=== Guest54827 is now known as awutz [02:19] I'm going to be setting maas again. It broke my system last time [02:19] I have it up and running now [02:20] what do you people suggest to approach the problem? [02:21] should i use the live cd? [02:22] anyone? === dendrobates is now known as dendro-afk [02:34] What problem? [02:34] "maas broke my system" is not a meaningful problem description [02:35] It broke apt-get [02:37] Telling us that something is broken is fairly useless. You need to tell us exactly what is failing, and what you expected it to do. [02:37] "Doesn't work" is a vague statement. Does it sit on the couch all day long? Does it procrastinate doing the dishes? Does it beg on the street for change? Please be specific! Define 'it' and what it isn't doing. Give us more details so we can help you without needing to ask basic questions like "what's the error message?". Ask me about , and . [02:40] password authenication failed [02:40] that was the error message [02:41] apt-get said that? [02:43] that is when I was trying to use maas [02:43] the error message for apt-get was it locked by a process and which it wasn't [02:46] I'm sorry, but I don't have the patience to get meaningful diagnostics out of you through dialogue. Please wait patiently for someone else to help. [02:46] you don't need to [02:46] I reinstall the system already [02:46] I did this a week ago [02:46] I just downloaded the ubuntu-cloud iso [02:47] I was wondering what is the preferred method to install maas [02:47] I did an apt-get install through ubuntu software center [02:47] Oh, I see. Sorry, I thought you were asking for help with your original issue, so it didn't happen again. [02:47] no [02:47] it didn't happen again [02:47] I'm sorry if you got that idea [02:47] no [02:49] Don't worry about it. [02:50] back to the issue [02:50] what is the preferred method to install maas? [02:50] I have the iso [02:51] I don't use maas, so I don't know. [02:51] thats fine [02:51] thank you. [03:35] hey, guys. i want ufw to allow connections from localhost to be able to access any port. help? [03:35] *any port on localhost [03:36] That's probably on by default [03:36] josheee12: please pastebin the output of "iptables-save -c"; I will check for you [03:37] momentarily. keep in mind this is an assumption it's not working, because my varnish cache server isn't turning up content after enabling ufw [03:41] http://pastebin.com/pQcNnnsK [03:48] any ideas? my issue is that i use varnish to front all my http connections. based on the host, you'll get content because varnish passes throgh to a different nginx instance. varnish keeps giving me an nginx 404 page, even with caching off. [03:51] Sorry, I was in a meeting [03:51] 74. [26:35261] -A ufw-before-input -i lo -j ACCEPT [03:52] ^^ that line says everything on loopback is allowed, and it has matched 26 packets (or more likely, flows) since ufw was last loaded [03:52] twb: Error: "^" is not a valid command. [03:52] any idea, then? [03:53] Have you checked logs? Have you tcpdumped? [03:53] You can establish quickly that ufw is not responsible, by turning ufw off entirely for five minutes [03:54] alright, thanks [03:54] can i pm you? [03:54] Sure, but technical discussion should remain in-channel [03:55] i'd just rather not give my domain out if i don't have to since i'm going to expose a current gaping security flaw on my server [03:55] Whatever [04:03] why is tasksel and byobu display screwed up in 12.04 on ec2? === pehden is now known as pehdens-sleeping [04:09] josheee12: so the problem is fixed? [04:16] I'm getting segment fault in libc-2.13.so when I mount -a [04:16] fstab: 192.168.1.92:/Volumes/StudentsHome/DragonNet /var/DragonNet nfs [04:19] That line is not complete [04:20] You need a mount options field like "default" or "intr,bg" [04:23] now it says missing helper program? [04:24] Is nfs-common installed? [04:25] Also "nfs" means NFS3; you may need NFS4 if the server is some OS X thing [04:25] (It is worth trying nfsv3 first) [04:26] works :) [04:46] Hey guys I have a question , when I do an fdisk -l im getting strange output. Dev disk/mapper/server-root for example? [04:46] anyone know what thats all about? [04:48] Do not do "fdisk -l". [04:48] Reliable information resides in /proc/partitions [04:49] And that is a devicemapper device, most likely an LVM LV. See also output from pvs, lvs, vgs commands [04:49] not sure what any of that means [04:49] but checking out /proc/partitions [04:50] partitions has lines i dont understand either. sr0, dm-0, dm-1 [04:50] sr0 is the CD drive. [04:51] ahh [04:51] dm-0 is again part of device mapper. /proc/partitions uses dm-0 rather than the LV name; you can resolve one into the other by looking in e.g. /dev/disks/ or /dev/mapper [04:51] dm-N will have a device minor number of N [04:51] wait a minute [04:51] is this because I installes server with LVM? [04:51] yes [04:51] which explains why I'm ignorant [04:51] as thats new to me [04:51] Otherwise instead of dm-0 you'd see something like sda2 [04:52] But its nothing I need to worry about. [04:52] well when I install manually [04:52] I only ever have two maybe 3 partitions [04:52] primary, swap, and maybe a backup partition [04:52] thats why it through me off [04:52] also I thought it was weird that server installed as 1 2 and 5 [04:53] That's ubuntu being silly [04:54] eventually I want top ut a larger harddrive into my server [04:54] When using MS-DOS disk labels ("partition table"), you have four "primary" partitions, of which one may begin "extended" partitioning. Linux numbers the extended partitions from 5. [04:54] and then eventually figure out a way to get it to to automatically back important files up to a repository [04:54] d-i partman for some reason defaults to allocating partitions as extended partitions immediately, so you typically see 1,5,6,7 instead of 1,2,3,5,6,... [04:55] KXTwo: etckeeper, and also keep /home (user data) on a separate fs from / (OS) [04:56] some one else was saying something about that [04:56] though currently im not entirely sure how to set that up but ll get to it [04:56] probably me [04:56] im reading the ubuntu guide [04:56] I think ti was your name does look familiar [04:56] but actually my server data is in var/www [04:57] So fix that [04:57] its a web server? [04:57] thats where I want it? [04:59] That's not where it belongs. [04:59] FHS clearly says it belongs in /srv; /var/www is a hysterical remnant for silly people [05:00] thats where the guide said to put it! [05:00] lol [05:01] what is FHS [05:01] The Filesystem Hierarchy Standard [05:02] https://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard [05:03] why does it matter so much? [05:03] var/www is where it looks fo the index.html? [05:03] it doesn't really matter [05:04] well if there is a learning opportunity or something to learn that will improve performance then I'm all about it, but if it doesnt matter lol [05:33] Has anyone here compared version ircd-seven-1.1.3 with Bahamut with Ubuntu, Debian or FreeBSD any comments? [05:36] Poindexter_: we support what's in the main apt archive. If you're running third-party irc daemons, you might have better luck talking to #freenode or something. [05:37] I tried that. It was useless. [05:37] Talking to #freenode was useless? [05:37] I didn't say that. [05:37] The people were non-descript.. [05:39] Setting up an IRC server is not a speciality with most webmasters. Especially with setting up a website interface to interact with an IRC server with clients. [05:40] Some issues have to do with PHP components of Linux and others have to do with AJAX additions. [05:41] Don't use PHP. http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/ [05:41] I am not a fan of PHP though. [05:43] anyone use ebox? [05:44] I have read tons of articles on .htaccess and the do's and dont [05:44] s [05:44] about it. [05:44] IMO htaccess falls into the "don't" category. [05:44] Just put everything in /etc/apache2 where it belongs; htaccess has too many hidden gotchas. [05:45] !anyone > KXTwo [05:45] KXTwo, please see my private message [05:45] I could imagine. Thanks for the heads up. [05:45] Poindexter_: IIRC it's best to disable the module entirely. [05:46] so uptight lol [05:46] whats the difference between ssh and ebox? [05:46] Um, ssh is a secure remote shell; ebox is a WHCP [05:48] Twb Freenode uses version ircd-seven-1.1.3 Have you installed and used the beast? [05:48] Poindexter_: I use ircd2 [05:49] What is the advantage? [05:49] It's in apt, so it (theoretically) receives security maintenance [05:49] That is abstruse. Explain please? [05:50] Well, do you understand why distros exist? [05:50] i've on ly ever used SSH, I do not know what WHCP is, or