summaryrefslogtreecommitdiffstats
path: root/perl-install/services.pm
Commit message (Collapse)AuthorAgeFilesLines
* Use saned description for saned.socket (there is no standalone service).Martin Whitaker2020-09-161-1/+1
|
* Use supplied name for looking up service descriptions.Martin Whitaker2020-09-161-6/+1
| | | | | Better to add new descriptions for socket-activated services rather than try to reuse existing ones.
* Fix comment.Martin Whitaker2020-09-161-1/+1
|
* Simplify and improve services::services().Martin Whitaker2020-09-161-55/+6
| | | | This filters out static services that can't be enabled or disabled.
* Show disabled systemd sockets and timers in GUI.Martin Whitaker2020-09-151-1/+1
| | | | | | | Several systemd units that the user may wish to enable or disable (e.g. fstrim.timer, saned.socket) are not wanted by the multi-user or graphical targets. So also show units that are wanted by the sockets or timers targets (better too much than too little).
* Fix GUI start/stop for systemd sockets and timers.Martin Whitaker2020-09-151-4/+1
|
* Include systemd sockets and timers in lists returned by services::services().Martin Whitaker2020-09-151-14/+24
|
* Support systemd sockets and timers in service manipulation routines.Martin Whitaker2020-09-151-10/+21
|
* Add --plain to systemctl list-unitsPascal Terjan2020-08-071-1/+1
| | | | | Since systemd 246 it adds some characters at the beginning of the line which break parsing when --plain is not set (mga#27046)
* Fix services::running_systemd().Martin Whitaker2020-03-211-1/+1
| | | | | | systemd has migrated to cgroups v2, so the old test no longer works. Instead, test for the existence of /run/systemd/system as described in https://www.freedesktop.org/software/systemd/man/sd_booted.html.
* Remove leftover debug message.Martin Whitaker2018-03-121-1/+0
|
* update service listThierry Vignaud2016-07-031-3/+3
|
* sort service categories, putting "Other" at endThierry Vignaud2016-06-221-1/+6
| | | | Resolves: mga#6286
* list a couple new servicesThierry Vignaud2016-06-221-1/+2
| | | | (because of previous commits)
* fix getting systemd services status at installThierry Vignaud2016-06-221-2/+13
| | | | | | | | | | | Some services such as shorewall are not "wantedby". However, as systemd is not running during installation, we failed to detect if whether they're enabled or not as the legacy fallback doesn't handle such services Let's try hard to query status of services that are not "wanted" but we're checking for. Thus fixing updating firewall status after configuration (mga#17288)
* silent rpm -qThierry Vignaud2016-06-221-1/+1
|
* update service list for current stuffThierry Vignaud2016-06-181-3/+3
|
* mosix/powertweak/... do not exist anymoreThierry Vignaud2016-06-181-3/+3
|
* fix wrong renaming of mandrake_firsttime serviceThomas Backlund2015-03-211-1/+1
|
* correctly recognize shorewall6 service + descriptionMaarten Vanraes2014-12-051-1/+2
|
* fix drakxservices not showing all services due to systemctl output changeColin Guthrie2014-10-201-2/+2
|
* kill $Id:$ tagsThierry Vignaud2014-06-071-1/+1
| | | | they're obsolete since switching from CVS/SVN to git...
* services.pm: Use no-reload mode when interacting with systemd.Colin Guthrie2014-01-141-4/+7
| | | | | | | | | | Basically, there seems to be an issue right now if we do lots of reloads. It seems to be somewhat racy and systemd gets generally confused when we do this during boot (which we do) and trigger a reload. The reload isn't strictly needed, as we do also explicitly start the services we enable (with --no-block so that we don't cause a deadlock) so all *should* be 99% OK. Using it at runtime might mean we don't have fully up-to-date enablement status, but being able to boot wins at the moment.
