bozonius | oh well, this is fun. My VM is locked up, it seems. I can ping it, but I can't get the mouse or keyboard to work? | 01:26 |
---|---|---|
gnarface | can you ssh into it? | 01:28 |
bozonius | the VM is jessie, and it was running fine... at least until a moment ago when I went to check on it. I got a email at 11:00 am PST saying that bareos (backup) could not talk to it | 01:29 |
bozonius | nope. | 01:29 |
gnarface | mouse & keyboard usually don't lock up on their own, usually its the video driver or Xorg that is actually freezing | 01:29 |
bozonius | Not even sure if sshd service is running | 01:29 |
gnarface | if there was a full i/o lock, it wouldn't respond to ping either | 01:29 |
bozonius | the clock stopped at 4:29pm PST | 01:30 |
gnarface | weird | 01:30 |
gnarface | did it go to sleep? could it do that? | 01:30 |
bozonius | checking my other VMs... | 01:30 |
bozonius | the rest are all working. Just devuan jessie | 01:31 |
bozonius | maybe. | 01:31 |
bozonius | with the system being totally unresponsive, it is hard to know what is going on (or not) | 01:32 |
bozonius | is sshd enabled by default on devuan jessie? | 01:32 |
bozonius | I don't think I've ever changed it... idk 100% though | 01:32 |
gnarface | it is not included by default | 01:33 |
gnarface | uh | 01:33 |
gnarface | its not included by default in a minimal install | 01:33 |
gnarface | one of the taskman selections might enable it though | 01:33 |
bozonius | a lot of snpp on vboxnet0, but not for that address... | 01:34 |
gnarface | well you can port scan it to look for an open ssh port if you're not sure | 01:34 |
gnarface | nmap -p 22 [address] | 01:35 |
bozonius | only open port is http | 01:35 |
bozonius | (80) | 01:35 |
gnarface | so its still serving web pages? | 01:35 |
gnarface | seems like its gotta be a video problem then | 01:35 |
bozonius | lemme see | 01:35 |
gnarface | you're running Xorg on that guest, right? | 01:36 |
bozonius | yep | 01:36 |
bozonius | I can see the desktop, but it's frozen | 01:36 |
gnarface | Xorg handles all the keyboard&mouse input, so if it freezes while running then your mouse and keyboard freeze too | 01:36 |
bozonius | I had this happen with centos, actually, long time ago, and it recovered after a long time | 01:36 |
gnarface | yea, sometimes low memory issue | 01:36 |
gnarface | sometimes it can get past it | 01:36 |
bozonius | 3G allocated to that VM | 01:36 |
bozonius | It wasn't maxed out last night when I last used it | 01:37 |
bozonius | (I do check htop frequently to check for memleaks, etc) | 01:37 |
gnarface | hmmm | 01:37 |
bozonius | 3G and (I think) 1G swap? | 01:37 |
bozonius | very little of the swap ever gets used though | 01:38 |
bozonius | oh... | 01:39 |
bozonius | the host (centos 6.10 now, after recent update), is reporting a memory issue | 01:39 |
bozonius | that was yesterday | 01:39 |
bozonius | and I ignored it because i thought it might be one of those one-off things | 01:40 |
* bozonius should NEVER ignore kernel errors | 01:40 | |
bozonius | bad, bad, bozonius! | 01:40 |
bozonius | There was a kernel update with the rest of the update to 6.10. For now, gnarface, I am going say that is suspect! | 01:41 |
bozonius | and it is definitely related to Vbox | 01:41 |
bozonius | /home/vbox/vbox-5.2.14/src/VBox/VMM/VMMR0/HMSVMR0.cpp(5773) void hmR0SvmAdvanceRipHwAssist(VMCPU*, CPUMCTX*, uint32_t) | 01:42 |
bozonius | (smoking gun) | 01:42 |
bozonius | OTOH, there was also an update to vbox recently, though before the 6.10 update on the host | 01:42 |
bozonius | gnarface: I will consult the vbox folks before going any farther here. This is NOT a devuan issue | 01:46 |
bozonius | bbl | 01:46 |
gnarface | ok, noted | 01:47 |
bozonius | btw, nix on the host kernel update -- I haven't rebooted since it was updated | 02:11 |
bozonius | (prob need to) | 02:11 |
fsdfa | Xfce4-Netload-Plugin won't accept the 8 alpha-numeric string assigned to my usb wifi dongle. How do I corrent this? | 03:07 |
gnarface | interesting | 03:09 |
gnarface | does it give you any reasons? | 03:09 |
gnarface | also, what is the string, exactly? | 03:09 |
fsdfa | Xfce4-Netload-Plugin: Error in initializing: Interface was not found. | 03:10 |
gnarface | when you say "alpha-numeric" does that include spaces? | 03:10 |
fsdfa | ls /sys/class/net eth0 eth1 lo wlxa854b22d868d\ | 03:10 |
fsdfa | "wlxa854b22d868d" | 03:10 |
gnarface | hmm | 03:10 |
gnarface | that looks weird too | 03:11 |
gnarface | i thought only symlinks go in that directory | 03:11 |
gnarface | fsdfa: what does 'ifconfig' report? | 03:13 |
