[00:01] and parted packages are on the Ubuntu 5.04 Server CD image (but not installed in the live session used by the installer) [00:05] JanC: so how do you get them on the live session? you keep trying to undo your claim that i was stating false information, whereas the issue is very simple: no, there are no usable parted binaries in the CD. i dont care if inside the dpkg seeds there is one package for it. [00:05] theres probably apache too somewhere in the seed repo for the initial install, so what? [00:06] the toolchain in 15.04 is broken to no avail for building fPIC libs, so cannot build a static parted here right now [00:06] may need to setup a gentoo vm for that [00:07] well, there is partman on the live image, why doesn't that work? [00:08] ugh [00:08] partman is terrible. ill seeif i can get around its limitations before i have access to a host that can build proper static execs of parted [00:08] i also need to script this installation so it can be done remotely [00:11] I'm sure you can install the parted udeb if you really want it [00:12] https://bugs.launchpad.net/ubuntu/+source/gcc-4.4/+bug/640734 [00:12] Launchpad bug 640734 in gcc-4.4 (Ubuntu) "crtbeginT.o needs to be recompiled with -fPIC" [Undecided,Confirmed] [00:12] JanC: can you tell me how that works? [00:12] btw that is a blocking bug for anyone compiling static execs [00:30] xedniv: udpkg -i /cdrom/pool/main/p/parted/parted-udeb [00:32] I tried using partman to set up an encrypted volume on lvm and that seemed to work fine too BTW [00:35] JanC: describe encrypted volume [00:35] did it contain a single partition or multiple? [00:36] ex. /boot, LVM(/, /etc, /var, /home, ...) [00:37] well, this was just a quick VM experiment, so I only tried unencrypted /boot + LVM+encrypted / [00:38] but, as said, you can install udebs with udpkg :) [00:45] yeah, LVM with SINGLE partition is easy with partman [00:45] otherwise it has never worked well for me === markthomas is now known as markthomas|away [02:08] is it recommended to use an encrypted file container such as LUKS to store ssh keys on an ubuntu server and have that container auto-mount on boot. or is there some other standard way of doing keeping ssh keys secure [03:36] If I see somewhat old versions and no open bugs in https://launchpad.net/ubuntu/+source/percona-xtrabackup does this mean updates are not coming soon, or maybe i looked in the wrong place? [03:40] i guess the upstream debian repo is not updating. [06:09] Good morning. === CiPi is now known as cipi === ashleyd is now known as ashd [07:52] if UBUNTU server with working static network configuration is suddenly disconnected and not able to reach the Gateway. How can i debug? Can any one please help on this. [07:52] Let's continue here ;) [07:56] lordievader: sure [07:56] lordievader: i have posted in ubuntu-server and did not get any replay. Routing table is fine in't not changed and compared with my backup configurations. [07:57] lodievader: any other idea? [08:01] Neo9: Could you pastebin the output of 'ip a' and 'ip r'? [08:01] lordievader: sure [08:05] lordievader: http://pastebin.com/s2aitiPS [08:07] Neo9: Would you mind giving the actual output? [08:07] lordievader: with actual ip? [08:08] Yes, [08:08] lordievader: okay. [08:10] lordievader: http://pastebin.com/24Z7erfu [08:11] Your routes are very much incorrect. (And that wasn't the output of 'ip a' and 'ip r', but never mind that) [08:12] lordievader: this is worked for earlier. [08:13] Oh wait, it is the retarted 'route' output that is throwing me off. [08:13] Can you pastebin the output of 'ip r'? [08:14] lordievader: ip r ..? would you reveal the command. [08:14] <_ruben> that is the command [08:14] Neo9: The full command would be 'ip route', 'ip r' is shorthand. [08:14] lordievader: okay. [08:15] I have a website hosted on a ubuntu server which does not have nameservers and mail server [08:15] so I want to use mx records of another host or another server [08:16] lordievader: http://pastebin.com/YMvQsiLp [08:17] Neo9: Can you ping 198.20.108.114? [08:17] lordievader: it is unreachable [08:18] Neo9: Could you give me the output of 'arp'? [08:19] lordievader: http://pastebin.com/44nXT2Eg [08:20] Neo9: Do you happen to have tcpdump installed? [08:20] On either 198.20.108.115 or 198.20.108.114 [08:22] lordievader: let me check [08:23] lordievader: and point1: i am able to ping the 198.20.90.115 from outside. but from 115 to 114 or any other is not working. [08:23] Neo9: Uhuh, I get the feeling arp is broken. [08:23] Hence the tcpdump to verify. [08:24] lordievader: it is installed in 114 [08:25] Neo9: Good, run 'sudo tcpdump -i any arp' then try to ping .114 from .115. Do you see any arp request incomming from the .115? [08:29] lordievader: http://pastebin.com/A7SYeXDa [08:30] No responses? There is your problem. [08:30] .114 cannot send a ping to .115 because it doesn't know it's mac address. [08:31] Now I am interested if .115 never receives the arp request or that .115 refuses to answer. [08:32] lordievader: understood, but how can i resolve it? [08:33] Neo9: First