longer load any console font or key map at boot by
604 default. Instead the kernel defaults will be left
605 intact. This is definitely the right thing to do, as no
606 configuration should mean no configuration, and hard-coding
607 font names that are different on all archs is probably a bad
608 idea. Also, the kernel default key map and font should be
609 good enough for most cases anyway, and mostly identical to
610 the userspace fonts/key maps we previously overloaded them
611 with. If distributions want to continue to default to a
612 non-kernel font or key map they should ship a default
613 /etc/vconsole.conf with the appropriate contents.
615 Contributions from: Colin Walters, Daniel J Walsh, Dave
616 Reisner, Kay Sievers, Lennart Poettering, Lukas Nykryn, Tollef
617 Fog Heen, Tom Gundersen, Zbigniew Jędrzejewski-Szmek
621 * journalctl gained a new --cursor= switch to show entries
622 starting from the specified location in the journal.
624 * We now enforce a size limit on journal entry fields exported
625 with "-o json" in journalctl. Fields larger than 4K will be
626 assigned null. This can be turned off with --all.
628 * An (optional) journal gateway daemon is now available as
629 "systemd-journal-gatewayd.service". This service provides
630 access to the journal via HTTP and JSON. This functionality
631 will be used to implement live log synchronization in both
632 pull and push modes, but has various other users too, such
633 as easy log access for debugging of embedded devices. Right
634 now it is already useful to retrieve the journal via HTTP:
636 # systemctl start systemd-journal-gatewayd.service
637 # wget http://localhost:19531/entries
639 This will download the journal contents in a
640 /var/log/messages compatible format. The same as JSON:
642 # curl -H"Accept: application/json" http://localhost:19531/entries
644 This service is also accessible via a web browser where a
645 single static HTML5 app is served that uses the JSON logic
646 to enable the user to do some basic browsing of the
647 journal. This will be extended later on. Here's an example
648 screenshot of this app in its current state:
650 http://0pointer.de/public/journal-gatewayd
652 Contributions from: Kay Sievers, Lennart Poettering, Robert
653 Milasan, Tom Gundersen
657 * The bash completion logic is now available for journalctl
660 * We don't mount the "cpuset" controller anymore together with
661 "cpu" and "cpuacct", as "cpuset" groups generally cannot be
662 started if no parameters are assigned to it. "cpuset" hence
663 broke code that assumed it it could create "cpu" groups and
666 * journalctl -f will now subscribe to terminal size changes,
667 and line break accordingly.
669 Contributions from: Dave Reisner, Kay Sievers, Lennart
670 Poettering, Lukas Nykrynm, Mirco Tischler, Václav Pavlín
674 * nspawn will now create a symlink /etc/localtime in the
675 container environment, copying the host's timezone
676 setting. Previously this has been done via a bind mount, but
677 since symlinks cannot be bind mounted this has now been
678 changed to create/update the appropriate symlink.
680 * journalctl -n's line number argument is now optional, and
681 will default to 10 if omitted.
683 * journald will now log the maximum size the journal files may
684 take up on disk. This is particularly useful if the default
685 built-in logic of determining this parameter from the file
686 system size is used. Use "systemctl status
687 systemd-journald.service" to see this information.
689 * The multi-seat X wrapper tool has been stripped down. As X
690 is now capable of enumerating graphics devices via udev in a
691 seat-aware way the wrapper is not strictly necessary
692 anymore. A stripped down temporary stop-gap is still shipped
693 until the upstream display managers have been updated to
694 fully support the new X logic. Expect this wrapper to be
695 removed entirely in one of the next releases.
697 * HandleSleepKey= in logind.conf has been split up into
698 HandleSuspendKey= and HandleHibernateKey=. The old setting
699 is not available anymore. X11 and the kernel are
700 distuingishing between these keys and we should too. This
701 also means the inhibition lock for these keys has been split
704 Contributions from: Dave Airlie, Eelco Dolstra, Lennart
705 Poettering, Lukas Nykryn, Václav Pavlín
709 * Whenever a unit changes state we'll now log this to the
710 journal and show along the unit's own log output in
713 * ConditionPathIsMountPoint= can now properly detect bind
714 mount points too. (Previously, a bind mount of one file
715 system to another place in the same file system could not be
716 detected as mount, since they shared struct stat's st_dev
719 * We will now mount the cgroup controllers cpu, cpuacct,
720 cpuset and the controllers net_cls, net_prio together by
723 * nspawn containers will now have a virtualized boot
724 ID. (i.e. /proc/sys/kernel/random/boot_id is now mounted
725 over with a randomized ID at container initialization). This
726 has the effect of making "journalctl -b" do the right thing
729 * The JSON output journal serialization has been updated not
730 to generate "endless" list objects anymore, but rather one
731 JSON object per line. This is more in line how most JSON
732 parsers expect JSON objects. The new output mode
733 "json-pretty" has been added to provide similar output, but
734 neatly aligned for readability by humans.