if there are any advantages disadvantages with ebox [05:51] KXTwo: WHCP is Web Hosting Control Panel. The idea is they provide a web interface so idiots can be sysadmins. Which doesn't work. [05:51] Distros exist because someone has a different idea about how things should be run. [05:51] so stick with ssh [05:51] Poindexter_: distros exist to take upstream code and *integrate* it, so it works as a cohesive whole, and to provide security and critical bugfix updates for it. [05:52] Poindexter_: so the argument for using ircd2 is that it's in Ubuntu, so it receives the standard benefits of being part of Ubuntu (as opposed to being third-party software) [05:52] Twb, that makes powerful sense. [05:52] KXTwo: exactly. If you are already comfortable with command-line use, there is zero point to ebox [05:53] twb: thanks, again i ask a lot of questions just for learning [05:54] KXTwo without questions there would never be answers. Keep asking. [05:55] KXTwo: no worries. It's in my best interests to reduce the idiocy on the greater internet [05:55] https://en.wikipedia.org/wiki/Endless_September [05:56] twb: I am far from an idiot [05:56] simply ignorant on all matters related to ubuntu server :0 [05:56] Everyone SAYS that :P [05:56] twb: well I have credentials lol [05:56] and most of my questions are fair [05:57] this guide is very informative [05:57] but chapter 6, network authentication, I am not sure what you use openLDAP for. [05:57] I'm an asshole; don't take it personally [05:57] twb: Im an asshole too, just not an idiot haha [05:58] LDAP is how you get hosts to have the same set of users and groups [05:58] (It is a lot more flexible than that, but that's it's primary use) [05:58] give me a an example of why I'd use it? [05:58] KXTwo: because you have >>1 host [06:00] as in multiple servers? [06:00] Yes [06:00] Or if you have one server and a hundred desktops [06:01] If you're familiar with Microsoft Active Directory, that is kerberized LDAP under the hood [06:01] ok well it sounds like for now I don't need LDAP [06:01] I just have one crappy web server lol [06:02] so will skip that chapter lol [06:02] now I do feel like an idiot, but I swear I'm smart haha [06:03] Sure; the server guide is a reference, not a recipe. You don't have to do everything it says [06:03] again im jjust reading for learning [06:03] I didnt realize I was in the optional section [06:03] i do want to to DNS though so gotta read this one [06:04] Eh, personally I don't like their use of bind; I prefer nsd3/unbound [06:05] hey twb, you definitely know your shit. I have a question about nmap. my phone also connects to my wireless but when I do nmap on my network it doesnt show up. I can map directly to it but never with just a nmap "network", do you know anytyhing about that? [06:05] twb: well i ahve to start somewhere! [06:06] KXTwo: you're doing what, nmap 10.1.2.0/24 ? [06:07] twb: yes and everything shows up but my phone, but if I type nmap "phone ip address" it gives a response [06:07] I cannot explain that [06:07] yah it seems no one can yet lol [06:13] Why would you want to use NMAP for port scanning your phone? [06:15] why not? [06:15] nmap is meant to find all devices on a network [06:15] other than my phone it works [06:15] it works if my phone is active too which is weird [06:15] Because everything you do with your phone is logged by the phone company. That is not smart. [06:16] Thats not true? [06:16] and even if it was [06:16] im not doing anything to be worried about [06:17] but my wireless provder cannot monitor what i do over wifi lol [06:17] Really? [06:17] Wireshark. [06:17] Yes really [06:18] again not that it would matter [06:18] because im not doing anything to care [06:18] I just connect to the router for downloading updates etc [06:18] more of just a curiosity as to why nmap doesnt discover ti [06:19] 16:15 it works if my phone is active too which is weird [06:19] KXTwo: that strongly implies that it's just the phone disconnects when idle, to save power [06:19] NMAP is a great tool. Its predacessor was SATAN System Admin tool. [06:20] twb: no it doesnt though. I can still ping it, and find it manually? [06:20] Hm, good point [06:20] yah I was right lol [06:20] yay* [06:20] the ? was because I expected you to tell me my assumption was wrong lol [06:21] For now I'm still inclined to write it off as the phone being strange [06:21] ok twb so I am reading this dns chapter and its talking about caching nameserver. it wants me to add my ISP's dns ip's. Why would I do that on a local DNS? [06:22] The phone company may BLOCK promiscuous MODE xmissions. [06:22] oh wait, reading further is always a good idea! [06:23] KXTwo: ultimately, there are DNS root servers. They contain the canonical references for e.g. *.com and *.org. You *could* point your caching resolver directly at them, but it is generally better for everyone if you point it at an intermediary caching resolver (your ISP) [06:24] The main reason not to, is if your ISP are pricks and do Naughty Things to records. [06:25] twb: maybe I am thinking completelyw rong about this. I just wanted a DNS for my local network, so say my network can be named something like personal.net [06:25] so I want to have server.personal.net [06:25] downlt.personal.net [06:25] uplt.personal.net [06:25] am I wrongly thinking about it? [06:25] Twb ISP's throttle bandwidth. That in itself is unforgiveable. [06:25] Poindexter_: if your ISP throttles per flow, get a different ISP. [06:26] KXTwo: for that, I use dnsmasq to provide combined DHCP and DNS services. [06:26] time warne cable definitely throttles [06:26] KXTwo: but you might want to separate network services from application services, and run an OpenWRT appliance for the former and an Ubuntu server for the latter. [06:26] My ISP is OK. I have no problem. But I did find fault with their email port number. I had to do a work around. [06:27] Poindexter_: they should be using submission port (587) [06:27] twb: I only want one server, its all I CAN run right nwo without virtualizing [06:27] Shrug. [06:27] so DNS by itself is not what I want to do that? [06:28] If you want DHCP clients' names to appear in the local domain, you need a DHCP server and a DNS server, and to make them talk to one another. [06:28] dnsmasq obviates the last by doing both jobs and talking to itself. [06:28] My entwork is all static ip's [06:28] but my router has dhcp? [06:29] Twb just curious, is there a way to flush the IP's DNS of your files. Windows uses ipconfig /flushdns command. What do you think? [06:29] Your "router" would be the thing I called an OpenWRT appliance. [06:29] Poindexter_: I don't understand the question. [06:30] ok so im back to needing to setup dns [06:31] I forgot I am talking with Ubuntu folks. Sorry for that question. But I suppose it still applies to the IP DNS server. Not the command ipconfig /flushdns. [06:33] so this dns was a bit harder than I thought lol [06:34] KXTwo: the concepts are difficult to convey, but operationally it is not that hard [06:35] I cant decide WHICH I need [06:35] caching [06:35] primary or secondary [06:35] im assuming primary [06:35] Well, kind of. You're hosting your own realm, but only for your local users [06:35] I have a huge .hosts file I use with my Windows box. When I add just one 127.0.0.1 It takes almost 5 minutes to fix the file. I have thousands of 127.0.0.0.1 entries in the .hosts file. When I ipconfig /flushdns it takes about 15 minutes for the computer to come back to normal. [06:35] kind of? [06:36] "primary server" usually means you are hosting somethig like example.net so the rest of the world can resolve it [06:36] You want something that resolves inside your LAN but nowhere else [06:36] so is that secondary? [06:36] I cant tell a difference in the description between primary and secondary [06:37] A secondary just keeps a backup copy of the primary's zonefiles [06:37] What you probably want is a caching resolver that also has the ability to host a local "fake" domain, although the term used for "fake" differs beween implementations. [06:38] Part of this is confused because bind -- the reference implementation -- does both jobs (serving zonefiles to the world, and caching the world's zonefiles for your LAN) [06:38] I think the guide is talking about setting up one to that works outside the local network [06:38] Probably, yes [06:38] Since usually this job would be done on your appliance router, not on your app server [06:39] ok lets start there [06:39] is me ewanting to do this kind of pointless? [06:39] do I need to have my own domain? [06:39] I was just doing this to learn how to do so [06:40] my lab had a local DNS, but we also had dozes of various machines we would log