try and answer the question I just asked ;) [08:37] lordievader: not understood what exactly you asking? it saying that "destination is unreachable" === cipi is now known as CiPi [08:38] Neo9: Does .115 receive the arp request? [08:40] lordievader: how can i know that? i am able to ping .115 from .114 but not vice versa. [08:40] lordievader: can i know how to check that? [08:40] Neo9: Run 'sudo tcpdump -i any arp' on .115 then ping from .114. [08:42] lordievader: oops; tcpdump is not installed in the .115. And since even ping 8.8.8.8 not works, it is not possible to install it now. [08:44] Hmm, I am mixing things up. Anyways I get the idea that your broadcast channel is wrong. [08:44] lordievader: might be. then how it's worked for me earlier? [08:46] I have no idea, what has changed since then? [08:47] lordievader: is there any issue with mac address? And how can i read MAC address of other machine with arp? [08:48] What arp does is shout in the broadcast channel who has ip X, if a machine holds ip X it responds through unicast that it has ip X. however that reply is never received by your problem machine. [08:49] So either the problem is, no other machines receive the arp requests, or they refuse to send the reply. [08:51] lordievader: okay. Thanks for your time. thanks alot. Bye. [08:53] frediz: any news on kimchi please? [12:09] How can I assume a pty as though it were my terminal? [12:09] So being able to read and write to it in realtime === Tm_Tr is now known as Guest13976 [15:26] I am doing my first major server transfer(from one host to another)I would love to be able to do some type of image deal and transfer the whole thing at once so I can shut down one server and update the other and take off. Is this possible with Ubuntu and not having physical access? [15:28] skittishtrigger: depends on whether transferring bare-metal or VM, and if VM, what hypervisor source and destination use. Also, after transfer, things like network might need reconfiguring (different MAC address, different device naming, etc) [15:29] It would be baremetal. I have not progressed to doing vm's yet [15:29] Its a fairly simple server with a webserver and several game servers, quassel, etc. [15:31] I was thinking with the custom config on a lot of things if it might be a better route to just make a list usable by ubuntu with everything that is installed on it and use that then re-set up everything? [15:32] skittishtrigger: so main difference will be hardware, and most drivers should auto-detect and load. Anything tied to NIC MACs will fail though... so think of /etc/udev/rules.d/70-persistent-net.rules [15:33] skittishtrigger: you could. debfoster is a useful too for creating the dependency list [15:34] ok, I will google that and look into it. luckily the only think network wise that is coded in is just the ips for client connections [15:34] Thank you so much for your help TJ- ! [15:41] I've installed 15.04 server on an old laptop...playing around with some stuff...Can someone tell me how to make it auto login a non-root account? Everything I find on google seems to be outdated. I honestly know extremely little about linux...I'm playing around with this to maybe turn into a digital picture frame [15:45] Hello, so I created a backup of /etc, /var, /usr, /tmp etc upon an upgrade and reinstall process. I was just wondering if there was any way to grab the postgresql database from one of those folder locations and insert it back into the newly installed OS [15:46] Or if there was a way to go about backing up the database from one of those folder locations and importing and upgrading it from 9.1 to 9.4 of postgresql === markthomas|away is now known as markthomas [16:47] I've installed 15.04 server on an old laptop...playing around with some stuff...Can someone tell me how to make it auto login a non-root account? Everything I find on google seems to be outdated. I honestly know extremely little about linux...I'm playing around with this to maybe turn into a digital picture frame [16:53] Synopsis: As I said earlier, edit the init system's ttyX configuration so it does "exec /sbin/getty -8 -a <$USER> 38400 tty6" [16:53] Synopsis: precise syntax depends on which init system it is [16:53] Synopsis: my example above is from a test I did on tty6 with upstart's /etc/init/tty6.conf [16:57] and as I told you earlier, those files don't exist for me [16:58] Synopsis: So find out where systemd stores it's equivalent, which are called unit files [17:00] Synopsis: "man systemd-units" should give all the details required to find them [17:01] ty, will look [17:01] Synopsis: "dpkg -L systemd" will list all the files the systemd package installs [17:02] Synopsis: "dpkg -L systemd | grep getty" should narrow down your search [17:04] the grep line returned a few files [17:07] http://pastebin.com/csM3rN4H [17:08] and btw, when i try to man systemd-units, i get no man entry [17:09] Synopsis: ahhh, no S. "man systemd.unit" [17:09] same thing, no man entry [17:10] i get one for just systemd [17:10] Synopsis: I've not the systemd source-code here. It looks