[Bug 902779] New: [Factory] Long boot delay "a start job is running for..."

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Bug ID: 902779 Summary: [Factory] Long boot delay "a start job is running for..." Classification: openSUSE Product: openSUSE Factory Version: 201410* Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening@forge.provo.novell.com Reporter: sb56637@gmail.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- Hi, My openSUSE Factory installation was taking at least 60 seconds longer than normal to boot. The delay always occurred on every boot. The delay occured before initializing the display manager with a message: "A start job is running for", and it would alternate between four different services: X display manager, Avahi, NetworkManager, and Modem Manager. I tried disabling the Modem Manager service, but it still kept hanging on Avahi and NetworkManager. I eventually tried disabling and later re-enabling the NetworkManager services, and it now boots quickly without hanging. I don't really know why this happened or what I did to fix it. Please let me know what additional information you need to prevent this bug from happening. Thanks a lot! -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Bernhard Wiedemann <bwiedemann@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |bwiedemann@suse.com, | |sb56637@gmail.com Flags| |needinfo?(sb56637@gmail.com | |) --- Comment #1 from Bernhard Wiedemann <bwiedemann@suse.com> --- So is this a laptop? Please also attach output from hwinfo --network Can you still reproduce the problem somehow? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #2 from Rahim Barbamas <sb56637@gmail.com> --- Hi, thanks for looking into this. Yes, this is a slightly older Acer 4810T x86-64 laptop with 4GB of RAM and a spinning hard drive. It does still appear to have some kind of conflict. After "Started Login Manager", there is a delay of several seconds, and then "A start job is running for" alternating between X Display Manager and NetworkManager, with a little ASCI animation of delay. I'm attaching my hwinfo. Between "Started Login Manager" and getting past the "Start job is running for" is about 16 seconds, which is a significant percentage of the total time that a systemd Linux system usually takes to boot on this laptop. So I would consider this to be a bug. I thought that systemd was supposed to have parallel boot processes, so this kind of waiting on one service for another to finish shouldn't happen during boot? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #3 from Rahim Barbamas <sb56637@gmail.com> --- Created attachment 611471 --> http://bugzilla.opensuse.org/attachment.cgi?id=611471&action=edit hwinfo --network -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Bruno Pesavento <bpesavento@infinito.it> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |bpesavento@infinito.it --- Comment #4 from Bruno Pesavento <bpesavento@infinito.it> --- I had even worse delay problems on a laptop with slow disk (SATA-I) due to systemd flushing the system journal to disk just when Xorg is supposed to start. See if disabling journal to disk helps, writing "Storage=volatile" in /etc/systemd/journald.conf . Not a bug proper, just a wrong default setting not suited to laptops. In my case "Storage=auto" and "SystemMaxUse=160M" gave a workable journal with little collateral damage. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Rémi Benoit <r3m1.benoit@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |r3m1.benoit@gmail.com -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Bernhard Wiedemann <bwiedemann@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|bnc-team-screening@forge.pr |systemd-maintainers@suse.de |ovo.novell.com | --- Comment #5 from Bernhard Wiedemann <bwiedemann@suse.com> --- so possibly a problem from the default journal config -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Sebastian Kuhne <sebastiankuhne1@googlemail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |sebastiankuhne1@googlemail. | |com, | |systemd-maintainers@suse.de Flags| |needinfo?(systemd-maintaine | |rs@suse.de) --- Comment #7 from Sebastian Kuhne <sebastiankuhne1@googlemail.com> --- Created attachment 618560 --> http://bugzilla.opensuse.org/attachment.cgi?id=618560&action=edit hwinfo --network Dear all, I am facing a similar issue on openSUSE 13.2 (fresh installation, no update). Attached is the ouput of hwinfo --network. Hope this helps. Best regards Sebastian -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #8 from Sebastian Kuhne <sebastiankuhne1@googlemail.com> --- Created attachment 618561 --> http://bugzilla.opensuse.org/attachment.cgi?id=618561&action=edit boot.log Additionally, I am attaching the boot.log file. Are there any more outputs required for analysis? Best regards Sebastian -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #10 from Thomas Blume <thomas.blume@suse.com> --- (In reply to Bruno Pesavento from comment #4)
Not a bug proper, just a wrong default setting not suited to laptops. In my case "Storage=auto" and "SystemMaxUse=160M" gave a workable journal with little collateral damage.