into for various projcts, including a half a dozen vm servers [06:40] you don't need to buy a domain from a registrar [06:41] But if you only have four boxes, and you don't have a FOSS firmware on your appliance router, it's probably easier to just hard-code things in /etc/hosts on all hosts [06:41] ok lets break that down al ittle [06:41] FOSS/ [06:42] ahh [06:42] free / open source software [06:42] I googled it lol [06:42] why do I need open source router? [06:42] I have a newer LINKSYS N router [06:43] KXTwo: so that you can set up your local domain on it [06:44] I think I can do that on my router? [06:44] I'm assuming the vendor's default firmware is too stupid to do so, otherwise we wouldn't be having this conversation [06:44] Twb have you used VBOX Virtual Box? [06:45] Poindexter_: yes. Use KVM or LXC. [06:45] no [06:46] I have set up lots of Guests with it. The most proficient Guest is Debian LXDE. [06:47] twb: I thought my linksys router could do it but I'm not sure how to do it [06:47] what was your suggestion about /etc/hosts? [06:47] KXTwo: you will need to ask your hardware supplier about that [06:47] KXTwo: /etc/hosts is where you can hard-code entries, instead of using DNS [06:47] would I do that for every machine? [06:48] I will Google KVM and LXC. I think I have seen this before but I will check it out again. [06:48] KXTwo: yes, and you'd have to keep them in sync all the time. Which is why DNS is used for >>1 host sites [06:48] on each node, put all other nodes in there? [06:48] KXTwo: right [06:49] in sync all the time? [06:49] KXTwo: like "10.1.2.4 yoyo.example.net www mail" "10.1.2.5 frisbee.example.net". You'd probably just copy-and-paste the stanza to each machine [06:49] It seems that Vbox is hands down over VMWare. Vmware has lots of issues. [06:49] KXTwo: well suppose you add another machine in three weeks, called "waffle". Now you have to go update ALL hosts files [06:50] Poindexter_: they're both closed-source software owned by oracle, so avoid them. [06:50] ok twb, so whats the reason why you say to do that instead of DNS? [06:50] (Technically vbox has a free core, but IME everyone uses the partially-non-free version.) [06:50] KXTwo: because otherwise you will have to learn how to use DNS [06:50] haha [06:50] I really thought DNS would be simple lol [06:51] I was looking at this page: http://virt.kernelnewbies.org/TechComparison [06:51] but I dont even really know how to set up a domain name for my local network lol [06:51] KXTwo: it's simple iff you've done it before [06:52] yah thats what I figured [06:52] im going to search for another guide [06:54] I ahve to set up a file server to so all my machines can copy data back and forth to others [06:55] KVM is Red Hat? [06:55] KXTwo: are they all unix hosts? [06:56] Poindexter_: kvm is linux. RH funds a lot (30%?) of linux development [06:56] twb, I dual boot a windows machine, and do want any windows host to communicate as well [06:56] Poindexter_: kvm should be documented in the Ubuntu Server Guide [06:56] but personally run mint and ubuntu [06:56] KXTwo: then you probably should deploy samba [06:56] samba file server [06:56] yes I have it installed [06:56] its on my list of things to do lol [06:57] I will download the ISO file when I find it. [06:57] ok here is a possibly dumb question [06:57] when setting up my dns is that HOW I set up a domain name for my local network? [06:58] or does that need to be done somewhere else? [06:58] I found a much better guide but it doesnt say anything, just says copy domain name here [06:58] Poindexter_: uh... apt-get install kvm [06:59] Twb I have a few choices here. http://sourceforge.net/projects/kvm/files/ [06:59] Do not install anything except via apt-get, or you WILL fuck up your system [07:00] Twb is there an ISO file? [07:00] yah why woudl you download the iso? [07:00] just install through apt-get or even aptitude [07:00] What if I am running Debian? [07:01] Poindexter_: ubuntu steals most of their packages from debian, so that advice goes double for debian [07:01] apt advanced packaging tool and is debian [07:03] It is always a new story every day with this stuff. [07:03] ubuntu is based off of debian [07:03] mint is based off of ubuntu lol [07:03] most modern/popular versions of linux are similar [07:03] I know that apt-get is a debian based download system. [07:04] Seeing that Ubuntu has decided to follow the system. [07:04] I think it is a good thing, since Red Hat had the old Packages. [07:05] red hat wow is that even still around? [07:05] lol [07:05] I iused redhat and slackware back in the 90's lol [07:05] KXTwo: yes, RH is still around. It is called "RHEL" and "Fedora" now. [07:05] RPM Red Hat Package Manager. [07:05] what about slackware? [07:05] slackware was the shit back in the day lol [07:06] AFAIK slackware does not have the same order of market penetration [07:06] Try distrowatch.com [07:06] linux has come so far in the past decade [07:06] it still ticks on, one guy only working on it I think. [07:08] That goes back a ways. I had a Dual SCSI RAID Pentium III server in those days with Red Hat 9. [07:08] ok question twb, I found a better written guide but when editing the named.conf.local he has only 3 octets in his network address, whats that all about? [07:09] KXTwo: read https://en.wikipedia.org/wiki/CIDR and https://en.wikipedia.org/wiki/Classful_network [07:10] im familiar with the difference between classful and classless actually [07:10] am I not making a connection I should? [07:10] Well he is probably just talking about a /24 [07:11] ok will the file except 4 octets with /24 [07:11] That brings me back to my old days at a Network Academy studying for CCNA classes. IP addresses. [07:12] Slash 24 [07:12] I just took network fundamentals lol [07:13] Using Pumpkin telnet software. [07:13] even at my lat the dns file has only 3 octets [07:14] lab* [07:15] 2 4 8 16 32 64 128 [07:16] also do not get why its in reverse order but ok [07:17] 128 64 32 16 8 4 2 [07:17] wasnt talking about that [07:18] and you are missing your 1's place :) [07:18] Carry the 1? [07:18] hehe [07:19] I always liked Anding the octets. [07:21] ok theres is a file /etc/bind/zones/domain.name.db I need to create [07:21] Dot decimal notation. [07:21] You may like this website though. [07:21] he says I have to put in the name of my DNS [07:21] http://www.subnetmask.info/ [07:28] okt his is weird [07:29] I dont knwo if i have a working dns server or not lol [07:32] I left the 1 out because it is not a binary number by itself. It needs a Zero to help it. [07:37] Always remember 1 is not a binary number. [07:39] KXTwo: use dig to interrogate DNS servers [07:40] <_ruben> Poindexter_: why not? [07:40] Ruben, one 1 is singular. Not Two. [07:41] Binary means two. [07:41] thats what the guid esaid [07:41] but I dont know how to read the response [07:41] im assuming yes [07:42] forgive me, the channel is dead [07:42] / replace example.com with your domain name, and 192.168.0.1 with the address of your new DNS server. [07:42] search example.com [07:42] nameserver 192.168.0.1 [07:42] wtf [07:42] http://pastebin.com/UFnQawP1 [07:42] guess it was a hint not to do what I was going to lol [07:43] 192.168.0.1 is your router address isn't it? [07:43] I think the dns server is working, but the guide doesnt tell me how to add files to it [07:43] not mine, that was the guides example [07:43] KXTwo: try dig @8.8.8.8 google.com [07:44] this should give you successfull dns lookups for google using the google 8.8.8.8 nameserver [07:44] to show a successful lookup [07:44] im testing if my own personal lan dns is working [07:45] right.. so then use dig @192.168.0.1 google.com [07:45] (if your local lan DNS ip is 192.168.0.1 ) [07:45] do you know which ip your local lan DNS use? [07:46] yes [07:46] 1.150 [07:46] is it working? [07:46] I guess? [07:46] you can verify using dig [07:46] I dont know what I'm looking for in the response [07:46] just says query time 380ms [07:47] then spits back my servers ip address [07:47] KXTwo: if the dns lookup is sucessful then you should see an ANSWER SECTION line like google.com. 300 IN A 173.194.32.14 [07:47] I dont think tis working [07:47] whatever i give up for tonight [07:47] this means that the DNS server tells your system that you can find google.com at the 173.194.32.14 [07:48] thought this would be simpler lol [07:48] if it was workign I should be able to ping my server by name from a node but I can't [07:48] KXTwo: which DNS server do your DNS server use? [07:49] maybe [07:49] bind if I udnerstand your question but I probably dont [07:51] I fllowed the instructions [07:51] and my name server is server.fashizzle.lan [07:51] so it would seem is hould be able to ping that