736 * We dropped all explicit sync() invocations in the shutdown
737 code. The kernel does this implicitly anyway in the kernel
738 reboot() syscall. halt(8)'s -n option is now a compatibility
741 * We now support virtualized reboot() in containers, as
742 supported by newer kernels. We will fall back to exit() if
743 CAP_SYS_REBOOT is not available to the container. Also,
744 nspawn makes use of this now and will actually reboot the
745 container if the containerized OS asks for that.
747 * journalctl will only show local log output by default
748 now. Use --merge (-m) to show remote log output, too.
750 * libsystemd-journal gained the new sd_journal_get_usage()
751 call to determine the current disk usage of all journal
752 files. This is exposed in the new "journalctl --disk-usage"
755 * journald gained a new configuration setting SplitMode= in
756 journald.conf which may be used to control how user journals
757 are split off. See journald.conf(5) for details.
759 * A new condition type ConditionFileNotEmpty= has been added.
761 * tmpfiles' "w" lines now support file globbing, to write
762 multiple files at once.
764 * We added Python bindings for the journal submission
765 APIs. More Python APIs for a number of selected APIs will
766 likely follow. Note that we intend to add native bindings
767 only for the Python language, as we consider it common
768 enough to deserve bindings shipped within systemd. There are
769 various projects outside of systemd that provide bindings
770 for languages such as PHP or Lua.
772 * Many conditions will now resolve specifiers such as %i. In
773 addition, PathChanged= and related directives of .path units
774 now support specifiers as well.
776 * There's now a new RPM macro definition for the system preset
779 * journald will now warn if it can't foward a message to the
780 syslog daemon because it's socket is full.
782 * timedated will no longer write or process /etc/timezone,
783 except on Debian. As we do not support late mounted /usr
784 anymore /etc/localtime always being a symlink is now safe,
785 and hence the information in /etc/timezone is not necessary
788 * logind will now always reserve one VT for a text getty (VT6
789 by default). Previously if more than 6 X sessions where
790 started they took up all the VTs with auto-spawned gettys,
791 so that no text gettys were available anymore.
793 * udev will now automatically inform the btrfs kernel logic
794 about btrfs RAID components showing up. This should make
795 simple hotplug based btrfs RAID assembly work.
797 * PID 1 will now increase its RLIMIT_NOFILE to 64K by default
798 (but not for its children which will stay at the kernel
799 default). This should allow setups with a lot more listening
802 * systemd will now always pass the configured timezone to the
803 kernel at boot. timedated will do the same when the timezone
806 * logind's inhibition logic has been updated. By default,
807 logind will now handle the lid switch, the power and sleep
808 keys all the time, even in graphical sessions. If DEs want
809 to handle these events on their own they should take the new
810 handle-power-key, handle-sleep-key and handle-lid-switch
811 inhibitors during their runtime. A simple way to achiveve
812 that is to invoke the DE wrapped in an invocation of:
814 systemd-inhibit --what=handle-power-key:handle-sleep-key:handle-lid-switch ...
816 * Access to unit operations is now checked via SELinux taking
817 the unit file label and client process label into account.
819 * systemd will now notify the administrator in the journal
820 when he over-mounts a non-empty directory.
822 * There are new specifiers that are resolved in unit files,
823 for the host name (%H), the machine ID (%m) and the boot ID
826 Contributions from: Allin Cottrell, Auke Kok, Brandon Philips,
827 Colin Guthrie, Colin Walters, Daniel J Walsh, Dave Reisner,
828 Eelco Dolstra, Jan Engelhardt, Kay Sievers, Lennart
829 Poettering, Lucas De Marchi, Lukas Nykryn, Mantas Mikulėnas,
830 Martin Pitt, Matthias Clasen, Michael Olbrich, Pierre Schmitz,
831 Shawn Landden, Thomas Hindoe Paaboel Andersen, Tom Gundersen,
832 Václav Pavlín, Yin Kangkai, Zbigniew Jędrzejewski-Szmek
836 * Support for reading structured kernel messages from
837 /dev/kmsg has now been added and is enabled by default.
839 * Support for reading kernel messages from /proc/kmsg has now
840 been removed. If you want kernel messages in the journal
841 make sure to run a recent kernel (>= 3.5) that supports
842 reading structured messages from /dev/kmsg (see
843 above). /proc/kmsg is now exclusive property of classic
844 syslog daemons again.
846 * The libudev API gained the new
847 udev_device_new_from_device_id() call.
849 * The logic for file system namespace (ReadOnlyDirectory=,
850 ReadWriteDirectoy=, PrivateTmp=) has been reworked not to
851 require pivot_root() anymore. This means fewer temporary
852 directories are created below /tmp for this feature.