Actually, there is a bug report open for changing the defaults. You might want to attach to bug #911347 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #34 from Dr. Werner Fink <werner@suse.com> --- (In reply to Sebastian Kuhne from comment #33) Nevertheless: # rpm -q --qf '%{DESCRIPTION}\n' -f /usr/lib/systemd/system/wpa_supplicant.service wpa_supplicant is an implementation of the WPA Supplicant component, i.e., the part that runs in the client stations. It implements key negotiation with a WPA Authenticator and it controls the roaming and IEEE 802.11 authentication/association of the wlan driver. Authors: -------- Jouni Malinen <jkmaline@cc.hut.fi> -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #43 from Sebastian Kuhne <sebastiankuhne1@googlemail.com> --- Created attachment 619941 --> http://bugzilla.opensuse.org/attachment.cgi?id=619941&action=edit journalctl_axb_root_2015_01_19.txt -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #44 from Sebastian Kuhne <sebastiankuhne1@googlemail.com> --- (In reply to Thomas Blume from comment #42)
(In reply to Sebastian Kuhne from comment #41)
OK, executed and attached, with udev_log="debug" and as root.
Unfortunately, the udev verbose log is still not visible in the logs. However, the problem is visible, after mounting swap and home, there is a delay of about 45s:
-->-- Jan 15 16:21:53 linux-1r76 systemd-fsck[430]: /dev/sda6: clean, 75903/18096128 files, 64059249/72373334 blocks Jan 15 16:21:53 linux-1r76 kernel: Adding 2104476k swap on /dev/sda1. Priority:-1 extents:1 across:2104476k FS Jan 15 16:21:55 linux-1r76 kernel: EXT4-fs (sda6): mounted filesystem with ordered data mode. Opts: (null) Jan 15 16:22:40 linux-1r76 kernel: usbcore: registered new interface driver snd-usb-audio Jan 15 16:22:40 linux-1r76 lvm[444]: WARNING: lvmetad is running but disabled. Restart lvmetad before enabling it! Jan 15 16:22:41 linux-1r76 lvm[444]: No volume groups found Jan 15 16:22:41 linux-1r76 lvm[446]: WARNING: lvmetad is running but disabled. Restart lvmetad before enabling it! Jan 15 16:22:41 linux-1r76 lvm[446]: No volume groups found --<--
Strange are the lvm messages. It seems you don't have any lvm setup. Can you confirm this? If so, please edit /etc/lvm/lvm.conf and change the use_lvmetad parameter to:
use_lvmetad = 0
In addition, please deactivate the lvmetad socket via:
systemctl disable lvm2-lvmetad.socket
I have tried to reproduce the problem with the latest factory build (20150107) but couldn't see any of the above problems. So I would recommend that you update your test system to latest factory before you do the next tests.
If it still doesn't work with the above settings, please provide the output of journalctl -axb again (please keep udev_log="debug" in /etc/udev/udev.conf).
Hi Thomas, use_lvmetad = 0 was already activated. I have executed "systemctl disable lvm2-lvmetad.socket", and the output of journalctl -axb attached (udev_log="debug"). Sorry for this issue. Many thanks for your help! Best regards Sebastian -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 --- Comment #46 from Sebastian Kuhne <sebastiankuhne1@googlemail.com> --- (In reply to Thomas Blume from comment #45)
(In reply to Sebastian Kuhne from comment #44)
use_lvmetad = 0 was already activated.
I have executed "systemctl disable lvm2-lvmetad.socket", and the output of journalctl -axb attached (udev_log="debug").
Sorry for this issue. Many thanks for your help!
Hm, still lvmetad was started:
--<-- Jan 19 09:05:04 linux-1r76 lvm[384]: WARNING: lvmetad is running but disabled. Restart lvmetad before enabling it! -->--
This shouldn't happen if lvm2-lvmetad.service and lvm2-lvmetad.socket are disabled. The only possibility, I can think of, is that it was started in the initrd. Did you run mkinitrd after disabling lvm2-lvmetad.socket?
However, the boot times look ok now. kdm gets started after 21s:
-->-- Jan 19 09:04:58 linux-1r76 kernel: Linux version 3.16.7-7-desktop (geeko@buildhost) (gcc version 4.8.3 20140627 [gcc-4_8-branch revision 212064] (SUSE Linux) ) #1 SMP PREEMPT Wed Dec 17 18:00:44 UTC 2014 (762f27a) [...] Jan 19 09:05:19 linux-1r76 display-manager[652]: Starting service kdm..done --<--
Can you confirm that this fixes your issue?
Good morning, no I didn't run mkinitrd after disabling lvm2-lvmetad.socket. But, as you said, the boot time looks OK now. I can confirm that this fixes my issue. I'll keep lvm2-lvmetad.socket disabled. Thank you for your great support. Best regards, Sebastian -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=902779 Sebastian Kuhne <sebastiankuhne1@googlemail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(sebastiankuhne1@g | |ooglemail.com) | -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@novell.com