but cannot [07:53] KXTwo: have you configured your DHCP server to tell all computers attached to your network to use this nameserver? [07:53] i havent gottehn that far [07:53] but everything is static thats on now anyway [07:53] KXTwo: please post a link to the guide you followed [07:55] xranby: he has a small local lan. He wants local machines to be able to resolve one another via DNS, to avoid having to edit /etc/hosts on each machine and keep them in sync. He doesn't need to host his local domain such that the world can see it. He has an ubuntu server and some kind of appliance router which runs default firmware. [07:55] xranby: ^^ that's the basic dump of what we've learned so far. [07:57] http://ubuntuforums.org/showthread.php?t=236093 [07:57] "Install Ubuntu dapper" -- that guide is very old [07:58] If you're going to follow "some web page I found", make sure it's up-to-date! [07:58] its for bind, should work in theory lol [07:58] Only because bind is so stale [07:59] If you were askng about zone "0.168.192.in-addr.arpa" before, that's how PTR records (reverse resoluton) works. [07:59] ill play with this more tomorrow and hopefully get it going out [07:59] it doesnt tell me where to add hosts [07:59] "getent hosts 1.2.3.4" corresponds to "dig -t PTR 4.3.2.1.in-addr.arpa" [07:59] or do the dhcp [08:00] twb ho the fuck do you remember so much [08:01] even commands l ol [08:01] no girlfriend [08:01] And https://en.wikipedia.org/wiki/Asperger_syndrome [08:01] I still have to look up commands that ive used a dozen times lol [08:01] <_ruben> a working brain helps... [08:01] hah [08:01] twb: thank you, for the summary. [08:01] I have one of those, a strong one in fact :) [08:02] KXTwo: what does she bench [08:02] I was referring to a working brain lol [08:02] KXTwo: i think you "simply" need to edit the /etc/bind/zones/example.com.db and /etc/bind/zones/rev.0.168.192.in-addr.arpa to add hosts and reverse records [08:02] my girlfriend is a petit little thing I don't think she could bench the bar lol [08:02] <_ruben> then you should be able to remember things as well [08:02] she does mostlyc ardio lol [08:02] xranby: that was my guess as well [08:03] _ruben: I think normals waste their memory on unimportant things like their address, phone number, which sock is the left sock [08:03] speaking of my massively intelligent brain, I need to rest it lol [08:04] KXTwo: for you lan setup start with creating two files /etc/bind/zones/fashizzle.lan.db and /etc/bind/zones/rev.1.168.192.in-addr.arpa [08:04] I did that xranby [08:05] The names of the files aren't important; as long as they're referred to by named.conf [08:05] I appreciate the added info but im about to crash so im not gaining anything right now, hopefully ill catch you guys tomorrow [08:06] KXTwo: ok good night [08:21] hi guys, im trying to find a good startup script for vnc4server but im really struggling, is anyone able to link me? [08:24] quick question: my site works when you go to https://site.site.com but not http://site.site.com (loads HTML without a stylesheet) [08:31] acapemont: probably something fishy in the html, a wrong path somewhere? [08:38] Kriz1: what are you trying to achieve? [08:38] Kriz1: generally servers should be configured by CLI, and individual apps can be tunnelled directly with ssh -X [08:39] twb, i just want vnc4server to start up at boot [08:39] morning o/ [08:40] it's live.leapmotion.com [08:40] Kriz1: serving a virtual head, or the same as what's shown on the local (say) vt7? [08:41] twb, yeah a virtual display [08:41] The latter is... nontrivial, especially if a GUI login is used. [08:41] Kriz1: well that should be a simple matter of writing an upstart job [08:41] where would i write that? [08:41] start on runlevel [2345] ; stop on runlevel [^2345] ; exec vnc4server -blah -blah [08:42] Kriz1: in /etc/init/.conf [08:42] There is a manpage about the file format; init(5) I think [08:42] so are all files in /etc/init ran at boot? [08:42] No [08:43] twb: I would kindly ask you to not bring AS into discussion where it doesn't belong [08:43] (or no-girlfriends) [08:43] Tm_T: he asked how I remember stuff [08:44] Kriz1: upstart is "event based" so jobs in there are initated when the relevant events are emitted [08:44] twb: ah, misread, my bad [08:44] Kriz1: in practice, most stuff in there *will* start at boot, but the details are a little more complex [08:44] Tm_T: no problem [08:44] * Tm_T read it that it was general guestion, not directed to one person === Gallomimia_ is now known as Gallomimia [10:24] How do I backup from Ubuntu Server to a Windows server? What is the easiest to get up and running? I was thinking about cygwin+openssh+rsync, but is there something better or easier? [10:25] tdn: you can't [10:25] I can't? [10:25] Not losslessly, anyway [10:25] Please elaborate. [10:26] e.g. hard links would not be preserved except on recent versions of Windows, xattrs and acls probably cannot be preserved at all [10:26] I suppose you could roll it up into e.g. squashfs loopback filesystem, and then push that onto the Windows host... [10:26] <_ruben> gzip + ftp ;) [10:27] cygwin also, at least historically, has some very bad race conditions that are triggered when rsyncing files to/from it over SSH. This condition isn't triggered when running rsync without SSH. [10:28] twb, this is a recent version of windows (2008 server R2) [10:28] I think hard link support was added in nt 6.1, possibly 6.0 [10:29] Anyway, I do not use xattrs, acls, nor do I use hardlinks in what I want to backup (svn repos) [10:29] Ah, well if it's svn the best way to back it up is git-svn ;-) [10:29] Also, I do not require it to be rsync. I was actually hoping for a more "boxed" solution. Like the dejadup that is default on Ubuntu Desktop? [10:30] I wouldn't trust Windows with backups, though [10:30] Best would be to flash that server with ubuntu [10:31] Hell it doesn't even have inodes [11:23] How do i get 7z in precise using command line? [11:29] Mischinka: sudo apt-get install p7zip p7zip-full [11:29] hello guys, i've a strange issue, i can see packets with tcpdump but they're not passing mangle table, pre routing chain in iptables.. any idea? [11:30] e_t_ what is the full command for? [11:30] Mischinka: Are you wanting to comress something with 7z? [11:30] Yes. [11:31] i did this: apt-get install p7zip withouth the p7zip full, and it seems to be working. [11:31] 7z a name_of_archive.7z files_to_add [11:32] Another question, How do I add my ape server to automatically start with nginx? and setup so i can have service aped restart commands? [11:42] How do i add a script to the service commands? and start with server? [12:05] Mischinka: you could have upstart start your app service after nginx starts. See http://upstart.ubuntu.com/cookbook/ [12:17] jdstrand: just to make sure, re bug 1001895, is there a good reason not to have lines like the /var/lib/libvirt/...monitor files in apparmor policy template-able instead of hard-coded in virt-aa-helper? [12:17] Launchpad bug 1001895 in libvirt "apparmor policy for libvirt can't cope with symlinked /var/lib/libvirt" [Low,Confirmed] https://launchpad.net/bugs/1001895 [12:21] hallyn: the thing is the the user is playing a trick on libvirt-- libvirt still believes they are in /var/lib/libvirt and is just moving things around and using symlinks. symlinks are resolved by apparmor for security reasons which is why the guest gets the denial [12:22] hallyn: there isn't much that can be done. /var/lib/libvirt is a compile-time choice [12:23] hallyn: to fix this in the manner that I think you are suggesting, you would need to update the xml to make this configurable, and the security drivers access this configuration option and the security drivers to update the policy [12:23] jdstrand: good point. Should I suggest he build his own pkg? [12:24] hallyn: that would be the easiest thing todo. it would take no insignificant effort to write said patch [12:25] I wasn't suggesting we do it :) if he wanted it i figured he would have to submit the patch [12:25] actually, all that would need to be added to the xml is the 'monitor' bits and the security drivers updated to look at it [12:27] Ok thanks. [12:27] (will comment and move to wishlist) === dendro-afk is now known as dendrobates [12:43] Here in Precise http://cdimage.ubuntu.com/releases/precise/release/ "This image is adjusted to work properly on Mac systems." is that image what I probably want to put Precise on an xserve? [12:50] I did a restart and got a guru meditation wtf? [12:58] sorry what ? [13:00] guru mediation? [13:05] Hi, I have forgotten my root password and I've tried adding "rw init=/bin/bash" and "rw init=/bin/sh" to the kernel startup lines but