[02:13] https://help.ubuntu.com/lts/serverguide/firewall.html#ip-masquerading those ufw nat instructions don't actually work [02:16] https://pastebin.com/sUrTm9ZK [02:25] deleting and readding the commit line several times fixed it? nice parser. [04:56] i have messup with my ubuntu 18.04 server with grub loader there is now only memory test entry are there there is no linux-version entry [04:56] how do i fix this problem ? [06:32] good morning [06:32] samba35: that seems that no kernel at all was found on your last update grub [06:33] samba35: there might be more options, but I'd try to boot from USB/CD [06:33] samba35: then chroot into your actual system [06:33] and then check why there is no kernel, fix that up [06:33] and eventually run sudo update-grub [06:33] that should then detect the kernel that you made available and update the boot tables as needed [06:34] then cross fingers, reboot and done [06:34] samba35: I think the live CD also has a boot-repair option that fixes most common issues [06:40] sorry cpaelzer [06:41] for what samba35? [06:41] i try to boot from dvd and mount it but it give erros [06:41] well then, disk broken maybe [06:41] what kind of errors [06:42] error with dpkg and unable to find some files [06:42] provide a pastebin of the commands you tried and maybe the logs and dmesg that was created along [06:42] hmm [06:42] are those dpdk errors after mount+chroot? [06:42] did you bind mount /proc, ss and dev ? [06:42] is it possible to boot from another harddisk and make broken disk as a slave and try to repiar ? [06:42] no [06:42] samba35: that is exactly what you already do with the CD [06:43] CD is the "other disk" [06:43] and you try to repair from there right [06:43] currenty i am on 16.04 and broken disk as a slave [06:43] will i able to fix this now ? [06:44] no one can tell you that in advance, but give it a try [06:44] before you chroot you should like [06:44] 1. mount the disk on /mnt [06:45] 2. for f in proc sys dev ; do mount --bind /$f /mnt/$f ; done [06:45] 3. chroot /mnt [06:45] then you are as if you'd be in your hard disk's system [06:45] not sure if you need network tricks, but that would be a start [06:45] there check /boot and if there is a kernel as it should be [06:45] then run sudo update-grub what does it say ... [06:45] ok [06:45] let me try [06:54] sda1 is mount as a /media/tesuser/some numbers [06:55] what is ss ? [06:58] mount /media/tesuser/637d0aee-9232-11e8-9c61-2ed34b5b4932/proc/ /current (from where i have booted ) that drive any mount point or old salave disk mount point ? [06:59] mount: /media/tesuser/637d0aee-9232-11e8-9c61-2ed34b5b4932/proc is not a block device [06:59] ok [07:00] done [07:00] with --bind [07:04] yeah it just means "be also available there for when I late chroot" [07:05] from where i have to run update-grub ? [07:06] salave /18.04 disk or currnt 16.04 disk [07:09] ok let me boot from cdrom and try [07:10] cpaelzer, thanks i will be back later after trying this [07:11] samba35: you want to do that "inside" the chroot which is on the disk you want to fix [07:11] ok [07:17] chroot: failed to run command '/bin/bash': No such file or directory [07:23] in old disk in /boot i can see initrd.img-4.15.0-42-generic.dpkg-bak [07:23] initrd.img-4.15.0-42-lowlatency.dpkg-bak [07:23] initrd.img-4.19.2-041902-generic.dpkg-bak [07:24] and when i run update-grub here i show currnt disk kernl 3.x [07:28] w /usr/sbin/grub-mkconfig: 37: /usr/share/grub/grub-mkconfig_lib: cannot create /dev/null: Permission denied [07:29] mount but why ? [07:29] a /usr/sbin/grub-probe: error: cannot find a device for / (is /dev mounted?). [07:36] cpaelzer, a /usr/sbin/grub-probe: error: failed to get canonical path of `/dev/mapper/ubuntu-root'. [07:36] now this error [07:38] hmm seems like lvm root [07:38] righ tnow I'm not sure without checking myself what special consideration apply [07:38] ok [07:38] your list of kernel/initrd mentions only backup files [07:39] no actual kernel/initrd there ?! [07:39] yes [07:39] well, whatever the reason is that is at least part of your issue [07:39] there was error with dpkg and i google some site and mess up [07:39] you'd want to install at least one kernel [07:39] yes [07:40] will booting from 18.04 dvd will help ? [07:40] yeah the case look half-broken already, this gets hard to fix-by-remote-hints [07:40] and rescue mode [07:40] booting a different DVD will not make it any different IMHO [07:40] ic [07:40] you still need to install a proper kernel on the main disk in the chroot [07:40] and then update-grub [07:41] and sort out whatever the lvm setup makes you need in additon [07:41] ok [07:46] thanks cpaelzer back soon thanks [09:02] Good morning [09:46] Does anyone know how to generate a control file from a control.in file? [09:48] kstenerud: there's no one way, make -f rules control ? [10:01] E: Sub-process /usr/bin/dpkg returned an error code (1) [10:01] how to fix this [10:29] Would this be the right place for some rsync things? I tried using --filter=":e- .gitignore" but unfortunately it still downloads all files instead of ignoring some. Why could that be? [10:37] damn it. forget about this. the ":" is a directory merge so the .gitignore of the sender will be used (and not mine as receiver). [11:09] good morning [11:27] I don't know when or how, but at some point, vsftpd and LDAP were working fine, and now it fails. with lftp client, it started doing this: ls: Login failed: 500 OOPS: cannot locate user entry:username [11:28] and if I change it to use the sshd pam service instead, it then says ls: Login failed: 530 Login