as if units/getty@service is what you want, it has a line "ExecStart=-/sbin/agetty --noclear %I 38400 linux" which is, I suspect, the one that needs editing to include "-a " [17:10] Synopsis: that is 15.04, right? [17:11] yes, 15.04 [17:11] Synopsis: according to packages.ubuntu.com that file is installed (as /usr/share/man/man5/systemd.unit.5.gz) [17:12] a zip file? i'll have to extract it, modify it, rezip it and replace? [17:12] Synopsis: No... all man-pages are compressed to save space. They are uncompressed on the fly by the 'man' tool [17:13] Synopsis: there's an ArchLinux guide on how to configure systemd for auto-login. https://wiki.archlinux.org/index.php/Automatic_login_to_virtual_console [17:14] Synopsis: that shows creating an over-ride to the package-installed unit file, which is the preferred way to do this kind of config change (we never want to change a package-installed file - that may prevent the package being upgraded later) [17:15] k, knocking some stuff out @ work real quick and i'll give that a try [17:21] Synopsis: For the man-page you've been typing a hyphen, not a full-stop. [17:23] i copied and pasted what you typed the first time, and then just backspaced the s off when you corrected yourself, didnt notice you also changed the - to a . [17:24] Synopsis: Sorry... that first command I took directly from a web-page of the man-page, and the distro there seems to replace the full-stop with a hyphen [17:27] no need to apologize :) you're still being very helpful [17:28] Synopsis: it's an intriguing option. Quite easy to implement once you find the correct file ;) [17:29] easy if you know what you're doing maybe...i ran the "systemctl edit getty@tty1" like on that link and copied the text into the file, rebooted and the system is hung at a blank screen with a blinking cursor [17:29] i can still ssh to it though [17:29] Synopsis: Best to test on something other than the primary tty, that is generally reserved for system [17:30] ok, so how do i undo what I did and which tty should i use? any # but 1? [17:30] Move that file to be @tty2, reboot, you should have a login prompt on tty1, and pressing Alt+F2 should get you onto tty2 [17:31] You'll probably need to edit its contents too, if they mention 'tty1' of course [17:34] k, moved that file, rebooted, have my login prompt again, but it didn't auto login, and if i alt-f2, it's stuck at the blinking cursor...i've got to just be missing something stupid [17:34] Synopsis: I see a possible confusion in those ArchLinux notes. They specify the file to manually edit inside the coloured code block that also shows the contents to use. *don't* include that filename in the file [17:35] Synopsis: in other words, the content you add starts with the "[Service]" line [17:35] Synopsis: just to be sure that systemd is creating the expected ttys... does Alt+F3 get you to a login on tty3? [17:36] yes, and i didnt copy the filename into the contents, just the [service] and below [17:37] Synopsis: I'm not sure than; the fact you don't get a login prompt on tty2 shows the unit file is being acted on... presumably something fails, so no process takes on tty2 [17:38] would it be logged somewhere so I could get an idea? [17:42] I think so, systemd is supposed to be quite verbose. I think "journalctl --unit=getty@tty2" *may* be what you need [17:43] nah...didnt return anything [17:43] Synopsis: I may have the syntax wrong; I was trying to decipher the man-page [17:44] i ran it without any parameters and found this [17:44] at the very end [17:44] try it for --unit=getty@tty1 .... if that shows something, at least we know its the correct syntax [17:44] nevermind, thats just the command i entered before that returned nothing [17:45] You know you can directly paste text from a console? if you use "pastebinit <( your-commands-nere )" the output will be sent to a pastebin and you'll be given the URL to pass on [17:45] didnt know that :) [17:45] or to send a plain file, "pastebinit /path/to/file" [17:46] We use that extensively for support, provided the PC has an internet connection [17:46] @tty1 did return 2 lines [17:46] Synopsis: OK ... try it for --unit=getty@tty3 as well . If that also returns something we know what we should be expecting for tty2 [17:47] nothing for 3 [17:47] Synopsis: And there's a login on tty3 (Alt+3) ? [17:47] yes [17:47] Synopsis: OK, that tells us tty1 is 'special' :) [17:48] Synopsis: ok, "systemctl --failed" [17:49] what am i doing wrong here? [17:49] picturesadmin@localhost:~$ pastebinit -bash: journalctl: No such file or directory [17:50] unrecogonized option --failed [17:50] oops [17:50] sorry, wrong base command [17:51] Synopsis: "pastebinit <( command )" ... notice the parenthesis () immediately after the < [17:51] ok, didnt realize those were literal [17:52] Synopsis: when I type commands inside double-quotes it shows exactly what is required, with the exception of the user-variable 'command' [17:52] http://paste.ubuntu.com/12142767/ [17:54] Synopsis: OK, that confirms the unit failed, so we know it tried. Now lets get detailed info on