they all lock up after displaying a error message. What can I do to fix this? More info @ http://ubuntuforums.org/showthread.php?t=1983781 [13:14] mikeeeeey: there shouldn't be a root password [13:16] hi. i'm upgrading to 12.04 with do-release-upgrade, and i see at one point during the process "Updating repository information WARNING: Failed to read mirror file". how can i further troubleshoot what this is actually talking about? [13:16] my google searching has revealed a number of similar discussions, but all from some time ago and none that seems to offer a clear resolution. [13:16] lunaphyte_: check the mirror [13:16] lunaphyte_: check the availability of the files it wants [13:18] ikonia: in sources.list, there are two hostnames referenced: us.archive.ubuntu.com and security.ubuntu.com. how can i determine which mirror it's complaining about, and which files? [13:18] it seems to retrieve a number of items from both during the process [13:18] i'll pastebin [13:18] lunaphyte_: it's us [13:18] lunaphyte_: security is for updates [13:20] http://dpaste.com/750770/ [13:20] see line 92 [13:21] it seems to, at least in some capacity, be able to talk to us.archive.ubuntu.com just fine. [13:22] lunaphyte_: odd, it complains and then carries on [13:22] yeah. [13:22] for me - I wouldn't continue, [13:22] it seems like it might work ok, but i would like to understand it before continuing. [13:22] right [13:22] the upgrades for me worked fine, except grub fails [13:23] i've had all kinds of issues with grub and updates over the years. :( [13:23] I never had any, till now [13:23] that's become almost expected at this point, for me anyway [13:24] ikonia: i know there shouldn't be one either, but I lost all the passwords for my sudo accounts, therefore I need root to reset them [13:30] Question : since ubuntu-vm-builder is near EOL, and I was told here the debian live-build was the new way to go, is there some other documentation than the debian one to achieve the same as with vm-builder ? [13:32] The Ubuntu server guide still mentions vm-builder as the way to create VMs [13:34] mikeeeeey: just boot into single suer mode [13:34] mikeeeeey: you've lost all the passwords for all your sudo accounts [13:43] ikonia: that was what i was going to do, but i get asked for the root password, but root is locked with "passwd -l" so it doesn't have one. [13:43] mikeeeeey: it shouldn't ask for the root password unless you have set it [13:43] if you've not set it, it shouldn't ask, did you set the root password ? [13:44] ikonia: the hoster set the password, i created a user for myself and gave it sudo, i then locked root [13:45] mikeeeeey: is it a proper ubuntu install, or a a virtual/xen/style thing [13:45] ikonia: it is a proper install, the machine is a Dell R210II [13:46] you're going to have to boot from a CD and remove the password field from the shadow file [13:46] you should always be able to edit your kernel cmdline and just add init=/bin/sh, then mount your rtoofs and chroot into it to reset the passwords [13:46] ogra_: this is what i have been trying to do, but it errors upon me all the time [13:46] ogra_: unless it's one of the virtualized / bastardised installs, which is why I was querying it [13:46] like, it does boot [13:47] iirc there were issues with sulogin (single user mode) and locked rootpw [13:47] it does drop me to the "shell" but i can't do anything [13:47] its like its locked [13:47] doesn't register anything [13:47] keybaord doesn't even work [13:47] you should be able to mount your rootfs [13:47] oh, no keyboard is special [13:47] ogra_: cannot set terminal process group (-1) inappropriate ioctl for device ubuntu [13:47] you can try something else: [13:47] this is the one i get with /bin/bash [13:48] add break=bottom intstead of init=/bin/sh [13:48] see if that helps it should stop the boot in the initrd right before it would switch to the rootfs [13:48] so everything should be mounted etc [13:48] okay [13:49] so replace rw init=/bin/bash with that? [13:49] should i replace anything else in the loader too? [13:49] uh, remove anything else* [13:49] or should i just add it at the end of the linux* line [13:49] well, use your default cmdline and just add break=bottom [13:50] no recovery mode? [13:50] nope [13:50] alright, thanks! going to try it [13:50] break=bottom will drop you into the initramfs shell [13:50] I'll be interested in how this works [13:50] with mounted rootfs under /root [13:51] ok, so then i just edit the /etc/shadow file? [13:51] or, /root/etc/shadow [13:51] just use the passwd command [13:51] alright, thanks [13:51] dont edit passwd or shadow directly, the tools usually have checks you dont want to miss [13:52] ok [14:01] Hi [14:02] Got a problem with DRBL and Ubuntu Server 12.04 [14:02] drbl? [14:02] /opt/drbl/sbin/drbl-ocs does not exist :( [14:03] bioman: that is not an ubuntu package [14:03] or it wouldn't be in /opt [14:04] the error is pretty clear "file does not exist" [14:04] and there is no drbl package at all [14:04] ikonia: using the same package as I used to with 10.10, and I had no problem. I know it's not an ubuntu package [14:05] ok, so what's the issue ? [14:05] you're using a package that worked on 10.10 - you're not using 10.10 any more [14:05] the error is "no such file/directory" - there you go, that's the issue [14:06] ikonia: /opt/drbl/sbin/drbl-ocs does not exist [14:06] correct, there is your problem [14:06] ikonia: so how to fix this ? [14:06] ask the drbl people? [14:06] bioman: contact the person who made the package and ask them to include the fix in the installer [14:06] ask the people that made that package? [14:06] patdk-wk: dunno is there are people... [14:07] bioman: where did you get the package ? [14:07] then, your going have to figure it out yourself [14:07] <_ruben> then ask the creatures who created it :) [14:07] I have to say that there was a missing dependency with 12.04. To boot via PXE, I had to install manually tftpd-hpa... [14:08] bioman: where did you get the package ? [14:08] ikonia: ubuntu's repository I think [14:08] drbl from ubuntu's repo / [14:08] ? [14:08] you just said it wasn't an ubuntu package [14:09] drbl is not an ubuntu package, dunno for tftpd-hpa. Is there a way to know it ? [14:10] where did converation about tftpd-hpa. come from ? [14:10] I thought we where talking about drbl [14:10] I'm asking where you got the drbl package from [14:10] bioman: tftpd-hpa provides a tftp *SERVER*, how/why would that be missing from what? [14:10] apt-cache madison [14:10] Daviey: is it an Ubuntu package ? [14:11] that tells you wherev a package comes from [14:11] -v [14:11] Daviey: sorry, am I missing something where is tftp coming into the equasion, the initial question was about drbl ? [14:12] ikonia: tftpd-hpa should have been installed as a drbl dependency I guess [14:12] ikonia: I *think* an unofficial package is missing the depends on tftpd-hpa.. which is not something we can resolve. [14:12] bioman, so talk to the creators of the drbl package and make them fix that bug :) [14:12] bioman: drbl shouldn't depend on a TFTP server [14:12] Daviey: totally, I just couldn't see how we'd got from a missing file in a DRBL package to a TFTP server [14:13] iirc DRBL is a hackish veariant of LTSP [14:13] ogra_: ahhh, thank you [14:13] so you will likely need a tftp server to do PXE booting of the thin clients [14:13] there is the missing link [14:13] yup, I see where you are going [14:13] thank you [14:13] ok, gonna try to find. Maybe #drbl exists... [14:13] nontheless this is the wrong forum for that bug :) [14:14] bioman: find the people who made the package [14:14] not drbl - the project, but the people who made the package you are using [14:14] bioman: I'll ask one more time "where did you get the package" [14:16] hey yooooooooooooooool [14:16] ikonia: I've added in /etc/apt/source.list : deb http://drbl.sourceforge.net/drbl-core drbl stable [14:16] nice [14:16] bioman: ok, so the team who maintain that package on sourceforge are who you need to talk to [14:17] bioman: you'll need a log a bug to them - not talk to them on IRC [14:17] bioman: please keep in mind that you told me this package was "the same one" you used on 10.10 - and you are not using 10.10 any more [14:17] ikonia: ok thank you. I'm really annoyed, I have 7 days to make it work :( [14:18] bioman: perhaps a little better planning and research would be well placed in future/going forward ? [14:18] ikonia: when I search with drbl-ocs package not found I find nothing [14:18] file* [14:19] apologies, I don't know what you mean [14:19] ikonia: that's normal, I'm french ^^ [14:19] you're English is good, I'm just not fully understanding [14:19] your [14:20] I mean, I've searched