854 * nspawn containers will now see and receive all submounts
855 made on the host OS below the root file system of the
858 * Forward Secure Sealing is now supported for Journal files,
859 which provide cryptographical sealing of journal files so
860 that attackers cannot alter log history anymore without this
861 being detectable. Lennart will soon post a blog story about
862 this explaining it in more detail.
864 * There are two new service settings RestartPreventExitStatus=
865 and SuccessExitStatus= which allow configuration of exit
866 status (exit code or signal) which will be excepted from the
867 restart logic, resp. consider successful.
869 * journalctl gained the new --verify switch that can be used
870 to check the integrity of the structure of journal files and
871 (if Forward Secure Sealing is enabled) the contents of
874 * nspawn containers will now be run with /dev/stdin, /dev/fd/
875 and similar symlinks pre-created. This makes running shells
876 as container init process a lot more fun.
878 * The fstab support can now handle PARTUUID= and PARTLABEL=
881 * A new ConditionHost= condition has been added to match
882 against the hostname (with globs) and machine ID. This is
883 useful for clusters where a single OS image is used to
884 provision a large number of hosts which shall run slightly
885 different sets of services.
887 * Services which hit the restart limit will now be placed in a
890 Contributions from: Bertram Poettering, Dave Reisner, Huang
891 Hang, Kay Sievers, Lennart Poettering, Lukas Nykryn, Martin
892 Pitt, Simon Peeters, Zbigniew Jędrzejewski-Szmek
896 * When running in --user mode systemd will now become a
897 subreaper (PR_SET_CHILD_SUBREAPER). This should make the ps
898 tree a lot more organized.
900 * A new PartOf= unit dependency type has been introduced that
901 may be used to group services in a natural way.
903 * "systemctl enable" may now be used to enable instances of
906 * journalctl now prints error log levels in red, and
907 warning/notice log levels in bright white. It also supports
908 filtering by log level now.
910 * cgtop gained a new -n switch (similar to top), to configure
911 the maximum number of iterations to run for. It also gained
912 -b, to run in batch mode (accepting no input).
914 * The suffix ".service" may now be ommited on most systemctl
915 command lines involving service unit names.
917 * There's a new bus call in logind to lock all sessions, as
918 well as a loginctl verb for it "lock-sessions".
920 * libsystemd-logind.so gained a new call sd_journal_perror()
921 that works similar to libc perror() but logs to the journal
922 and encodes structured information about the error number.
924 * /etc/crypttab entries now understand the new keyfile-size=
927 * shutdown(8) now can send a (configurable) wall message when
928 a shutdown is cancelled.
930 * The mount propagation mode for the root file system will now
931 default to "shared", which is useful to make containers work
932 nicely out-of-the-box so that they receive new mounts from
933 the host. This can be undone locally by running "mount
934 --make-rprivate /" if needed.
936 * The prefdm.service file has been removed. Distributions
937 should maintain this unit downstream if they intend to keep
938 it around. However, we recommend writing normal unit files
939 for display managers instead.
941 * Since systemd is a crucial part of the OS we will now
942 default to a number of compiler switches that improve
943 security (hardening) such as read-only relocations, stack
944 protection, and suchlike.
946 * The TimeoutSec= setting for services is now split into
947 TimeoutStartSec= and TimeoutStopSec= to allow configuration
948 of individual time outs for the start and the stop phase of
951 Contributions from: Artur Zaprzala, Arvydas Sidorenko, Auke
952 Kok, Bryan Kadzban, Dave Reisner, David Strauss, Harald Hoyer,
953 Jim Meyering, Kay Sievers, Lennart Poettering, Mantas
954 Mikulėnas, Martin Pitt, Michal Schmidt, Michal Sekletar, Peter
955 Alfredsen, Shawn Landden, Simon Peeters, Terence Honles, Tom
956 Gundersen, Zbigniew Jędrzejewski-Szmek
960 * The journal and id128 C APIs are now fully documented as man
963 * Extra safety checks have been added when transitioning from
964 the initial RAM disk to the main system to avoid accidental
967 * /etc/crypttab entries now understand the new keyfile-offset=
970 * systemctl -t can now be used to filter by unit load state.
972 * The journal C API gained the new sd_journal_wait() call to
973 make writing synchronous journal clients easier.
975 * journalctl gained the new -D switch to show journals from a
978 * journalctl now displays a special marker between log
979 messages of two different boots.
981 * The journal is now explicitly flushed to /var via a service
982 systemd-journal-flush.service, rather than implicitly simply
983 by seeing /var/log/journal to be writable.
985 * journalctl (and the journal C APIs) can now match for much
986 more complex expressions, with alternatives and
989 * When transitioning from the initial RAM disk to the main
990 system we will now kill all processes in a killing spree to
991 ensure no processes stay around by accident.