* simplify: just use tooltips for service descriptionsThierry Vignaud2013-11-241-19/+1
|
* workaround a gtk+3 regression (bgo#711873)Thierry Vignaud2013-11-241-1/+1
| | | | See https://bugzilla.gnome.org/show_bug.cgi?id=711873
* fix "undefined value for mandatory argument..."Thierry Vignaud2013-11-241-1/+1
|
* perlish ->window => ->get_windowThierry Vignaud2013-11-241-1/+1
|
* use Gtk3 namespace instead of Gtk2Thierry Vignaud2013-11-241-13/+13
|
* use (my|u)gtk3 instead of *tk2Thierry Vignaud2013-11-241-4/+4
|
* Add chrony support to drakclock mga#11092David Walser2013-11-171-0/+1
| | | | | | | | | | | v2 (by Colin Guthrie) * Fix external variable reference (my -> our + $ when using it) * Drop systemd cgroup filesystem check as the path has changed now (prefer "systemctl is-active foo.service" instead) * Add iburst to chrony config server definitions (triggers sync at startup) * Don't use ntpdate with chrony - rather chronyc waitsync * Don't bother checking for ntp in finish-install - just use chrony
* Fix path after UsrMoveNicolas Lécureuil2013-09-031-1/+1
|
* list 'shorewall-ipv6' along 'shorewall'Thierry Vignaud2013-04-281-1/+1
|
* simplifyThierry Vignaud2013-03-041-3/+1
|
* drop unused parameterThierry Vignaud2013-03-041-1/+1
|
* adapt to new help layout (mga#5634)Thierry Vignaud2012-12-281-1/+1
|
* (has_systemd) fix checking systemd presence after systemd merge of subpackagesThierry Vignaud2012-08-131-1/+1
|
* drakxservices: when listing services ensure disabled services that can be ↵Colin Guthrie2012-07-261-0/+18
| | | | enabled are shown (mga#6843)
* (systemd_unit_exists) explain why we use -lThierry Vignaud2012-06-031-0/+1
|
* (systemd_unit_exists) factorizeThierry Vignaud2012-05-231-4/+6
|
* services: Use no-block mode for service start/stop under systemd.Colin Guthrie2012-04-291-1/+1
| | | | | | | | | | | | | | This is perhaps the same issue as the 'cpufreq' bug on first boot, but by changing things to use systemctl directly, I have inadvertantly bypased the solution and reintroduced the bug. As there is likely no reason to block, we will simply use no-block mode unconditionally. Longer term, we may need to decide under what context we are restarting/starting a given service (i.e. in an interactive GUI vs from a script) and use the appropriate mode with appropriate user feedback. https://bugs.mageia.org/show_bug.cgi?id=4439
* perl_checker cleanupsThierry Vignaud2012-04-251-2/+2
|
* services: Read systemd service information.Colin Guthrie2012-04-241-16/+82
| | | | | | | | | | | | | | | | | | | | | | This adds support for two different methods of working with systemd. The first, and more complete method is to query systemd for information about whether a service is enabled at boot. This mode requires that we are running under systemd and thus only works for machines in interactive mode post-install. For the installer, we also support a more rudamentary method which scans for unit files which contain simple install rules (i.e. only those services which are part of the typical multi-user or graphical targets (corresponding to runlevels 3 and 5 in the old days). This method will not show all services but it should be enough for the installer. https://bugs.mageia.org/show_bug.cgi?id=3253 https://bugs.mageia.org/show_bug.cgi?id=3740 https://bugs.mageia.org/show_bug.cgi?id=4910 https://bugs.mageia.org/show_bug.cgi?id=5122
* services: Use systemctl to determin if a service is active under systemd.Colin Guthrie2012-04-241-1/+5
|
* services: Check for systemd unit when determining if a service exists.Colin Guthrie2012-04-241-1/+1
| | | | This could be made more complex, but it'll do for now.
* services: Use the appropriate tool to start/stop servicesColin Guthrie2012-04-241-1/+5
|
* services: Whitespace and commentsColin Guthrie2012-04-241-7/+7
|
* services: Use systemd utility functions to determin how to handle ↵Colin Guthrie2012-04-241-2/+6
| | | | enabling/disabling of services.
* services: Add a couple systemd untility functions.Colin Guthrie2012-04-241-0/+8
|
* services: Do not rely on /var/lock/subsys/$service to determin if a service ↵Colin Guthrie2012-04-241-1/+1
| | | | is running or not.
* services: WhitespaceColin Guthrie2012-04-241-3/+3
|