in Google yet, without success [14:21] http://sourceforge.net/projects/drbl/ [14:21] click on "tracker" file a bug [14:23] bioman: I wouldn't expect this to be fixed and released in 7 days though [14:24] use LTSP, it is fully supported in ubuntu ;) [14:24] ikonia: that's what I'm afraid of... [14:24] Where is "sudo command" logged in 12.04? I'm not finding /var/log/secure [14:24] ogra_: have a look, the latest bug is 2010 [14:25] ogra_: thank you but I MUST use DRBL :( [14:25] bioman, i really dont care what their latest bug is, thats the place to file your bugs if you have probs with their packages [14:25] if you want to use something that is supported and gets regular bugfixes, use LTSp [14:26] urthmover: /var/log/auth.log [14:26] henkjan: ok thanks checking now [14:27] Okay, I'm trying to host the data from a wordpress installation on a remote Windows2003 server. I tried mounting /usr/share/wordpress and /var/lib/mysql as cifs mounts to network shares, but that's no good. No symbolic links, not to mention mysqld doesn't like the redirect. for mysql, I can redirect just the database folder, but not sure what to do about wordpress. [14:28] can i create a virtual 'disk' and store it in a file on the cifs share? [14:28] henkjan: I'm seeing the auth attempt for root level privliges....but not seeing the bash command itself....in /var/log/auth.log is there somewhere else the "sudo command" will show up similar to a .bash_history ? [14:29] USER=root ; COMMAND=/bin/ls [14:29] is what i get in my logs [14:29] henkjan: looking again [14:30] ok. I thank you all for your help/advices :) [14:30] henkjan: all my logs appear to have stopped on the May 3rd. I'll check my other test system and see how this works thanks henkjan [14:56] Did anyone successfully use LXC in 12.04? I just installed the packages, but the containers I spawn does not have a network connection. They do have an eth0 card, but it has no ipv4, and it seems that the ipv6 does not have routes to anyone. I'm using the default lxc.conf [14:59] Are there any graphical web authoring packages recommended specifically for the lighttpd web server package? === Guest83754 is now known as wolferz [15:04] how can i totally remove all php from a server? [15:04] What is the right place to ask about Install CD customization? https://help.ubuntu.com/community/InstallCDCustomization is based on hardy and i386 - I need to customize a CD for amd64 - have some questions. [15:05] stgraber: how's your queue of patches for quantal and precise for lxc now? Trivial fix needed for bug 1000174 ... [15:05] Launchpad bug 1000174 in lxc "lxc-net fails to start with bash" [Medium,Triaged] https://launchpad.net/bugs/1000174 [15:08] does Ubuntu have some kind of magic ability to detect when a file system change is made on 1 machine, and will replicate to another? [15:08] much like mysql replication [15:08] I could write a cron that does an rsync at a specific interval, bt would prefer more magic :) [15:08] tash: Look at DRBD [15:12] e_t_: cool stuff [15:14] how can i totally remove package installs? so i can build from source instead? [15:14] any easy way? [15:14] !purge [15:14] To purge all removed but not yet purged packages, use the following command: dpkg -l | awk '/^rc/{print $2}' | sudo xargs dpkg -P [15:14] that is crap. [15:14] sudo apt-get remove --purge foobar [15:15] just sudo apt-get purge [15:15] no need for the "remove" anymore [15:15] ogra_: you and your new fangled shortcuts. [15:15] haha [15:15] hm.. ok i did that and apache is still there [15:15] lol [15:15] well, new since lucid :) [15:16] ogra_: I find it easier to use dselect. [15:16] lol [15:16] rm :) [15:16] * ogra_ never thought Daviey was *that* old :) [15:17] dselect isn't that old! [15:17] 10-12 years ? [15:47] back here :) [15:48] Sorry for the bothering (DRBL and 12.04 server). The fact is -I've checked- there's no package of DRBL for 12.04 yet [15:48] Gonna use 10.10 server and try to update to 3.2.16 kernel [15:49] What's DRBL? [15:49] Ah [15:49] Kinda like LTSP [15:49] Jeeves_: http://drbl.sourceforge.net/ [15:51] The fact is that whe have a 10.10 server DRBL, but whe're having problems with new hardware which is not recognized... [15:51] I found some tuto to update to 3.2.x [15:52] Hope it'll work... [15:52] Why wouldn't it? [15:52] Jeeves_: because 10.10 is made to run 2.6.3x [15:55] It's made to run on the linux kernel [15:55] :) [15:57] Jeeves_: :) [15:58] which is the latest version of module-init-tools on 10.10 server please ? [15:58] what* [15:59] in 10 minutes I install 10.10 server on a VM :) suspense ^^ [16:03] bioman: 10.10 ?! why would you have based a server on 10.10? [16:03] bioman: I'd recommend an immediate upgrade to 11.04 [16:04] bioman: and then consider going all the way to 12.04 [16:05] SpamapS: sorry, typo, it's 10.10... Cannot use 12.04 as there is no DRBL package yet [16:05] SpamapS: holly s*hit 11.10 [16:06] bioman: whats the advantage of DRBL over LTSP? [16:07] SpamapS: dunno. I MUST install DRBL [16:07] SpamapS: source : my boss [16:08] bioman: looks like drbl is in Debian now.. [16:08] bioman: you could just build the Debian package for 12.04 [16:08] SpamapS: dunno how to. I'm an Archlinux user ;) [16:08] bioman: actually its already even in quantal .. [16:09] bioman: so why are you doing this with Ubuntu? [16:09] SpamapS: my boss... [16:10] SpamapS: do you know how to build this package for 12.04 please ? [16:12] bioman: apt-get install ubuntu-dev-tools && backportpackage drbl [16:12] bioman: that *should* get it done [16:14] SpamapS: there's nothing else to do ? [16:15] bioman: backportpackage --build drbl .. I forgot that it doesn't build by default [16:15] bioman: pretty much thats it. You'll probably have to install some build dependencies [16:16] SpamapS: will I be told which dependencies I'll have to install ? [16:16] bioman: no it will pose you a riddle, which you must solve in 20 minutes or the build dependencies change and it starts all over again [16:16] bioman: sorry that was very snarky ;) [16:16] bioman: yes it will show them to you :) [16:18] SpamapS: ^^ You said the drbl is yet in Debian. This command "backportpackage --build drbl" will download the .deb drbl package by itself ? [16:18] Or should I add something in source.list ? [16:20] bioman: it will download the source [16:20] bioman: and build it [16:20] bioman: why are you asking us about this package [16:21] bioman: it's hosted on sourceforge [16:21] ikonia: its in Ubuntu now too [16:21] as of quantal [16:21] SpamapS: then why is bioman saying it's not and downloaded from sourceforge ? [16:21] SpamapS: ok, thank you. I try... [16:22] ikonia: because he's an arch user and he needs our help to find stuff in Ubuntu. :) [16:22] !info drbl [16:22] Package drbl does not exist in precise [16:22] what's the package name ? [16:23] ikonia: hello ;) I added a repo in sources.list but the latest version that sourceforge gives is only for 11.10 [16:23] bioman: yes, we discussed that [16:23] ikonia: its in quantal [16:24] ikonia: and what do you think about SpamapS's fix ? [16:24] ikonia: hence the suggestion to use backportpackage [16:24] SpamapS: what is quantal please ? [16:24] SpamapS: ooh quantal [16:24] sorry, I was reading 12.04 in my head [16:24] bioman: the next release of ubuntu [16:25] !info quantal [16:25] Package quantal does not exist in precise [16:25] !info drbl quantal [16:25] drbl (source: drbl): diskless remote boot, and a disk cloning utility. In component universe, is extra. Version 1.10.31-2 (quantal), package size 2022 kB, installed size 6886 kB [16:25] cool [16:25] * SpamapS never knows how to use th ebots :-P [16:25] ok, so version 1.10 is in quantel [16:29] SpamapS: so it will build a .deb that I'll install with dpkg -i ? [16:42] hallyn: I uploaded all my changes to quantal, will look at SRUing some of them once the package currently in -proposed is accepted in -updates [16:44] stgraber: ok, i pushed the fix to quantal, let me know when you're queuin up the next sru candidate, or just pull in the trivial fix yourself. thanks [16:45] hallyn: I'll go through all our quantal changes to get the list of stuff to SRU so it should get on my list [16:45] cool. thanks [16:51] SpamapS: backportpackage --build drbl says error: Please specify either a working dir or an upload target! === dendrobates is now known as dendro-afk [16:57] SpamapS: please help me... You've told too much or nothing ;) [16:58] How do I set up Ubuntu Server with a static IP (class B)? [16:59] mr-rich, man interfaces [17:01] bioman: sorry I had to go do something else for a bit. ;) [17:01] bioman: did you try reading the man page? ;) [17:01] SpamapS: np [17:02] SpamapS: not yet... [17:02] bioman: hm, I'd have