993 * Three new specifiers may be used in unit files: %u, %h, %s
994 resolve to the user name, user home directory resp. user
995 shell. This is useful for running systemd user instances.
997 * We now automatically rotate journal files if their data
998 object hash table gets a fill level > 75%. We also size the
999 hash table based on the configured maximum file size. This
1000 together should lower hash collisions drastically and thus
1001 speed things up a bit.
1003 * journalctl gained the new "--header" switch to introspect
1004 header data of journal files.
1006 * A new setting SystemCallFilters= has been added to services
1007 which may be used to apply blacklists or whitelists to
1008 system calls. This is based on SECCOMP Mode 2 of Linux 3.5.
1010 * nspawn gained a new --link-journal= switch (and quicker: -j)
1011 to link the container journal with the host. This makes it
1012 very easy to centralize log viewing on the host for all
1013 guests while still keeping the journal files separated.
1015 * Many bugfixes and optimizations
1017 Contributions from: Auke Kok, Eelco Dolstra, Harald Hoyer, Kay
1018 Sievers, Lennart Poettering, Malte Starostik, Paul Menzel, Rex
1019 Tsai, Shawn Landden, Tom Gundersen, Ville Skyttä, Zbigniew
1024 * Several tools now understand kernel command line arguments,
1025 which are only read when run in an initial RAM disk. They
1026 usually follow closely their normal counterparts, but are
1029 * There's a new tool to analyze the readahead files that are
1030 automatically generated at boot. Use:
1032 /usr/lib/systemd/systemd-readahead analyze /.readahead
1034 * We now provide an early debug shell on tty9 if this enabled. Use:
1036 systemctl enable debug-shell.service
1038 * All plymouth related units have been moved into the Plymouth
1039 package. Please make sure to upgrade your Plymouth version
1042 * systemd-tmpfiles now supports getting passed the basename of
1043 a configuration file only, in which case it will look for it
1044 in all appropriate directories automatically.
1046 * udevadm info now takes a /dev or /sys path as argument, and
1047 does the right thing. Example:
1049 udevadm info /dev/sda
1050 udevadm info /sys/class/block/sda
1052 * systemctl now prints a warning if a unit is stopped but a
1053 unit that might trigger it continues to run. Example: a
1054 service is stopped but the socket that activates it is left
1057 * "systemctl status" will now mention if the log output was
1058 shortened due to rotation since a service has been started.
1060 * The journal API now exposes functions to determine the
1061 "cutoff" times due to rotation.
1063 * journald now understands SIGUSR1 and SIGUSR2 for triggering
1064 immediately flushing of runtime logs to /var if possible,
1065 resp. for triggering immediate rotation of the journal
1068 * It is now considered an error if a service is attempted to
1069 be stopped that is not loaded.
1071 * XDG_RUNTIME_DIR now uses numeric UIDs instead of usernames.
1073 * systemd-analyze now supports Python 3
1075 * tmpfiles now supports cleaning up directories via aging
1076 where the first level dirs are always kept around but
1077 directories beneath it automatically aged. This is enabled
1078 by prefixing the age field with '~'.
1080 * Seat objects now expose CanGraphical, CanTTY properties
1081 which is required to deal with very fast bootups where the
1082 display manager might be running before the graphics drivers
1083 completed initialization.
1085 * Seat objects now expose a State property.
1087 * We now include RPM macros for service enabling/disabling
1088 based on the preset logic. We recommend RPM based
1089 distributions to make use of these macros if possible. This
1090 makes it simpler to reuse RPM spec files across
1093 * We now make sure that the collected systemd unit name is
1094 always valid when services log to the journal via
1097 * There's a new man page kernel-command-line(7) detailing all
1098 command line options we understand.
1100 * The fstab generator may now be disabled at boot by passing
1101 fstab=0 on the kernel command line.
1103 * A new kernel command line option modules-load= is now understood
1104 to load a specific kernel module statically, early at boot.
1106 * Unit names specified on the systemctl command line are now
1107 automatically escaped as needed. Also, if file system or
1108 device paths are specified they are automatically turned
1109 into the appropriate mount or device unit names. Example:
1111 systemctl status /home
1112 systemctl status /dev/sda
1114 * The SysVConsole= configuration option has been removed from
1115 system.conf parsing.
1117 * The SysV search path is no longer exported on the D-Bus
1120 * The Names= option is been removed from unit file parsing.
1122 * There's a new man page bootup(7) detailing the boot process.
1124 * Every unit and every generator we ship with systemd now
1125 comes with full documentation. The self-explanatory boot is
1128 * A couple of services gained "systemd-" prefixes in their
1129 name if they wrap systemd code, rather than only external
1130 code. Among them fsck@.service which is now
1131 systemd-fsck@.service.