incorrect. [11:35] you need to checkout which ldap searches are being done, and what their result is [11:35] peetaur2: ^ [11:46] I basically did that. [11:46] but restarting nslcd fixed it .... :/ [11:46] why should getent, ssh, etc. all work fine using nslcd, but vsftpd needed it restarted? silly thing [11:47] ah, the cache? [11:48] nslcd does caching, yes [16:53] rbasak: hey, I just saw something interesting in a debian package [16:54] rbasak: debian builds samba with glusterfs support, so their samba-vfs-modules package ships /usr/lib/x86_64-linux-gnu/samba/vfs/glusterfs.so [16:54] and that links with libglusterfs.so.0, which comes from glusterfs-common [16:54] rbasak: but their samba-vfs-modules package has no dependency on glusterfs-common. Instead, they have a Recommends [16:54] root@oriented-mastodon:~# dpkg -s samba-vfs-modules|grep gluster [16:54] Recommends: glusterfs-common, libcephfs2 (>= 12.2.8), libdbus-1-3 (>= 1.9.14) [16:55] is that a valid trick? [16:55] shlibs didn't seem to have picked up the linkage between /usr/lib/x86_64-linux-gnu/samba/vfs/glusterfs.so (samba-vfs-modules) and /usr/lib/x86_64-linux-gnu/libglusterfs.so.0 (glusterfs-common) [16:58] in the package's description: [16:58] Note: The runtime dependencies of vfs_ceph, vfs_glusterfs and vfs_snapper are [16:58] moved to Recommends. [17:01] ahasenack: so if you install samba-vfs-modules (but not it's recommends) and then configure samba to use gluster, does it segfault/crash? [17:01] nacc: it would definitely fail to load that module. I think it still starts up, but I would have to check [17:01] Plugin modules are weird [17:02] It might be reasonable to leave it as a recommends or even a suggests [17:02] also, what does ubuntu do, since it installs recommends by default? What does it do when it cannot find a recommends? [17:03] I think it will not take any action on a recommends it cannot fulfil. But that might need checking. [17:03] "The Depends field should be used if the depended-on package is required for the depending package to provide a significant amount of functionality." [17:03] From policy [17:04] For something like samba-vfs-modules, glusterfs may not be treated as "significant". [17:04] right [17:04] but the question is, what would britney do? :) [17:04] It doesn't know so it doesn't care. [17:05] I mean, at migration time [17:05] if it would treat recommends as depends or not [17:05] or "whatever apt would do" is the answer [17:05] Oh [17:05] I see what you're asking now. I don't know. Good question. [17:06] I'll find out [17:06] we could drop that bit of delta if this works and is acceptable [21:13] coreycb: jamespage: is that nova test error anything you triggered or could fix https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-disco/disco/amd64/n/nova/20181206_155403_69eef@/log.gz ? [21:13] the libvirt upload I did certainly didn't trigger that [21:14] was a s390x only one line change to vfio handling [21:14] so I wonder if there is anything going on that you know avoiding that I have to debug and find out on my own tomorrow [21:14] * cpaelzer is lazy before going to bed and hope the world fixes my issues [21:15] I was thinking the other day. It would be awesome if apt update would notify if you needed to reboot or not like it does on the desktop. [21:16] evit: motd is updated with the reboot needed notice [21:17] cpaelzer: I've been trying to figure it out all day. I made a little progress at https://bugs.launchpad.net/ubuntu/+source/nova/+bug/1807262 [21:17] Launchpad bug 1807262 in nova (Ubuntu) "stein unit tests fail with sqlalchemy.exc.NoSuchTableError: migration_tmp" [High,Triaged] [21:19] sdeziel, I'm talking about after you do a manual apt update && apt upgrade [21:20] @sdeziel, I disabled MOTD [21:20] cpaelzer: something changed w sqlite [21:20] evit: I guess you could have a post-apt hook to do what the motd in question was doing [21:21] coreycb: if you don't find something towards your EOD [21:22] would you mind filing a force-badtest for now to unblcok me and potentially others [21:22] that would give you tim eto sort it out [21:22] evit: if you are looking for something to let you know which services would need a restart (more frequent than reboot-needed), you may want to look at https://github.com/simondeziel/check-deleted-libs/blob/master/check-deleted-libs [21:22] thanks for the bug reference, I'll refresh tomorrow what your last state on this was [21:26] cpaelzer: haven't done that before but will look into it [21:27] thanks coreycb I'd appreciate that [21:28] ok, cu tomorrow where I'll recheck the latest updates to that bug [21:51] Silly question. I've updated my .bashrc to include my FQDN with /H but I don't see it upon logging back in. Any ideas? [21:52] I mean \ [21:52] \H sorry [21:52] evit, to include it in $PS1, you mean? [21:53] lordcirth, Yes [21:53] evit, what line exactly did you put in .bashrc? [21:54] PS1='${debian_chroot:+($debian_chroot)}\[\033[01;32m\]\u@\H\[\033[00m\]:\[\033[01;34m\]\w\[\033[00m\]\$ ' [21:54] else [21:54] PS1='${debian_chroot:+($debian_chroot)}\u@\H:\w\$ ' [21:54] fi [21:54] unset color_prompt force_color_prompt [21:55] hostname produces www. Is there somewhere else I need to change that? [22:00] debian_chroot? [22:00] ah, nvrm [22:00] evit, you might have run into this bug: https://bugs.launchpad.net/ubuntu/+source/bash/+bug/1276796 [22:00] Launchpad bug 1276796 in bash (Ubuntu) "Long hostname placeholder for PS1 (\H) fails in Bash" [Undecided,Confirmed] [22:01] A workaround is to use $(hostname -f) instead of \H [22:06] lordcirth, Gracias! =) [22:07] np