quidnunc | I'm getting kicked back to lxdm greeter after a few seconds after entering my user/pass. I don't see anything .xsession-errors, where should I be looking to debug? | 00:01 |
---|---|---|
TJ- | quidnunc: which GPU and driver? | 00:01 |
TJ- | quidnunc: there were some problems of that nature with nvidia very recently | 00:01 |
quidnunc | Intel | 00:05 |
quidnunc | TJ-: Intel | 00:05 |
TJ- | quidnunc: did you check that your $USER owns $HOME/.{X,ICE}authority ? | 00:05 |
quidnunc | TJ-: they do | 00:06 |
slidinghorn | quidnunc: try moving to a TTY terminal (Ctrl + Alt + F1) login using your username & pass, and then try startx | 00:06 |
TJ- | oh and I think they're under /run/user/$UID/ for Gnome/gdm | 00:06 |
quidnunc | TJ-: What are under /run/user ? | 00:07 |
TJ- | quidnunc: the files I mentioned. It's a tmpfs so they don't persist after a reboot, which prevents the commone issue of users running root commands and causing those to be owned by 'root' and thus preventing fresh log-ins | 00:09 |
quidnunc | slidinghorn: I have an X session active with another user with a different config, is that going to kill that session? | 00:09 |
slidinghorn | quidnunc: in that case do startx -- :1 instead | 00:20 |
quidnunc | slidinghorn: So that fails and also kills my keyboard | 00:41 |
quidnunc | so I have to reboot | 00:41 |
quidnunc | slidinghorn: .local/share/xorg/Xorg.log.1 says "failed to start service org.freedesktop.login1" | 00:42 |
quidnunc | slidinghorn: and some errors about failing to open /dev/fb0 and /dev/dri/card(?) | 00:42 |
slidinghorn | quidnunc: those errors could be the key to figuring out the issue...can you try again and post the errors? (pastebin if multiple lines) | 00:52 |
=== brainwash_ is now known as brainwash | ||
quidnunc | slidinghorn: https://paste.ubuntu.com/p/r2h3FPyJCd/ | 01:19 |
quidnunc | sorry it took so long, difficult to do this in terminal and because I'm logged in as another user | 01:20 |
slidinghorn | you said this is lxdm? | 01:26 |
quidnunc | yes | 01:26 |
quidnunc | slidinghorn: Last time, when I upgraded to 17.10 there was a problem with the session names containing spaces | 01:27 |
quidnunc | but that doesn't seem to be the problem here | 01:27 |
slidinghorn | quidnunc: can you pastebin this log: /home/quid/.local/share/Xorg.1.log | 01:29 |
quidnunc | slidinghorn: That is the pastebin for that log | 01:29 |
slidinghorn | ok, just wanted to make sure, as it referenced it | 01:29 |
slidinghorn | quidnunc: this may be a stupid question, but are you able to log in if the other user doesn't have an active x session? | 01:35 |
quidnunc | slidinghorn: No I tried | 01:36 |
slidinghorn | in that tty terminal, can you cd to /dev/ and do ls -l fb0 | 01:40 |
quidnunc | crw-rw---- 1 root video 29, 0 Apr 15 20:31 fb0 | 01:40 |
slidinghorn | okay...now: sudo usermod -a -G video quid (assuming quid is the username) | 01:40 |
slidinghorn | after you do that, log out, back in and see if you're able to load an X session | 01:41 |
quidnunc | slidinghorn: Okay, I can try that, I just want to look at the Xorg.log when logging in through the greeter without startx first | 01:46 |
slidinghorn | that's fine - I'm hoping adding your user to the video group would allow you to do that anyway...the startx thing was moreso a fallback in case there was still a problem | 01:47 |
quidnunc | slidinghorn: I have my doubts, the user that works isn't part of video | 01:48 |
quidnunc | Okay, it doesn't update Xorg.1.log that way | 01:49 |
quidnunc | I think something is wrong with systemd | 01:50 |
quidnunc | slidinghorn: Adding to video group doesn't work (at least with user switching). I don't want to try startx because it kills my system | 01:57 |
slidinghorn | thinking... | 01:59 |
quidnunc | It's okay, I think I'm going to give up for now... | 01:59 |
quidnunc | Maybe I'll try to get gdm (which has other issues) working | 02:00 |
quidnunc | thanks for the help, good night | 02:04 |
mattfly1 | wow im so happy with the daily build | 02:40 |
mattfly1 | just wish hibernation could work now | 02:40 |
donofrio | two machines next to each other why would one load 16 updates that the left machine doesn't pulldown (stated "timeout" a few times before I lost like 10 built in repo checks) on left system right system pulled 16 | 03:16 |
donofrio | both 18.04 fwiw | 03:17 |
donofrio | same sources.list | 03:18 |
flocculant | 9392 | 04:40 |
flocculant | stupid mouse focus | 04:41 |
slidinghorn | change your PIN :P | 04:41 |
flocculant | ha ha | 04:41 |
=== kallesbar_ is now known as kallesbar | ||