1133 * The HaveWatchdog property has been removed from the D-Bus
1136 * systemd.confirm_spawn= on the kernel command line should now
1139 * There's a new man page crypttab(5) which details all options
1140 we actually understand.
1142 * systemd-nspawn gained a new --capability= switch to pass
1143 additional capabilities to the container.
1145 * timedated will now read known NTP implementation unit names
1146 from /usr/lib/systemd/ntp-units.d/*.list,
1147 systemd-timedated-ntp.target has been removed.
1149 * journalctl gained a new switch "-b" that lists log data of
1150 the current boot only.
1152 * The notify socket is in the abstract namespace again, in
1153 order to support daemons which chroot() at start-up.
1155 * There is a new Storage= configuration option for journald
1156 which allows configuration of where log data should go. This
1157 also provides a way to disable journal logging entirely, so
1158 that data collected is only forwarded to the console, the
1159 kernel log buffer or another syslog implementation.
1161 * Many bugfixes and optimizations
1163 Contributions from: Auke Kok, Colin Guthrie, Dave Reisner,
1164 David Strauss, Eelco Dolstra, Kay Sievers, Lennart Poettering,
1165 Lukas Nykryn, Michal Schmidt, Michal Sekletar, Paul Menzel,
1166 Shawn Landden, Tom Gundersen
1170 * "systemctl help <unit>" now shows the man page if one is
1173 * Several new man pages have been added.
1175 * MaxLevelStore=, MaxLevelSyslog=, MaxLevelKMsg=,
1176 MaxLevelConsole= can now be specified in
1177 journald.conf. These options allow reducing the amount of
1178 data stored on disk or forwarded by the log level.
1180 * TimerSlackNSec= can now be specified in system.conf for
1181 PID1. This allows system-wide power savings.
1183 Contributions from: Dave Reisner, Kay Sievers, Lauri Kasanen,
1184 Lennart Poettering, Malte Starostik, Marc-Antoine Perennou,
1189 * logind is now capable of (optionally) handling power and
1190 sleep keys as well as the lid switch.
1192 * journalctl now understands the syntax "journalctl
1193 /usr/bin/avahi-daemon" to get all log output of a specific
1196 * CapabilityBoundingSet= in system.conf now also influences
1197 the capability bound set of usermode helpers of the kernel.
1199 Contributions from: Daniel Drake, Daniel J. Walsh, Gert
1200 Michael Kulyk, Harald Hoyer, Jean Delvare, Kay Sievers,
1201 Lennart Poettering, Matthew Garrett, Matthias Clasen, Paul
1202 Menzel, Shawn Landden, Tero Roponen, Tom Gundersen
1206 * Note that we skipped 139 releases here in order to set the
1207 new version to something that is greater than both udev's
1208 and systemd's most recent version number.
1210 * udev: all udev sources are merged into the systemd source tree now.
1211 All future udev development will happen in the systemd tree. It
1212 is still fully supported to use the udev daemon and tools without
1213 systemd running, like in initramfs or other init systems. Building
1214 udev though, will require the *build* of the systemd tree, but
1215 udev can be properly *run* without systemd.
1217 * udev: /lib/udev/devices/ are not read anymore; systemd-tmpfiles
1218 should be used to create dead device nodes as workarounds for broken
1221 * udev: RUN+="socket:..." and udev_monitor_new_from_socket() is
1222 no longer supported. udev_monitor_new_from_netlink() needs to be
1223 used to subscribe to events.
1225 * udev: when udevd is started by systemd, processes which are left
1226 behind by forking them off of udev rules, are unconditionally cleaned
1227 up and killed now after the event handling has finished. Services or
1228 daemons must be started as systemd services. Services can be
1229 pulled-in by udev to get started, but they can no longer be directly
1230 forked by udev rules.
1232 * udev: the daemon binary is called systemd-udevd now and installed
1233 in /usr/lib/systemd/. Standalone builds or non-systemd systems need
1234 to adapt to that, create symlink, or rename the binary after building
1237 * libudev no longer provides these symbols:
1238 udev_monitor_from_socket()
1239 udev_queue_get_failed_list_entry()
1240 udev_get_{dev,sys,run}_path()
1241 The versions number was bumped and symbol versioning introduced.
1243 * systemd-loginctl and systemd-journalctl have been renamed
1244 to loginctl and journalctl to match systemctl.
1246 * The config files: /etc/systemd/systemd-logind.conf and
1247 /etc/systemd/systemd-journald.conf have been renamed to
1248 logind.conf and journald.conf. Package updates should rename
1249 the files to the new names on upgrade.
1251 * For almost all files the license is now LGPL2.1+, changed
1252 from the previous GPL2.0+. Exceptions are some minor stuff
1253 of udev (which will be changed to LGPL2.1 eventually, too),
1254 and the MIT licensed sd-daemon.[ch] library that is suitable
1255 to be used as drop-in files.