expected more from an Arch user. :) Anyway, create a dir somewhere and pass --workdir=/that/dir [17:03] people still use ip classes? [17:04] patdk-wk: only to prove that they don't understand CIDR :) [17:04] or they had too many cisco classes [17:05] SpamapS: I was just trying to find how-to specify a workdir ;) [17:05] I so wish we could peek 5 years into the future *JUST* at the internet for 5 seconds to see how much IPv6 is not being used [17:05] maybe it will be up to 5% by then [17:07] SpamapS: error : debsign: gpg error occured! Aborting... debuild: fatal error at line 1271: running debsign failed backportpackage: Error: debuild returned 29. [17:10] SpamapS: gpg: /tù/debsign.7ZmHk2X1/drbl_1.10.31-2~precise.dsc: clearsign failed: secret key not available [17:11] I have lighttpd running on my server and created a generic page to test. My keywords are not being published in Google ie I cannot search for the website. What do I need to do? webpage is http://wolferz.sleepingwolf.net [17:20] ask google? we are not google [17:21] and you do know google has stopped using keywords recently [17:21] and is depending on content, that you lack [17:22] patdk-wk: I did not know that... Is there no way to make the site visible to search? [17:22] dunno, I'm not google, and I have no interest in increasing my search ratings [17:22] http://googlewebmastercentral.blogspot.com/2009/09/google-does-not-use-keywords-meta-tag.html [17:22] from years ago [17:22] wonder what that keyword thing was from a few weeks ago [17:24] SpamapS: Googled without luck :'( [17:24] patdk-wk: thank you for the link [17:24] bioman: you just need a gpg key [17:25] bioman: though you should be able to skip that somehow [17:25] SpamapS: that's what I understood (I'm french ^^). How do I please ? [17:27] bioman: gpg --gen-key [17:27] * SpamapS can't see how to tell backportpackage to pass options to not sign the package.. which is annoying :-P [17:33] SpamapS: it fails [17:34] "it needs 284 bytes more" [17:36] bioman: are you on a VM? [17:37] SpamapS: yep [17:40] bioman: sudo apt-get install haveged [17:40] bioman: will generate entropy for you [17:41] hey [17:42] i have a custom c http server, sitting behind a few things (nginx, other proxyies...etc) [17:42] but sometimes it segs, or hangs [17:42] what tool should i use to monitor this, and restart it if need be? [17:43] SpamapS: key generated, but dunno where it is stored... [17:45] SpamapS: same error: debsign: gpg error occured! Aborting... [17:56] bioman: argh [17:56] hello all [17:56] bioman: wait, did you end up with a .dsc file in your work dir? [17:56] SpamapS: no [17:57] bioman: really? I did when I failed to sign [17:57] jamespage, i was trying hard to trouble you for java6 dep needed by hadoop in ubuntu [17:57] oh wait no I did sign [17:57] i finally got it up and working : [17:57] :) [17:57] hello SpamapS [17:58] check_call(['debuild', '--no-lintian', '-S', '-nc', '-sa'], cwd=srcdir) [17:58] No way to change that. *hrm* [17:59] SpamapS: have to leave for 20 minutes. Will you be there ? [17:59] bioman: ok so use this instead... 'pull-lp-source drbl quantal && dpkg-source -x drbl_1.10.31-2.dsc && cd drbl-1.10.31 && debuild binary' [17:59] bioman: I'll be in and out [18:00] SpamapS: hope you'll be there :) === bioman is now known as bioman|afk === Lcawte is now known as Lcawte|Away === matsubara is now known as matsubara-lunch [18:20] I am having a problem with my server i tried to install ubuntu 12.04 on it and now every time i boot i have 1 drive failure and another degraded [18:20] i have switched out the drives i get the same i have changed out the perc controllers and the raid controllers same thing === koolhead17 is now known as koolhead17|afk [18:28] I have a couple questions regarding Dovecot/Postfix installation and implemination in a virtual host environment. I have a VPS running Ubuntu Server 11.10 64bit [18:28] Was wondering if anyone can offer their assistance === bioman|afk is now known as bioman [18:57] Please how to know the version of a package ? [19:05] I have an issue. I've installed on a VM Ubuntu Server 10.10. I've installed kernel 3.3.6. From grub, the default does not boot (black screen). If I choose troubleshooting mode, it starts. If I then choose "resume" it runs fine. What's going on please ? [19:06] why are you using 10.10? [19:06] iirc it's not even supported anymore [19:07] 10.10 support ended 2012-04-10 [19:07] 10.04 server support ends 2015-04 [19:07] don't use non-LTS versions unless you know what you're doing ;) [19:09] RoyK: sorry, typo, it's 11.10 [19:10] RoyK: after -update and -upgrade, I had kernel 3.0.0 [19:10] RoyK: I thought I could install 3.3.6 [19:10] oh - custom compiled kernel? [19:11] why did you do that? anything that doesn't work with the default kernel? [19:12] RoyK: no, download from here : http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.3.6-precise/ [19:12] ok, still, why? [19:12] RoyK: it's for a DRBL. I need the "latest" kernel because of hardware detection [19:13] ok [19:14] RoyK: what's the difference between generic and generic-pae please ? [19:15] bioman: pae is "physical address expansion", to be able to address >3GB on 32bit machines [19:16] RoyK: ok, thanks. I've choosen generic. But why do I have a black screen when I launch default ? [19:16] bioman: are you running 32bit? [19:16] bioman: no idea, might be a bug [19:16] bioman: is this fixed in precise? [19:17] if so, an upgrade might be easier to get support for than some special kernel [19:17] (also, if running 32bit, why?) [19:19] zul: jdstrand: the pc-0.12 machine type is become quite a headache in precise. I wonder what we can do to help people. Do we just announce that people should update? Do we provide a script? do we automatically use pc-1.0 even though people say 0.12? [19:19] (the latter is likely bad...) [19:20] what's pc-0.12? [19:21] older machine type... [19:21] (in libvirt) [19:21] k [19:21] hallyn: I wonder if we just declare 0.12 as bad and then maybe downgrade them to 0.11? [19:22] is 0.11 better? [19:22] well, I don't know-- I assume it is closer than 12 [19:22] err [19:22] closer to 0.12 than 1.0 [19:22] we could output a warning message [19:23] (or log it) [19:23] and then a script to migrate [19:23] thing is oneiric is supposed to default to 0.14, but i never see that anywhere... of course that may just be bc it works for them so there's no bugs :) [19:23] eg: [19:23] hm - where would we output the mssage? [19:23] (sorry, will wait for your example :) [19:24] RoyK: 32bits because the laptop have 2Go of RAM [19:24] bioman: ic === matsubara-lunch is now known as matsubara [19:24] "XML machine type is spcified as 0.12, but this is a known bad machine type. Using 0.14 instead." [19:25] bioman: but I think upgrading to precise would be better if the kernel there supports what you need (currently at 3.2.0-24-generic-pae or so) [19:25] hallyn: I was thinking libvirt on stderr [19:25] RoyK: ok thanks :) [19:25] hallyn: it could be logged as well, but logging is pretty quiet by default [19:25] s/as well/instead/ [19:25] jdstrand: well we could make this the log-no-matter-what loglevel :) [19:26] true. this is only for precise? [19:26] that's wher ei've been seein gthe bugs, yes. [19:26] that's not to say the same bugs don't exist on oneiric [19:27] what is weird is that 0.12 worked on earlier releases [19:27] right it probably is only precise [19:27] and the latest bug, problem is, was reported on migration from oneiric->precise [19:27] that is where I hit stuff [19:27] so our message would show up on the precise server, but it needs to be updated on the oneiric server [19:28] I had vms created at some point that default to 0.12, then upgraded libvrt to precise and they don't work well [19:28] (must have been created during natty?) [19:29] possibly [19:29] of course it's possible that there's another option - fixing the 0.12 bugs in precise :) but i'm not optimistic on that [19:29] (there appear to be many) [19:29] I just wonder why there are so many 0.12 bugs on precise [19:29] do you know what would be a reason why someone woul dprefer 0.12 to 0.14 (on oneiric) or 1.0 (on precise)? [19:30] i.e. is there any downside to automatically updating, with a log msg? [19:30] automatically upgrading/downgrading probably would break stuff-- if someone really wanted 0.12, they couldn't get it [19:30] hallyn: no idea [19:30] hallyn: might be better to suggest 0.14 or 1.0 and havea tool to help [19:31] hallyn: another option is not to patch libvirt at all, but to provide the tool and in postinst unconditionally upgrade to 0.14 [19:31] hallyn: but only when upgrading from below a particular version [19:32] that will allow new machines to use 0.12 if