slidinghorn | I've installed sntp to sync my clock with an NTP server, but I've stumbled upon a couple questions: 1) Does this have a systemd unit to run? (man page doesn't specify) 2) The package page (https://packages.ubuntu.com/bionic/sntp) shows the source package as being the regular ntp package. Am I missing something here? | 05:14 |
slidinghorn | cancel the 1st question...the 2nd I'm still curious about | 05:15 |
alkisg | slidinghorn: https://www.meinbergglobal.com/english/faq/faq_37.htm | 05:19 |
alkisg | In other words, you already have an ntp client, and why would you ever want to use sntp instead. | 05:19 |
alkisg | But if you have specific reasons to, it's there | 05:19 |
slidinghorn | alkisg: I don't need the full functionality of the full ntp package (which isn't installed by default, by the way) | 05:20 |
alkisg | It's managed by systemd nowadays | 05:20 |
alkisg | And you can control the aspects with timedatectl | 05:21 |
alkisg | timedatectl => Network time on: yes | 05:21 |
alkisg | Don't you already have that running? | 05:21 |
slidinghorn | alkisg: then it's doing a crappy job, haha - also, if I go to Time & Date Settings from my clock in my panel, it isn't set to update. Setting it to do so alerts me that I need to enable an ntp client | 05:22 |
alkisg | You may or may not have found some bugs to report there. Which desktop environment is that? | 05:23 |
slidinghorn | I'm running Studio 18.04 - so it's XFCE | 05:23 |
alkisg | OK, then you'd need to file a bug report against its time/date settings, so that they support the new systemd method of managing time | 05:24 |
alkisg | When you say it's doing a crappy job, how is that? | 05:24 |
slidinghorn | I'll often happen to watch my clock skip a couple minutes...just a bit ago, it went from 12:59 to 1:02 | 05:25 |
alkisg | If the correct time is near 1:02, then it did a good job | 05:25 |
alkisg | ntp clients are supposed to gradually go near the correct time, not instantaneously | 05:26 |
slidinghorn | being off by a few minutes at any point irks me though...it's obviously nothing major, but I don't like the idea of my clock being off | 05:28 |
alkisg | If you installed ntpd, you told systemd-timesyncd to stop running | 05:29 |
alkisg | So in that case, maybe that's what broke your setup | 05:29 |
slidinghorn | i didn't | 05:29 |
alkisg | What's the output of this? systemctl status systemd-timesyncd.service | 05:29 |
slidinghorn | I installed sntp after I saw that happen...only pulled in sntp & libopt25 | 05:29 |
slidinghorn | it's running | 05:30 |
alkisg | And the messages say that? It synced? When? | 05:30 |
slidinghorn | 2 days ago | 05:30 |
slidinghorn | presumably on startup | 05:30 |
alkisg | It's supposed to do an initial sync, AND periodically | 05:31 |
alkisg | So if you think you found a bug there, it should be reported | 05:31 |
slidinghorn | on it... | 05:31 |
alkisg | I haven't heard any issues that it's a bad ntp client or anything | 05:31 |
slidinghorn | Would this be directly against systemd? I don't see a package on launchpad for timesyncd | 05:36 |
alkisg | Yes, against systemd; although so far I got "progressively syncs with the correct time", which is what the protocol says that the clients should do | 05:38 |
alkisg | slidinghorn: man timesyncd.conf for customization | 05:41 |
slidinghorn | reading my conf file now...says max interval is 2048 seconds | 05:41 |
slidinghorn | ...actually it's commented out... | 05:42 |
* BionicMac sees the light... I accidently stumbled onto an entry in kern.log that something called "apparmor" was putting a profile in place and DENY that application access to resolv.conf... interestingly enough I have been troubleshooting that app ( clamav freshclam ) and boom... now I see... hmmm | 07:45 | |
* BionicMac dances and starts learning about AppArmor. | 07:45 | |
BionicMac | so freshclam is not to blame.. neither is my dns broken. AppArmor has it in jail. =) | 07:46 |
BionicMac | what a relief. I see many profiles in place and several of them I have been trying to trace down the issue... and here it is. | 07:47 |
vbotka | BionicMac, FWIW, to automate your config, I added configuration of apparmor to https://galaxy.ansible.com/vbotka/linux-postinstall/ recently. | 08:10 |
BionicMac | vbotka: Thank you kindly. | 08:22 |
BionicMac | vbotka: audit(1523870863.242:163): apparmor="ALLOWED" operation="open" profile="/usr/bin/freshclam" | 09:31 |
BionicMac | ^^ That is a beautiful sight. =) | 09:31 |
BionicMac | That being said, time for sleep. Thanks again vbotka. I need to fine tune ansible variables, playbook & inventory. | 09:33 |
vbotka | BionicMac, you're welcome. Let me know if you have any troubles, or questions about the role. | 09:35 |