1257 * systemd and logind now handle system sleep states, in
1258 particular suspending and hibernating.
1260 * logind now implements a sleep/shutdown/idle inhibiting logic
1261 suitable for a variety of uses. Soonishly Lennart will blog
1262 about this in more detail.
1264 * var-run.mount and var-lock.mount are no longer provided
1265 (which prevously bind mounted these directories to their new
1266 places). Distributions which have not converted these
1267 directories to symlinks should consider stealing these files
1268 from git history and add them downstream.
1270 * We introduced the Documentation= field for units and added
1271 this to all our shipped units. This is useful to make it
1272 easier to explore the boot and the purpose of the various
1275 * All smaller setup units (such as
1276 systemd-vconsole-setup.service) now detect properly if they
1277 are run in a container and are skipped when
1278 appropriate. This guarantees an entirely noise-free boot in
1279 Linux container environments such as systemd-nspawn.
1281 * A framework for implementing offline system updates is now
1282 integrated, for details see:
1283 http://freedesktop.org/wiki/Software/systemd/SystemUpdates
1285 * A new service type Type=idle is available now which helps us
1286 avoiding ugly interleaving of getty output and boot status
1289 * There's now a system-wide CapabilityBoundingSet= option to
1290 globally reduce the set of capabilities for the
1291 system. This is useful to drop CAP_SYS_MKNOD, CAP_SYS_RAWIO,
1292 CAP_NET_RAW, CAP_SYS_MODULE, CAP_SYS_TIME, CAP_SYS_PTRACE or
1293 even CAP_NET_ADMIN system-wide for secure systems.
1295 * There are now system-wide DefaultLimitXXX= options to
1296 globally change the defaults of the various resource limits
1297 for all units started by PID 1.
1299 * Harald Hoyer's systemd test suite has been integrated into
1300 systemd which allows easy testing of systemd builds in qemu
1301 and nspawn. (This is really awesome! Ask us for details!)
1303 * The fstab parser is now implemented as generator, not inside
1306 * systemctl will now warn you if .mount units generated from
1307 /etc/fstab are out of date due to changes in fstab that
1308 haven't been read by systemd yet.
1310 * systemd is now suitable for usage in initrds. Dracut has
1311 already been updated to make use of this. With this in place
1312 initrds get a slight bit faster but primarily are much
1313 easier to introspect and debug since "systemctl status" in
1314 the host system can be used to introspect initrd services,
1315 and the journal from the initrd is kept around too.
1317 * systemd-delta has been added, a tool to explore differences
1318 between user/admin configuration and vendor defaults.
1320 * PrivateTmp= now affects both /tmp and /var/tmp.
1322 * Boot time status messages are now much prettier and feature
1323 proper english language. Booting up systemd has never been
1326 * Read-ahead pack files now include the inode number of all
1327 files to pre-cache. When the inode changes the pre-caching
1328 is not attempted. This should be nicer to deal with updated
1329 packages which might result in changes of read-ahead
1332 * We now temporaritly lower the kernel's read_ahead_kb variable
1333 when collecting read-ahead data to ensure the kernel's
1334 built-in read-ahead does not add noise to our measurements
1335 of necessary blocks to pre-cache.
1337 * There's now RequiresMountsFor= to add automatic dependencies
1338 for all mounts necessary for a specific file system path.
1340 * MountAuto= and SwapAuto= have been removed from
1341 system.conf. Mounting file systems at boot has to take place
1344 * nspawn now learned a new switch --uuid= to set the machine
1345 ID on the command line.
1347 * nspawn now learned the -b switch to automatically search
1350 * vt102 is now the default TERM for serial TTYs, upgraded from
1353 * systemd-logind now works on VT-less systems.
1355 * The build tree has been reorganized. The individual
1356 components now have directories of their own.
1358 * A new condition type ConditionPathIsReadWrite= is now available.
1360 * nspawn learned the new -C switch to create cgroups for the
1361 container in other hierarchies.
1363 * We now have support for hardware watchdogs, configurable in
1366 * The scheduled shutdown logic now has a public API.
1368 * We now mount /tmp as tmpfs by default, but this can be
1369 masked and /etc/fstab can override it.
1371 * Since udisks doesn't make use of /media anymore we are not
1372 mounting a tmpfs on it anymore.
1374 * journalctl gained a new --local switch to only interleave
1375 locally generated journal files.
1377 * We can now load the IMA policy at boot automatically.
1379 * The GTK tools have been split off into a systemd-ui.
