[00:47] Hello! [00:47] I want to install a driver for my nvidia geforce 8400 cs video card [00:48] I am following this tutorial http://ubuntuguide.net/install-latest-nvidia-graphics-drivers-in-ubuntu-linux but it seems different [00:48] Perhaps it was donne for an older ubuntu server version [00:48] Could someone help me please? [00:48] I already downloaded NVIDIA-Linux-x86_64-340.102.run file from nvidia website [01:08] kuwer: that guide looks eight years old or so. it's probably fair to ignore it entirely. [01:09] I always thought you just install the normal nvidia package and be done with it [01:09] unless you want to do something odd, like use cuba in your applications for like bitcoin [01:09] then your going have hell and be mostly on your own, cause that crap is not fun [01:09] patdk-l2: well, yes. I need Cuba, OpenCL [01:09] I don't think anyone does bitcoin mining on gpus anymore [01:09] atleast I gave up on it, and it takes a lot to do that [01:10] sarnold, I know a guy that just setup 12 amd video cards for it this week [01:10] I am not mining bitcoins but other coinis [01:10] patdk-l2: yikes [01:10] yeah, Bytecoin, Ethereum, well there is hundreds [01:11] this wiki page suggests 'sudo ubuntu-drivers devices' is a good starting point https://help.ubuntu.com/community/BinaryDriverHowto/Nvidia [01:11] what's funny is it jumps from that right into installing nvidia-nnn -- maybe nvidia users are forced to install the new driver updates by hand? dunno.. [01:11] they have to [01:11] anyway based on this wiki the first thing I'd try is "apt-get install nvidia-375" [01:12] cause the ones in ubuntu don't have the cuba stuff in them [01:12] get a intel phi :) [01:12] The correct driver for my video card is NVIDIA-Linux-x86_64-340.102 [01:15] I could use wget http://us.download.nvidia.com/XFree86/Linux-x86_64/340.102/NVIDIA-Linux-x86_64-340.102.run [01:15] chmod u+x NVIDIA-Linux-x86_64-190.42-pkg2.run [01:15] ups [01:16] chmod u+x 340.102/NVIDIA-Linux-x86_64-340.102.run [01:16] chmod u+x NVIDIA-Linux-x86_64-340.102.run [01:16] sudo apt-get install linux-headers-$(uname -r) build-essential [01:16] Is that necessary? [01:17] probably; you have to compile portions of the drivers on your system [01:20] How about https://askubuntu.com/questions/66328/how-do-i-install-the-latest-nvidia-drivers-from-the-run-file [01:28] Used chmod +x driver [01:28] then sudo ./driver.run [01:29] And http://imgur.com/a/Vk5iu [01:29] kuwer: pastebin the logfile? [01:29] (the pastebinit tool from the pastebinit package can make that easy) [01:29] sarnold: this is the log file [01:30] is that the WHOLE log file? or just the last few lines? [01:30] last few lines [01:31] good good :) pastebin the whole thing and then we might stand a chance of spotting the error [01:31] how can I send out from the server this line? [01:32] apt-get install pastebinit ; pastebinit /var/log/nvidia-installers.log [01:33] paste.ubuntu.com/24812224 [01:34] Perhaps this driver was not meant to server [01:34] WARNING: You do not appear to have an NVIDIA GPU supported by the 340.102 NVIDIA Linux graphics driver installed in this system [01:35] are you sure that your card is supported by this driver/ I hear they are aggressive about dropping support for 'old' hardware [01:36] This is not the latest driver [01:36] I went to nvidia and check for my video card model driver [01:36] And this is the one it pointed me to [01:38] As http://imgur.com/a/vHo8c [01:39] Maybe I should Run 'make oldconfig && make prepare' on kernel src to fix it."; [01:40] I don't think that would help [01:43] kuwer: the patches on https://gist.github.com/tpruzina look promising [01:44] kuwer: from your pastebin, the actual errors are starting on lines 3588 and then in 3598 [01:44] all look like hotplug cpu related [01:44] kuwer: so try making the same changes that the patch that disables the hotplug support .. [01:45] Well, something I got to say is I am running this Ubuntu Server on a VirtualBox machine just to test. [01:46] dunno how that would work [01:46] Probably I will be doing the same procedure soon on a ssd driver I am waiting [01:51] I don´t think there is path for my driver version [01:52] did you pass the pci through? [01:52] what do you mean, sorry [01:53] in order to use devices inside vms you have to use pci passthrough [01:53] I've never tried it before [01:54] Oh, not really. [01:54] Let me check that [01:58] Not sure where to set PCIE [01:58] Checking [02:10] Don´t think my BIOS has BIOS that has the IOMMU === zzz is now known as zero === zero is now known as zzz === med_ is now known as Guest6511 === JanC_ is now known as JanC [07:42] Good morning [09:12] coreycb: checklist via PPA for pike b2 - https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/2808/+packages === ogra_ is now known as ogra === ogra_ is now known as ogra === ogra is now known as ogra_ [13:03] jamespage: looks like you need zunclient now [13:03] zul: and some [13:03] zul: https://launchpad.net/ubuntu/artful/+queue?queue_state=0&queue_text=python [13:03] ahaha [15:19] Hey nacc, do you have any example of a preseed copy? A file copy from USB/CDROM to the newly installed system if possible. Google isn't being useful to me. [15:24] hashwagon: https://help.ubuntu.com/lts/installation-guide/example-preseed.txt down at the advanced options [15:25] hashwagon: that's how you run any old shell command [15:26] mwhudson: jamespage: is it expected that there is no python-celery-sphinx? I think the new celery is loading a module from it, but there's only python3-celery-sphinx, afaict [15:26] s/celery-sphinx/sphinx-celery/ [15:26] hashwagon: a more complex example: https://github.com/gc3-uzh-ch/openstack-tools/blob/master/etc/ubuntu-preseed.cfg#L489 [15:27] tomreyn: nice, thanks! [15:27] and here you got a copy https://lists.ubuntu.com/archives/ubuntu-installer/2007-December/000130.html [15:28] oops this one doesn't work, sorry [15:31] this one uses 'cp' an is said to work https://help.ubuntu.com/community/InstallCDCustomization/PreseedExamples [15:38] nacc: hmm that's odd [15:38] I would expect both versions [15:38] * jamespage looks [15:45] tomreyn, nacc, thanks guys. [15:51] nacc: no idea [16:08] jamespage: yeah, it's surprising to me -- i also only see python3 in debian (afaict) [16:08] mwhudson: np, thanks [16:09] nacc: apparently that's right but I have no idea why there is no py2 [16:09] maybe the maintainer was being visionary! [16:10] jamespage: yeah, and i think celery won't build without the py2 -- or I have to dig a bit into the internals [16:27] jamespage: i'm looking to see if it's easy to add -- it seems like it's unexpected to only be py2 (and the debian side is just a trival stdeb generated thing (it seems). I wonder if it's just oversight [16:57] jamespage: http://paste.ubuntu.com/24816633/ it builds. I'm verifying the autopkgtest now [17:03] jamespage: tests pass, and it is exactly what fails with celery (import'ing sphinx_celery) :) [17:52] Is it normal for a 16.04 server preceed install to have GRUB_CMDLINE_LINUX_DEFAULT="quiet splash" in /etc/default/grub? This prevents me from booting to a tty unless I remove quiet and splash. [17:56] sed command to resolve that: http://downloads.linux.hpe.com/SDR/project/minnow/current/Ubuntu_16.04_RAID1_preseed.cfg - odd behavior though.. [18:19] I created a new apache2 vhost at a specific subdomain, and now my default vhost is being clobbered. What gives? [18:21] creator typo [18:21] The vhost in question: https://paste.pound-python.org/show/aZwN2OyZqjb3ee74YiZQ/ [18:22] Note there are no wildcards in that vhost, and yet it's answering to ALL requests, not just ones directed at mantis.lavacano.net [18:22] I've currently disabled that site configuration. [18:23] Google says something about a NameVirtualHost directive but it doesn't appear to actually do anything [18:25] "If multiple virtual hosts contain the best matching IP address and port, the server selects from these virtual hosts the best match based on the requested hostname. If no matching name-based virtual host is found, then the first listed virtual host that matched the IP address will be used. As a consequence, the first listed virtual host for a given IP address and port combination is the default virtual host for that IP and port combi [18:25] https://httpd.apache.org/docs/2.4/mod/core.html#virtualhost [18:25] that's my guess.. [18:26] sarnold: Assuming 000-default.conf is parsed before 100-mantis.conf, then *:80 should come first and take priority, no? [18:26] https://httpd.apache.org/docs/2.4/vhosts/ suggests using apachectl -S to dump the parsed configuration as a useful troubleshooting tip [18:26] jerichowasahoax: I'd hope so :) [18:29] jerichowasahoax: instead of you'd usually use [18:29] using hostnames there is "not recommended" [18:29] https://httpd.apache.org/docs/2.4/mod/core.html#virtualhost [18:30] tomreyn: But that IP address points to a couple different FQDNs, as I'm sure you noticed while you were checking my DNS records [18:30] jerichowasahoax: if you mean the alias, there's only one [18:30] mantis.lavacano.net is an alias for andariel.lavacano.net. [18:30] tomreyn: Not the CNAME, the IP itself [18:31] tomaw: and andariel.lavacano.net when contacted directly should be serving different content [18:31] er [18:31] oh your mean PTR records? [18:31] tomreyn: see above [18:31] that's what ServerName is for [18:32] see the example in the docs i pointed to [18:32] ServerName isn't working either [18:32] VirtualHost uses an ip address, ServerName an FQDN [18:32] or now it is [18:33] * jerichowasahoax makes a mental note: any virtualhost that isn't is practically invalid [18:34] if thisis a general statement then i'd disagree [18:34] i know the technical explanation is "explicitly defined addresses take priority over wildcards" [18:34] which is why "mantis.lavacano.net" was clobbering things [18:34] but if I don't overgeneralize I'll just keep screwing that up over and over and over [18:35] VirtualHost defines the ip address + port to service this ServerName FQDN on. [18:35] tomreyn: apachectl -S was putting "149.56.132.23:80" over "*:80" [18:35] ahh because it's more specific.. [18:35] sarnold: yes [18:35] sigh stupid complicated mess [18:36] to be fair, the google results and docs weren't entirely clear on exactly what i was trying to put where [18:36] all they really said was "VirtualHost some.domain.com" and "VirtualHost some.domain2.com" [18:37] it didn't occur to me that they were assuming those names had different addresses [19:39] Do you folks have any recommendations to *automatically* do some kernel cleanup, so /boot doesn't keeep filling up? [19:39] it'd be nice to only keep 2-3 kernels... [19:39] how about: sudo apt autoremove [19:41] SuperLag: do you have /etc/kernel/postinst.d/apt-auto-removal on your system? it should auto-generate a list of packages to keep in /etc/apt/apt.conf.d/01autoremove-kernels that works with /etc/apt/apt.conf.d/01autoremove [19:42] sarnold: I wasn't aware of that. I'll investigate. [19:42] thank you, btw! === unreal_ is now known as unreal === oerheks_ is now known as oerheks