BionicMac | vbotka: I will be editing ~/.ansible/roles/vbotka.linux-postinstall/vars/main.yml next. I may need a few hints along the way. Take care now. I will pick this up again after some sleep. For now, getting clamav/freshclam out of jail and downloading updated virus definitions feels like an accomplishment. | 09:39 |
BionicMac | I feel like AppArmor has been the cause of an intermittent vpn connection issue also. That will be another victory. I'm out. | 09:40 |
=== JanC_ is now known as JanC | ||
quidnunc | Why can't I install avahi-daemon? journalctl -xe says "Failed to activate service 'org.freedesktop.Packagekit': timed out'" | 16:21 |
lotuspsychje | TJ-: was that bug you found related on this ^ | 16:26 |
quidnunc | "systemctl status avahi-daemon.service" says "Failed to connect to bus: No such file or directory" | 16:28 |
TJ- | Not that I recall, this looks like a different bug | 16:30 |
quidnunc | and gnome depends on avahi-daemon | 16:31 |
quidnunc | running avahi-daemon --debug seems fine | 17:16 |
=== simonizor is now known as Guest4960 | ||
=== Guest4960 is now known as simonizor | ||
jluc | hello | 19:43 |
jluc | where is the bug tracker to follow the issues before +1 release ? | 19:43 |
jluc | is it may be https://launchpad.net/ubuntu/+milestone/ubuntu-18.04 ? | 19:45 |
jluc | https://bugs.launchpad.net/ubuntu | 19:45 |
enyc | Hrrm | 20:06 |
enyc | I wonder why ubunt 18.04 apparently supposedly prefers swapfile over swap partition | 20:07 |
nacc | enyc: as in, what does that, or why that was the choice? | 20:11 |
enyc | nacc: as in, why the change in that regard | 20:16 |
flocculant | enyc: http://blog.surgut.co.uk/2016/12/swapfiles-by-default-in-ubuntu.html | 20:18 |
flocculant | been about since 17.04 btw - not for 18.04 | 20:19 |
enyc | flocculant: coo i wonder if thats ' pritucere of an XMS or EMS card =) | 20:22 |
roothorick | WINS lookup isn't working. libnss-winbind and winbind are installed, winbindd is running, "wins" is at the end of the appropriate line in nsswitch.conf, wbinfo -N is able to resolve the host, but trying to ping a host that needs to be looked up by WINS gives a useless "System Error" message and trying to ssh to the host gives "Could not resove hostname <x>: No such file or directory" | 20:22 |
roothorick | and I can't for the life of me find a way to get more useful information. No messages appear in the system journals. | 20:24 |
flocculant | enyc: I have no idea what that means | 20:26 |
enyc | flocculant: the artcile poi pointed out (which im reading furethr into its comments{, starts with picture of a 4mb ram on ISA-card | 20:27 |
roothorick | There are multiple 16.04 machines on the same network which are having no problems resolving WINS hosts... | 20:27 |
flocculant | enyc: oh right | 20:28 |
flocculant | tbh not that bothered how swap is served up - it still uses swap partition here as it already exists | 20:28 |
roothorick | upon further investigation, it's just not consistent. Right now only one host does that... I don't get it | 20:57 |
CoJaBo | So, one of my systems still shows this: Welcome to Ubuntu Bionic Beaver (development branch) (GNU/Linux 4.15.0-15-generic x86_64) | 23:23 |
CoJaBo | The other one says beta now; is there a reason for the difference? | 23:24 |
nacc | CoJaBo: says beta where? | 23:41 |
gabefair | Is @flocculant here? His https://wiki.ubuntu.com/LightDM page has not been updated since Ubuntu switched to systemd rather than Upstart | 23:48 |
nacc | gabefair: ... flocculant --^ | 23:49 |
CoJaBo | nacc: I'm not sure; I just tabbed back, and it's not there anymore | 23:49 |
nacc | CoJaBo: well 'beta' is some point in time | 23:50 |
nacc | CoJaBo: it should always say development branch in the above, until release, iirc | 23:50 |
CoJaBo | I'm not 100% sure I wasn't just looking at the login prompt for a non-ubuntu system or something.. | 23:50 |
nacc | CoJaBo: also possible | 23:52 |
CoJaBo | Very strange tho, as I don't have any local systems that aren't some ubuntu, and I don't think I've closed any tabs. Maybe I'm just going ınsane. | 23:52 |
valorie | CoJaBo: in my case, it's always a possibility | 23:53 |
valorie | :-) | 23:53 |
CoJaBo | The server install crashed again last night; no idea why, but I hope a bug that's been fixed | 23:54 |
CoJaBo | I hadn't updated it since before the alpha was declared, so there's that.. | 23:55 |
valorie | try the daily? | 23:56 |
CoJaBo | Yeh, it was still in daily | 23:56 |
CoJaBo | I'm used prealphas being called nightlies; why did ubuntu have to go and be different lol | 23:56 |
CoJaBo | What's kinda bizarre and a bit disturbing is that the Bionic server crashing took down my laptop, and the laptop is running 17.10 stable. how tho | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!