1381 Contributions from: Andreas Schwab, Auke Kok, Ayan George,
1382 Colin Guthrie, Daniel Mack, Dave Reisner, David Ward, Elan
1383 Ruusamäe, Frederic Crozat, Gergely Nagy, Guillermo Vidal,
1384 Hannes Reinecke, Harald Hoyer, Javier Jardón, Kay Sievers,
1385 Lennart Poettering, Lucas De Marchi, Léo Gillot-Lamure,
1386 Marc-Antoine Perennou, Martin Pitt, Matthew Monaco, Maxim
1387 A. Mikityanskiy, Michael Biebl, Michael Olbrich, Michal
1388 Schmidt, Nis Martensen, Patrick McCarty, Roberto Sassu, Shawn
1389 Landden, Sjoerd Simons, Sven Anders, Tollef Fog Heen, Tom
1394 * This is mostly a bugfix release
1396 * Support optional initialization of the machine ID from the
1397 KVM or container configured UUID.
1399 * Support immediate reboots with "systemctl reboot -ff"
1401 * Show /etc/os-release data in systemd-analyze output
1403 * Many bugfixes for the journal, including endianess fixes and
1404 ensuring that disk space enforcement works
1406 * sd-login.h is C++ comptaible again
1408 * Extend the /etc/os-release format on request of the Debian
1411 * We now refuse non-UTF8 strings used in various configuration
1412 and unit files. This is done to ensure we don't pass invalid
1413 data over D-Bus or expose it elsewhere.
1415 * Register Mimo USB Screens as suitable for automatic seat
1418 * Read SELinux client context from journal clients in a race
1421 * Reorder configuration file lookup order. /etc now always
1422 overrides /run in order to allow the administrator to always
1423 and unconditionally override vendor supplied or
1424 automatically generated data.
1426 * The various user visible bits of the journal now have man
1427 pages. We still lack man pages for the journal API calls
1430 * We now ship all man pages in HTML format again in the
1433 Contributions from: Dave Reisner, Dirk Eibach, Frederic
1434 Crozat, Harald Hoyer, Kay Sievers, Lennart Poettering, Marti
1435 Raudsepp, Michal Schmidt, Shawn Landden, Tero Roponen, Thierry
1440 * This is mostly a bugfix release
1442 * systems lacking /etc/os-release are no longer supported.
1444 * Various functionality updates to libsystemd-login.so
1446 * Track class of PAM logins to distuingish greeters from
1449 Contributions from: Kay Sievers, Lennart Poettering, Michael
1454 * This is an important bugfix release for v41.
1456 * Building man pages is now optional which should be useful
1457 for those building systemd from git but unwilling to install
1460 * Watchdog support for supervising services is now usable. In
1461 a future release support for hardware watchdogs
1462 (i.e. /dev/watchdog) will be added building on this.
1464 * Service start rate limiting is now configurable and can be
1465 turned off per service. When a start rate limit is hit a
1466 reboot can automatically be triggered.
1468 * New CanReboot(), CanPowerOff() bus calls in systemd-logind.
1470 Contributions from: Benjamin Franzke, Bill Nottingham,
1471 Frederic Crozat, Lennart Poettering, Michael Olbrich, Michal
1472 Schmidt, Michał Górny, Piotr Drąg
1476 * The systemd binary is installed /usr/lib/systemd/systemd now;
1477 An existing /sbin/init symlink needs to be adapted with the
1480 * The code that loads kernel modules has been ported to invoke
1481 libkmod directly, instead of modprobe. This means we do not
1482 support systems with module-init-tools anymore.
1484 * Watchdog support is now already useful, but still not
1487 * A new kernel command line option systemd.setenv= is
1488 understood to set system wide environment variables
1489 dynamically at boot.
1491 * We now limit the set of capabilities of systemd-journald.
1493 * We now set SIGPIPE to ignore by default, since it only is
1494 useful in shell pipelines, and has little use in general
1495 code. This can be disabled with IgnoreSIPIPE=no in unit
1498 Contributions from: Benjamin Franzke, Kay Sievers, Lennart
1499 Poettering, Michael Olbrich, Michal Schmidt, Tom Gundersen,
1504 * This is mostly a bugfix release
1506 * We now expose the reason why a service failed in the
1507 "Result" D-Bus property.
1509 * Rudimentary service watchdog support (will be completed over
1510 the next few releases.)
1512 * When systemd forks off in order execute some service we will
1513 now immediately changes its argv[0] to reflect which process
1514 it will execute. This is useful to minimize the time window
1515 with a generic argv[0], which makes bootcharts more useful
1517 Contributions from: Alvaro Soliverez, Chris Paulson-Ellis, Kay
1518 Sievers, Lennart Poettering, Michael Olbrich, Michal Schmidt,
1519 Mike Kazantsev, Ray Strode
1523 * This is mostly a test release, but incorporates many
1526 * New systemd-cgtop tool to show control groups by their
1529 * Linking against libacl for ACLs is optional again. If
1530 disabled, support tracking device access for active logins
1531 goes becomes unavailable, and so does access to the user
1532 journals by the respective users.