the failure: "systemctl status getty@tty2 | tee /dev/stdout | pastebinit" [17:54] Synopsis: that command lets you see the output as well as sending it to pastebin [17:54] http://paste.ubuntu.com/12142782/ [17:55] all it did was give me the pastebin url, didnt see output on screen [17:57] Synopsis: oh... silly me, typos again ... should have had in there " | tee /dev/stderr | " [18:00] k...that showed me the output :) [18:00] Synopsis: I think the cause is ArchLinux buts agetty in a different place to Ubuntu. I also notice that ExecStart= doesn't seem to be followed immediately by a hyphen as is shown in the ArchLinux guide - not sure if it is 100% vital but doubt it'd be there if it isn't needed. The correct path is ExecStart=-/sbin/agetty ... [18:01] s/buts/puts/ [18:03] so run systemctl edit getty@tty2 and fix that path? [18:05] Synopsis: Yes [18:06] Then you should be able to test it immediately with "systemctl start getty@tty2" - no reboot required [18:06] ok, that worked, when i switched to tty2 it logged the user in, but i think i want it on tty1 since i want that to start my "slideshow" [18:08] Synopsis: I think it would be better to keep tty1 reserved for the system since it is 'special' and just have a script/command the makes tty2 the active VT [18:08] i'm ok with that if you can point me in the direction to make it happen [18:10] * TJ- groans :D [18:10] I really walked into that! [18:10] hehe [18:13] The manual command is "chvt X" where X would be 2 (like pressing Alt+F2). I need to check if there's another/better way to set it from boot [18:13] k === Lcawte|Away is now known as Lcawte === Lcawte is now known as Lcawte|Away [18:20] according to http://unix.stackexchange.com/questions/115482/how-to-change-the-default-tty-after-boot, i can just add the command to /etc/rc.local [18:20] Synopsis: yes, that's the 'manual' way I was talking about, but I seem to recall there was a way to configure it on the kernel command line [18:21] Synopsis: but go with /etc/rc.local ... add "chvt 2" *before* the "exit 0" line. rc.local *must* return 0 (success) to its caller [18:22] perfect, works like a charm...couple more questions and all that will be left is for me to disassemble this laptop and build a "frame" for it :) [18:22] Don't break it after all this! [18:23] would i put my "slideshow" command in rc.local as well? [18:24] http://paste.ubuntu.com/12143037/ [18:24] If you're not worried about it failing and not being restarted. Otherwise, create a custom systemd unit that starts the application [18:24] thats what i was getting at...and can i make it restart periodically even when no problems so it can look for new files? [18:26] Yes... you could even use a separate background process using 'inotifywait' that 'watches' the media directly and restarts fbi/a systemd service when it sees changes in the directory [18:26] s/directly/directory/ [18:26] ooooh...pretty! tell me how! :P [18:27] I'm leaving that as an exercise for you! I've got things of my own to do :D [18:27] :) === ashleyd is now known as ashd [19:07] hey tj, still around? [19:30] Who maintains https://cloud-images.ubuntu.com/locator/ec2/releasesTable? There's a trailing comma (invalid in JSON) at the end of the list of lists. [19:35] Never mind, I found https://cloud-images.ubuntu.com/releases/streams/v1/com.ubuntu.cloud:released:aws.json [19:37] good find, i was searching https://help.ubuntu.com/community/UEC/Images [21:14] smoser ping, you around [21:38] When I reboot from a session it loads to a blank screen. CTL+ALT+DEL works to reboot. Sometimes it boots back to the blank screen, sometimes to the GRUB menu, and even once it went straight to the login. What's going on??? I would love to have it just boot to login. [21:39] I'm setting up Ubuntu Server 14.04.3 [21:43] you can disable the gui and make it boot straight to terminal by editing grub_cmdline_linux_default to say "text" in /etc/default/grub [21:44] if that's what you're attempting to do [21:55] @larsi, that is what I would like, there is no GUI installed yet. However, I think the problem is in the BIOS or GRUB. It the computer still reboots to a blank screen after changing the value to "text". [21:59] dweller_: 'text' tells the system not to auto-start the display manager. Has no effect on -server [21:59] dweller_: it sounds like a problem with the system... possibly hardware [22:26] Where are postgresql database files located? I just want to see if it is possible to restore adatabase (I didn't back it up to a file) that might have been backed up from one of the folders I archived before I installed ubuntu 15.04 [22:26] Or am I SOL if I didn't back the databse up to a file before wiping and installing 15.04? [22:29] Somewhere in /var/lib would be my guess [22:51] maxb well, I do see several files there, but I assume that copying a 9.1 database over to a 9.4 postgresql would be a bad idea in general === medz is now known as samsn [23:46] Any way to install postgresql 9.1 on 14.04 now that the only one listed in te repos is 9.4? [23:46] I sorta need to grab it before I can do anything.