fsdfa | bash: ifconfig: command not found | 03:14 |
gnarface | sorry, i meant: /sbin/ifconfig | 03:14 |
fsdfa | eth0: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500 ether 00:15:c5:a9:e2:e8 txqueuelen 1000 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 0 bytes 0 (0.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 16 eth0:avahi: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500 inet 169.254.12.171 netmask 255.255 | 03:14 |
gnarface | ok that's interesting too | 03:15 |
gnarface | because an ubuntu tutorial addressing this same issue suggested that 'wlxa854b22d868d' string should have shown up in this output as well | 03:15 |
fsdfa | i don't think it listed evgerything | 03:15 |
fsdfa | the copy/paste | 03:15 |
gnarface | well try this to be sure: /sbin/ifconfig -a | 03:15 |
gnarface | oh | 03:15 |
gnarface | well you can paste it to me directly or you can paste it to paste.debian.net | 03:16 |
gnarface | but i don't think it'll change my evaluation | 03:16 |
gnarface | at this point, it *seems* like you just need to manually enter "wlxa854b22d868d" into the config panel, and that's what my google searches suggest the fix is, but since you have inferred that is what you were doing when you ran into this problem, something else must be wrong here.. | 03:17 |
gnarface | ... | 03:17 |
fsdfa | I already tried that and I get the error | 03:18 |
gnarface | did it work before? maybe that device requires firmware you don't have installed. | 03:18 |
fsdfa | hmm now it works :/ | 03:18 |
fsdfa | thanks | 03:18 |
gnarface | heh, i didn't do anything | 03:18 |
fsdfa | exit | 03:19 |
gnarface | lol you're welcome | 03:19 |
gnarface | now i'm more confused than he is | 03:19 |
Corin | Hey, so, since Ascii was a lot faster than Jessie, should the next release be expected shortly after Buster? | 05:56 |
Corin | Also, wow, way bigger channel than when I last came here. | 05:57 |
gnarface | Corin: don't expect anything and you won't ever be disappointed | 06:29 |
gnarface | i don't think there's any verdict or consensus or projections for Beowulf yet | 06:30 |
gnarface | from the talk i've heard, there is a lot more work to do than there was for ascii, but the same was said of ascii compared to jessie | 06:31 |
gnarface | in #debian they would say "sooner if you help :)" | 06:31 |
gnarface | some people have already reported limited success upgrading to Beowulf, but stuff is still known to be broken there too | 06:32 |
gnarface | for the moment you're probably better off running ascii | 06:33 |
gnarface | i've got a bunch of games working on ceres though, too | 06:33 |
gnarface | jessie -> ascii -> beowulf -> ceres | 06:33 |
Corin | Well, I'm still using Jessie right now, and fine with that, but yeah, I was planning to upgrade to ascii in the next couple days. | 06:38 |
Corin | But congrats on speeding up the process a bit. Can't imagine how hard it is to make everything work without systemd. | 06:40 |
eyalroz | Can someone help me "debug" why Cinnamon is crashing on startup (with ASCII)? | 08:55 |
gnarface | eyalroz: only maybe. got an error message? | 09:23 |
gnarface | eyalroz: afk again... it's better typically to just ask the whole question instead of asking permission, that way if someone takes you up on it while you're afk, they have enough information to at least try to help | 09:25 |
gnarface | it could be a lot of things | 09:25 |
gnarface | permissions, odd display hardware... | 09:25 |
gnarface | these days usually it is a conflict with video card drivers and compositing though | 09:26 |
gnarface | that's where i'd place my bet at least, given no other information | 09:26 |
gnarface | but a relevant error message from dmesg or the Xorg log would go a long way | 09:27 |
gnarface | if you're looking in the Xorg log (usually /var/log/Xorg.0.log or ~/.local/share/xorg/Xorg.0.log) then pay attention specifically to lines containing "(EE)" | 09:28 |
]BFG[ | we should make a fuck systemd song similar to weird AI windows 95 | 09:38 |
]BFG[ | fuck systemD� | 09:39 |
]BFG[ | fuck systemD�� | 09:39 |
]BFG[ | fuck systemD�� | 09:39 |
]BFG[ | fuck systemD� | 09:39 |
aitor | good morning | 09:41 |
eyalroz | gnarf_at_gollum: (and everybody else) - the question also appears here: https://unix.stackexchange.com/questions/454179/ | 10:45 |
eyalroz | The ghist of it is as follows: | 10:45 |
eyalroz | Cinnamon crashes on startup, and works only in fallback mode. | 10:46 |
eyalroz | My system has on-board graphics, and an nVIDIA card which is not used for graphics/display at all (only for CUDA work). | 10:46 |