1534 * If a group "adm" exists, journal files are automatically
1535 owned by them, thus allow members of this group full access
1536 to the system journal as well as all user journals.
1538 * The journal now stores the SELinux context of the logging
1539 client for all entries.
1541 * Add C++ inclusion guards to all public headers
1543 * New output mode "cat" in the journal to print only text
1544 messages, without any meta data like date or time.
1546 * Include tiny X server wrapper as a temporary stop-gap to
1547 teach XOrg udev display enumeration. This is used by display
1548 managers such as gdm, and will go away as soon as XOrg
1549 learned native udev hotplugging for display devices.
1551 * Add new systemd-cat tool for executing arbitrary programs
1552 with STDERR/STDOUT connected to the journal. Can also act as
1553 BSD logger replacement, and does so by default.
1555 * Optionally store all locally generated coredumps in the
1556 journal along with meta data.
1558 * systemd-tmpfiles learnt four new commands: n, L, c, b, for
1559 writing short strings to files (for usage for /sys), and for
1560 creating symlinks, character and block device nodes.
1562 * New unit file option ControlGroupPersistent= to make cgroups
1563 persistent, following the mechanisms outlined in
1564 http://www.freedesktop.org/wiki/Software/systemd/PaxControlGroups
1566 * Support multiple local RTCs in a sane way
1568 * No longer monopolize IO when replaying readahead data on
1569 rotating disks, since we might starve non-file-system IO to
1570 death, since fanotify() will not see accesses done by blkid,
1573 * Don't show kernel threads in systemd-cgls anymore, unless
1574 requested with new -k switch.
1576 Contributions from: Dan Horák, Kay Sievers, Lennart
1577 Poettering, Michal Schmidt
1581 * This is mostly a test release, but incorporates many
1584 * The git repository moved to:
1585 git://anongit.freedesktop.org/systemd/systemd
1586 ssh://git.freedesktop.org/git/systemd/systemd
1588 * First release with the journal
1589 http://0pointer.de/blog/projects/the-journal.html
1591 * The journal replaces both systemd-kmsg-syslogd and
1592 systemd-stdout-bridge.
1594 * New sd_pid_get_unit() API call in libsystemd-logind
1596 * Many systemadm clean-ups
1598 * Introduce remote-fs-pre.target which is ordered before all
1599 remote mounts and may be used to start services before all
1602 * Added Mageia support
1604 * Add bash completion for systemd-loginctl
1606 * Actively monitor PID file creation for daemons which exit in
1607 the parent process before having finished writing the PID
1608 file in the daemon process. Daemons which do this need to be
1609 fixed (i.e. PID file creation must have finished before the
1610 parent exits), but we now react a bit more gracefully to them.
1612 * Add colourful boot output, mimicking the well-known output
1613 of existing distributions.
1615 * New option PassCredentials= for socket units, for
1616 compatibility with a recent kernel ABI breakage.
1618 * /etc/rc.local is now hooked in via a generator binary, and
1619 thus will no longer act as synchronization point during
1622 * systemctl list-unit-files now supports --root=.
1624 * systemd-tmpfiles now understands two new commands: z, Z for
1625 relabelling files according to the SELinux database. This is
1626 useful to apply SELinux labels to specific files in /sys,
1629 * Output of SysV services is now forwarded to both the console
1630 and the journal by default, not only just the console.
1632 * New man pages for all APIs from libsystemd-login.
1634 * The build tree got reorganized and a the build system is a
1635 lot more modular allowing embedded setups to specifically
1636 select the components of systemd they are interested in.
1638 * Support for Linux systems lacking the kernel VT subsystem is
1641 * configure's --with-rootdir= got renamed to
1642 --with-rootprefix= to follow the naming used by udev and
1645 * Unless specified otherwise we'll now install to /usr instead
1646 of /usr/local by default.
1648 * Processes with '@' in argv[0][0] are now excluded from the
1649 final shut-down killing spree, following the logic explained
1651 http://www.freedesktop.org/wiki/Software/systemd/RootStorageDaemons
1653 * All processes remaining in a service cgroup when we enter
1654 the START or START_PRE states are now killed with
1655 SIGKILL. That means it is no longer possible to spawn
1656 background processes from ExecStart= lines (which was never
1657 supported anyway, and bad style).
1659 * New PropagateReloadTo=/PropagateReloadFrom= options to bind
1660 reloading of units together.
1662 Contributions from: Bill Nottingham, Daniel J. Walsh, Dave
1663 Reisner, Dexter Morgan, Gregs Gregs, Jonathan Nieder, Kay
1664 Sievers, Lennart Poettering, Michael Biebl, Michal Schmidt,
1665 Michał Górny, Ran Benita, Thomas Jarosch, Tim Waugh, Tollef
1666 Fog Heen, Tom Gundersen, Zbigniew Jędrzejewski-Szmek