the user really knows what he/she is doing. and it could arguably be justified as just part of upgradign to precise [19:33] where are these pc-0.xx things documented? [19:34] i don't know, if it's not safe to do it unconditionally at domain start time, then it's not safe to do it unconditionally at upgrade [19:34] and that doesn' thelp if they have some script that downloads existing xml templates with 'pc=0.12' in them (as i actually have) [19:34] they'll just be creating new problem vms after the upgrade [19:35] I think I'm fine with a patch to libvirt to spit out a warning any time a vm with pc=0.12 gets started or defined. but is there then another way we can publicize this, sort of like post-release release notes? [19:36] there is the NEWS file, but that seems heavy handed. you could also update the release notes so post release upgraders see it [19:37] actually, there is a way to mark those. you might ask skaet (idr off-hand). it might just be the SRU process makes it show up [19:37] in that case, what exactly are the release notes? :) [19:37] ok [19:37] 'virsh capabilities' doesn't give any hint of differences between pc-0.12 and pc-0.?? [19:38] all right i'll comment the bug for now and think about it for a half day [19:38] thanks jdstrand [19:38] (and then talk to skaet :) [19:38] sure thing. it is not an easy problem [19:40] jdstrand: actually... if we wanted to offer a script for users to automatically update their VMs, where should that live? Or should we just offer it online and advertise users can download it if they want it? (I don't want to pollute /usr/bin with that...) [19:40] oh, i guess under /usr/share/doc/libvirt-bin [19:41] hallyn: we have precedent for this. see /usr/sbin/libvirt-migrate-qemu-disks [19:41] hallyn: in that case, we were actually upgrading during postinst-- but that was because a security update would have otherwise broken everyone's vms [19:42] well, not everyone's, but quite a few :) [19:42] yeah, this is not quite so crucial. i really don't want to pollute sbin with that [19:42] RoyK: please how can I remove 3.3.6 packages ? I've installed 3.2.0 but grub displays and load 3.3.6 [19:42] do you think /usr/share/doc/libvirt-bin would be ok? [19:43] RoyK: I've tried dpkg -r with no luck [19:43] hallyn: it would, but I think the warning message that is output should be clear on where and how to use it. the advantage of putting it in /usr/sbin or similar is you can do a man page, etc [19:43] I mean you could do a manpage that talks about using a script in /usr/share/doc too... [19:44] jdstrand: ok, thanks. I'll let that process for a half day [19:44] bioman: hm.. should work. they're just packages, right? [19:44] RoyK: yes, they are .deb [19:44] zul: your libvirt package builds fine for me on precise. But I get test failures, all apparently from "memory unit='KiB'>219136 hm. i misread the error :) [19:44] I guess the testscripts need to be updated. sorry :) [19:45] dpkg --reemove --purge [19:45] s/ree/re/ [19:45] bioman: try that [19:46] RoyK: conflict --remove --purge [19:47] just --purge it, then [19:47] I was thinking apt-get remove --purge [19:48] dpkg --purge linux-headers : no installed package matching linux-headers [19:49] bioman: that's a metapackage [19:49] bioman: dpkg -l | grep 3.3.6 [19:52] RoyK: thanks :) [19:52] smoser: ping [19:52] kirkland, here. [19:52] smoser: I'm having trouble running ntpdate from within aws [19:52] smoser: have you seen anything like this? [19:52] smoser: this is 12.04; I *know* I've used ntpdate before [19:53] smoser: having trouble triaging this, though I'd ask if anyone else is having trouble [19:54] kirkland, i just ran: [19:54] $ sudo ntpdate ntp.ubuntu.com [19:54] 21 May 19:53:46 ntpdate[4352]: adjust time server 91.189.94.4 offset -0.101924 sec [19:54] smoser: hmm [19:54] (in ec2) [19:54] smoser: all ports open? [19:55] http://paste.ubuntu.com/999650/ [19:55] smoser: PERMISSION 950047163771 default ALLOWS udp 0 65535 GRPNAME default [19:57] smoser: looks like that's all udp ports are open, right? [19:57] i dont think it means that. [19:57] i just checked tcp [19:58] by running 'nc -l 1234' and then trying to connect from outside, and its shutoff [19:58] (but using 8000 rather than 1234 connects as expected) [19:59] smoser: hmm, okay [20:18] kirkland, i'm not saying something isn't broken. but i didn't see it just now. are you able to reproduce ? [20:38] Is it possible to remove a LVM group during installation? Keep getting "Partition in user" error [20:38] Ubuntu Server 12.04 [20:39] I get the error when trying to "Delete volume group". Frustrating! [20:49] so i accidentally added a route that looks like this: host-216-153-17 [20:49] but it wont let me delete it [20:49] i get error [20:49] host-216-153-17: Unknown host === Lcawte|Away is now known as Lcawte [21:26] esuave: that'll be its reverse dns name - type netstat -rn to view the ip address and remove that [21:35] W: Failed to fetch bzip2:/var/lib/apt/lists/partial/127.0.0.1:9999_debian_dists_sid_main_i18n_Translation-en Hash Sum mismatch [21:35] Yay.. not unique to Ubuntu.. :-P [21:36] SpamapS: local repo? [21:36] RoyK: approx [21:37] RoyK: its just skew on my upstream mirror caused by apt not being atomic on updates [21:37] Something we're trying to solve this cycle actually [21:38] ic [21:40] hrm no this seems almost like something else [21:40] getting it on all mirrors [21:43] Translation-en.bz22012-May-21 16:08:073.8Mapplication/x-bzip [21:43] Release2012-May-21 10:26:02227.4Kapplication/octet-stream [21:43] Looks like Translation-en.bz2 was updated but Release was not === Lcawte is now known as Lcawte|Away [22:17] going to give maas another whirl I think. [22:19] I'm wondering how flexible maas is though, I know it uses cobbler as the pxe manager. I was told, though, that you're not supposed to create new systems using cobbler but only through maas. Which means you have to do it manually mac by mac in the web gui [22:20] jkyle: no, the default pxe boot target for machines maas doesn't yet know about get auto enlisted into MAAS> [22:20] you just need to accept them. [22:23] Daviey: alright, so I can just spin up my nodes set to pxe boot and click accept. cool [22:23] jkyle: yup [22:24] jkyle: if you have a list of mac's.. you can add them through the API aswell fwiw. [22:24] but really, just doing a power on.. makes more sense IMO. [22:25] cool, I'll have to start reading the API. I suspect this is a fairly customizable system by just leveraging the underlying utils, like cobbler. I shouldn't have to worry about "breaking" maas by modifying on that level, I'd assume. [22:26] e.g. custom late_commands, postflight scripts, etc. [22:26] jkyle: no, should be good. Note, that more exciting things are in the works. [22:26] jkyle: what custom late commands are on your mind? [22:26] Daviey: there's some funky stuff I have to do with networking to get bonding working...though I should verify the issues haven't been patched [22:27] othe rthan that, standard stuff, like RAID configuration, etc. [22:27] jkyle: MAAS provides a metadata service which runs user-data supplied on first boot.. which is the place where people would put mot late_commands content [22:27] but then, funky networking might be more of a challenge.. if default dhcp doesn't cut it for you. [22:28] jkyle: RAID configuration should happen in the commissioning environment, which happens before the install [22:28] yeah, mine run on first boot. I have a runonce script that adds itself to the init, then executes all scripts in a certain directory in lexical order [22:28] (note that, as shipped.. the comissioning enviroment is not as accessible to customisations as we wanted.. but still viable.) [22:28] Daviey: should [22:28] . . . [22:29] ;) [22:30] jkyle: anyway.. you didn't buy me a drink at UDS.. :( [22:30] these use the megacli utility, so takes a bit of fudging. also the utility doesn't work (officially) for 3.x kernels. but if you spoof the arch, seems to do ok. makes for a quirky config [22:30] Daviey: I didn't see yah! (/me thinks...did i?) [22:31] Daviey: I was getting kp's on our cloud, missed all but like 1/2 a day [22:31] spent most of that with Kyle [22:39] jkyle: if you want to watch an ill-prepared talk, http://www.youtube.com/watch?feature=player_detailpage&v=m3utPU99Wgg#t=2079s [22:39] that was me at UDS. [22:40] Daviey: the best talk of the day! [22:40] lol [22:40] roaksoax: yah, it was a good trip. [22:41] Daviey: haha besides the trip, the talk was good, you definitely have ankerman skills man [22:41] roaksoax: thanks :) [22:43] * koolhead17 wants to have that crowd connector skill too [23:44] www.wegotoyourpc.com catchy aint it? [23:47] pehden: not sure this is the right place :/