[00:22] Hmm, that's a purely wishlist item compared to my other problem at the moment, apparently one of the storage servers has a dead drive, but the 4-drive BTRFS raid10 pool won't even mount with `-o degraded` [00:24] And meanwhile another server is giving "Reading package lists... Error!" on an apt update, yikes, everything's blowing up at work this Friday night . . . [01:55] Hrmm. The label of the BTRFS pool is "heartland", and that's how it was being mounted via its fstab entry. But, running `sudo mount -o ro,degraded LABEL=heartland /media/heartland` is definitely not working, but it's not spitting out any errors either, it's just failing to actually do anything. [02:03] A reboot later and surprisingly the same mount command worked fine this time, huh. [02:15] Now my problem is that the `btrfs replace` command doesn't appear to be working. `sudo btrfs replace start 4 /dev/sde /media/heartland` should be what I need as I understand it, but running `btrfs replace status` just returns "Never started" [02:25] I can't just add a device because I can only mount it read-only; attempting to mount it using only `-o degraded` gives "wrong fs type, bad option, bad superblock on /dev/sdd" etc [05:42] Okay, looks like that's a known issue and I need a new enough kernel to be able to mount a degraded 4-disk raid10 array. Unfortunately that "new enough" kernel is "literally the latest", heh. Time to call it a week and come back and make a custom live disk with kernel 4.14.2 on Monday. [10:36] Good morninig [10:36] morning even [14:58] if you had to take an educated guess: of all servers connected to the internet how much would you think run a webserver? [18:08] <__mapo__> hi, i've tried to install ubuntu server 16.04.03 but I had a problem in the tasksel step, so I had to skip that step,install grub and reboot the system. Now to fix that missing step all I have to do is apt-get install ubuntu-server? [18:09] __mapo__: you don't need to ask in multiple channels :) [18:11] sudo tasksel install ubuntu-server maybe. google that [18:11] <__mapo__> :) I wrote here because alkig pointed me to for server-specific issues [18:13] did he? I must have missed that! [18:14] It's generally ultra quiet in here weekends [18:17] <__mapo__> compdoc: sudo tasksel install ubuntu-server did nothing, sudo tasksel install server ended up with an "apt-get failed (100)" error, which is basically my problem at hand [18:17] <__mapo__> should we continue here or go back to ubuntu? === hhee is now known as rh10 [18:21] you have run 'sudo apt-get update' ? [18:23] <__mapo__> yes [18:25] mgith be easier to just start from scratch [18:25] Probably. [18:26] You can modify /etc/apt/sources.list to include only CD sources and update from that but it's a pain. [18:26] When i install, I only consider the install a success if there are no errors. I do it over until I figure out what Im doing wrong [18:26] Like, I had to boot the CD in recovery mode to even get mounting to copy the files and do all that. [18:27] And yeah, if I hadn't been using the server for other things I would've just done a reinstall [18:28] <__mapo__> I just came from a fresh install and it all started with a dependency error in the tasksel step during installation, which prevented me to install the task I wanted, including the "basic ubuntu server" [18:30] <__mapo__> compdoc: with your criteria my install was a fail, but whatever I tried I just couldn't get this step to complete successfully [18:31] Bad disc? [18:35] <__mapo__> you mean the installation disc? how can it be the cause of a dependency error? [18:37] If you've tried repeatedly and it's refusing to get past it, that's something to look in to. [18:38] <__mapo__> I totally agree, but at this point I'm not sure I'm willing to re-install the machine again [18:42] When I installed my home server I wound up spending about a day banging my head against getting it to work [18:43] <__mapo__> how do I clean the apt package list? [23:31] Greetings! I've been ripping my hair out all week trying to figure out how to name serve my own domain. I tried installing bind but after struggling with it for hours and days, I couldn't get it working and so that led me to dnsmasq which led me to