eyalroz | Still it seems like nouveau is loaded (despite my best efforts); although I'm not sure nouveau is the cause of the Cinnamon crash. | 10:46 |
gnarface | eyalroz: i'm not that other guy, just fyi. i don't know if nouveau is the cause, but if it is an optimus laptop usually those pass through a Intel video card. so maybe what cinnamon is complaining about is the intel driver, or perhaps something missing from it | 10:50 |
gnarface | Intel video uses the mesa opengl libraries | 10:50 |
gnarface | they may not have been installed by default | 10:50 |
gnarface | but i think they'd be required for compositing to work | 10:50 |
gnarface | cinnamon's "fallback mode" means without compositing, am i right? | 10:51 |
aitor | gnarface: i've never heard about Cinnamon fallback mode, but about Gnome fallback mode (classic session) no longer included since Gnome 3.8 | 11:23 |
aitor | i'm finishing simple-netaid today | 11:24 |
aitor | time to take a coffee, bbl :) | 11:24 |
cr1mson_king | Devuan ASCII doesn't start the ssh-agent on login, right? I'm having to start it manually otherwise when cloning a repository through git, it doesn't ask for authentication and fails. | 14:01 |
cr1mson_king | I suppose on systemd distros they use gnome-keyring to start it, so it shows a graphical window asking for the private key passphrase. | 14:01 |
gnarface | well neither the ssh server or client packages auto-start it | 14:02 |
gnarface | that doesn't mean something else won't | 14:02 |
gnarface | maybe if you can find an alternate gui keyring widget? | 14:02 |
cr1mson_king | I'm reading online people start ssh-agent through their .profile files, in a way that it dies when user logs out to protect data. | 14:04 |
gnarface | seems possible | 14:08 |
gnarface | i think it drops a pid file by default if it's backgrounded without any command being run | 14:09 |
gnarface | you should be able to kill that from the bash logout script | 14:09 |
gnarface | maybe not, maybe it makes an environment variable | 14:10 |
gnarface | oh yea, the man page says it is SSH_AGENT_PID | 14:11 |
cr1mson_king | Yep, i'll probably do that | 14:16 |
cr1mson_king | the gitlab key has a standard name id_rsa, and it works without the agent, but the github key is named differently, and (idk?) it doesn't ask for password without the agent... | 14:17 |
cr1mson_king | Maybe the agent is already running, but the github key is not added. I need to try that. | 14:18 |
gnarface | well its really easy to see if the agent is running | 14:20 |
gnarface | ps aux |grep agent | 14:21 |
cr1mson_king | good, i'll try that at home | 14:24 |
fsmithred | I think you can add it to the startup apps in your desktop | 14:34 |
fsmithred | or, just run 'ssh-add /path/to/key' | 14:35 |
fsmithred | if that fails, you can run 'ssh-agent ssh-add /path/to/key' | 14:36 |
fsmithred | cr1mson_king, ^^^ | 14:36 |
gnarface | he wants to make sure it dies on logout too though | 14:37 |
cr1mson_king | fsmithred, sure, that would be easier | 14:37 |
gnarface | will the startup apps do that? | 14:37 |
gnarface | it backgrounds automatically so there isn't a window to close to imply exiting it | 14:38 |
fsmithred | in xfce settings, I see GPG Password Agent listed (not checked here) | 14:40 |
fsmithred | only ways I know to kill it are to kill it or log out | 14:41 |
fsmithred | I guess I don't need it. I use ssh every day, and I hardly ever have to invoke the agent. | 14:42 |
gnarface | i usually just open a terminal with it directly | 14:43 |
gnarface | then it exits automatically when you close that terminal | 14:43 |
fsmithred | that makes sense. | 14:43 |
fsmithred | I tend to leave my terminals open. I know I'm going to use them again soon. | 14:44 |
cr1mson_king | good point, trusting excessive automation is a recipe for trouble. Always best to have control, when it doesn't use too much of our time, I think. | 14:50 |
cr1mson_king | couldn't find snapd in the repositories. What happened to it? | 16:18 |
MinceR | my guess is it depends on systemd | 16:18 |
cr1mson_king | damn... | 16:18 |
gnarface | what is it? | 16:23 |
MinceR | https://packages.debian.org/stretch/snapd | 16:24 |
cr1mson_king | funny that flatpak doesn't have systemd listed as dependency, I thought it was even more closely tied to it, because Red Hat's involvement | 16:29 |
cr1mson_king | actually, it has libpam-systemd, which brings systemd | 16:30 |
eyalroz | gnarface: It's not a laptop, nor does it have "Optimus" - at least not that I know of. And - yes, I think "fallback mode" means without compositing. | 17:34 |