Network Manager & dhcp-client and then finally resolvconf. I've read through dozens of setup guides, articles and likely hundreds of posts. So, before I start bleeding from the eyes and/or head, [23:31] what is the Ubuntu "right way" to serve up my own DNS master domain? Is bind the proper way or is it one of these other methods? My router serves up DHCP so I guess I wouldn't want to run dhcp-client I don't think.; v16.04 LTS [23:33] apb1963: a domain for the Internet at large? [23:33] apb1963: does your server have a publicly reachable IP address? [23:33] yes [23:33] Use bind [23:33] ok, in that case the most recent error I'm getting is: rndc[31343]: rndc: connect failed: 127.0.0.1#953: connection refused [23:34] Sounds like bind hasn't started up [23:34] Right. And it won't. Because of that error. [23:34] it's catch-22 [23:34] You don't use rndc to start bind [23:34] Only to signal it once it's running [23:35] I don't use it at all. I simply startup bind. What it does with rndc is beyond me. [23:35] let me paste it for you.... hang on [23:35] apb1963: have you created a zone file for your domain? [23:36] rbasak, https://pastebin.com/csesAXn9 [23:36] TJ-, Yes [23:37] Looks like you're missing some of bind's log. [23:37] Try sylog. [23:37] syslog [23:39] apb1963: did you use named-checkconf and named-checkzone ? [23:41] tj- yes. all ok [23:41] named[31329]: open: /etc/bind/named.conf: permission denied [23:41] in syslog [23:41] ll /etc/bind/named.conf [23:41] -rw-r--r-- 1 root bind 463 Nov 8 05:00 /etc/bind/named.conf [23:43] apb1963: how about the directory: "ls -ld /etc/bind" [23:43] drw-r-Sr-- 3 root bind 4096 Nov 25 13:25 /etc/bind [23:44] That looks pretty screwed up. [23:44] What should it be? [23:44] I'd have to fire up a container and check. [23:45] 755 or perhaps 750 I expect. [23:45] That looks correct; I've jsut compared it [23:45] drwxr-sr-x 2 root bind 4096 Nov 25 23:45 /etc/bind [23:45] In a Xenial container [23:46] hmm... I don't know the difference between s and S [23:46] apb1963: is there another service listening on port 53? [23:46] S is missing the executable bit. [23:46] tcp 0 0 127.0.1.1:53 0.0.0.0:* LISTEN 4106/dnsmasq [23:47] nobody 4106 1176 0 Nov18 ? 00:00:07 /usr/sbin/dnsmasq --no-resolv --keep-in-foreground --no-hosts --bind-interfaces --pid-file=/var/run/NetworkManager/dnsmasq.pid --listen-address=127.0.1.1 --cache-size=0 --conf-file=/dev/null --proxy-dnssec --enable-dbus=org.freedesktop.NetworkManager.dnsmasq --conf-dir=/etc/NetworkManager/dnsmasq.d [23:48] Seems to be tied to networkmanager perhaps? [23:49] Yes, it is. Not sure if that'd cause named not to start though. Depends what interfaces you've told named to bind to [23:50] I don't understand. Your bind is broken because your /etc/bind permissions are screwed up. Fix those first. [23:51] It's an apparmor issue I think [23:51] No, it's a permission issue on /etc/bind. [23:51] You might *then* have an issue with something else binding to 127.0.1.1:53, but you're not even getting that far. [23:53] ok, reinstalling just to verify I didn't hose it myself... or if I did. [23:53] ll /etc/bind [23:53] ls: cannot access '/etc/bind': No such file or directory [23:53] there's proof :) [23:54] I guess it was me [23:55] drwxr-sr-x 2 root bind 4096 Nov 25 15:54 /etc/bind/ [23:56] Active: active (running) since Sat 2017-11-25 15:55:40 PST; 11s ago [23:56] Boom [23:56] I'm curious.. why doesn't apt-get purge clean out config files and dirs? [23:57] I would have expected /etc/bind to be purged... in the purge. [23:57] oh. Thanks for your help guys!!!! I should have come here before spending dozens of hours on it. Oh well :) [23:57] Purging the bind9 package does clean out /etc/bind for me in a Xenial container. [23:58] You're welcome :) [23:58] it doesn't if you add any files [23:58] Plus it should also delete the dir itself. [23:59] dpkg: warning: while removing bind9, directory '/etc/bind' not empty so not removed [23:59] right [23:59] so, nice warning... cool. But not the documented behavior of purge [23:59] The packaging system doesn't know if files from other packages are left in there. [23:59] It's by design. [23:59] Purge will only remove stuff shipped or managed by the package itself.