eyalroz | I can try to install mesa libraries,, but IIANM, there's some incompatibility between some nvidia-specific packages and mesa packages (perhaps w.r.t. OpenGL?) | 17:35 |
eyalroz | But beyond my specific case - it seems to me that: | 17:39 |
eyalroz | 1. All relevant libraries should be installed as dependencies for Cinnamon, or alternatively, it should not default to trying to use something it doesn't depend on; and | 17:39 |
eyalroz | 2. Even if Cinnamon needs a library and doesn't find it, it shouldn't "crash", there should just be a message saying "Cinnamon started in fallback mode since the foo library was not found at /path/to/libfoo.so" or something along those lines. And if Cinnamon doesn't do it, some other mechanism should. | 17:39 |
B3ta1 | Has anyone gotten Office 2013/wine working? | 17:43 |
furrywolf | I'd answer that question by going and looking at winehq's compatability reports... | 17:51 |
eyalroz | B3ta1: I remember the answer to be "Yes but not perfectly" | 17:53 |
B3ta1 | Acutally, on Linux Mint 2013 was without issue for me | 17:56 |
B3ta1 | thanks for replying | 17:56 |
B3ta1 | I'm not comparing them, but they are both debian derived | 17:57 |
furrywolf | I don't get the point of why you've said anything you've said. | 17:59 |
furrywolf | bbl | 18:01 |
eyalroz | Continuing my previous comment regarding mesa and nVIDIA drivers conflicting: | 18:44 |
eyalroz | If I ask to install agi glx-alternative-mesa, one of the dependencies is nvidia-installer-cleanup, | 18:45 |
eyalroz | and that, I suspect, will remove my nvidia drivers | 18:45 |
eyalroz | ... which I need for CUDA work | 18:45 |
eyalroz | gnarface: ... so I'm hesitant about trying it. | 18:46 |
gnarface | it would be likely that they can't be installed concurrently, yea.... | 18:47 |
muep | eyalroz: for opengl drivers, there is a quite special thing with there being multiple things that can be libGL.so and possibly some other drivers as well | 19:06 |
muep | er, other libraries | 19:07 |
eyalroz | muep: Well, I don't need an nvidia-driver-linked library being libGL.so, but - I do need the nvidia drivers installed and working | 19:07 |
muep | so with FOSS drivers you will usually want to use libGL.so from mesa and with nvidia non-free driver the libGL.so from nvidia drivers | 19:07 |
eyalroz | but it seems if I try to install glx-alternative-mesa, this will sweep all of nvidia's stuff away | 19:08 |
eyalroz | (and I'm not even sure it will resolve my problem...) | 19:08 |
muep | the libGL.so from nvidia drivers is essential if you want to use the nvidia driver otherwise | 19:08 |
muep | I would guess that it has all the shared compilers and other interesting parts of the driver | 19:09 |
muep | likely the kernel module of the nvidia driver is a pretty thin shim that just provides the driver code (outside the kernel) channels for communicating with the GPU hardware | 19:10 |
muep | hm, meant shader compilers in the message before my last one | 19:11 |
eyalroz | muep: I don't think the ligGL.so from nvidia is essential, because I don't want to use OpenGL with the nvidia card, I want to use CUDA. | 19:40 |
eyalroz | muep: I never link with it. | 19:40 |
muep | eyalroz: okay. I do not really know how CUDA works but my guess would have been that it can require also the right libGL.so to be present | 20:01 |
eyalroz | muep: CUDA does not involve OpenGL at all, AFAICT. I never include nor link against anything with GL or gl in its name. | 20:05 |
muep | that does not mean that it does not need the infrastructure that also backs their opengl drivers | 20:08 |
muep | but I do not know. I just wonder if you positively know that all the things you run with CUDA are things that do not need nvidia's libGL.so | 20:10 |
aitor | hi | 21:53 |
aitor | FreeCAD is cool, it's a great project | 21:55 |
aitor | developed in Qt/C++ and python | 21:57 |
aitor | nobody here? | 22:01 |
golinux | Maybe nobody that cares about CAD | 22:03 |
aitor | :) | 22:03 |
* golinux is always here | 22:03 | |
aitor | golinux is like god, he/she is always here | 22:06 |
golinux | aitor: There is no god. Certainly not me. ;) | 22:50 |
aitor | oh my god! | 22:53 |
MinceR | yes? | 22:56 |
aitor | we can be heroes, just for one day | 23:20 |
aitor | https://www.youtube.com/watch?v=bsYp9q3QNaQ | 23:20 |
debdog | hehe, having gods certainly worked out better than just having one | 23:21 |
golinux | It's all just human imagination. | 23:22 |
aitor | great david bowie :) | 23:24 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!