=== airtonix_ is now known as airtonix [01:11] New bug: #714932 in squid (main) "package squid 2.7.STABLE7-1ubuntu12 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/714932 [01:33] I have a DHCP server set to a static ip that keeps using DHCP to assign it's own ip address [01:33] There is no DHCP server that I know of. [01:33] running [01:34] anzenketh: to paraphrase: you have a server. It serves DHCP to the network. Its *own* network configuration *should* be static, but it is trying to use DHCP, and you want to know why. [01:34] anzenketh: is that correct? [01:34] Basicly [01:35] OK, is network-manager (NetworkManager) installed? [01:35] It is a server so it should not be [01:35] No GUI [01:37] Double checked nope network-manager is not installed [01:37] network-manager is a daemon, not a GUI. [01:37] But OK. [01:37] Now, pastebin your /etc/network/interfaces [01:38] http://pastebin.com/U3ib1EcV [01:40] Now pastebin the output of "ip a" [01:41] http://pastebin.com/pqUyVR6A [01:50] I was in here earlier asking about DHCP and why a static ip system was accepting a DHCP address [01:50] I forgot who was working with me. [01:52] Well anyone can really help me I have a DHCP server that should have a static ip set that every one in a while it is taking a dynamic IP address [01:53] http://pastebin.com/pqUyVR6A that is my ip a [01:53] That doesn't really make sense anzenketh [01:53] anzenketh: couple of problems there [01:53] anzenketh: i would first find the rogue DHCP server [01:54] Oh I already know what it is it is a crappy modem/router [01:54] anzenketh: well turn it off [01:54] Can't [01:54] Becouse "It Is" [01:54] anzenketh: why? [01:55] Becouse even though I told it to turn off it is not [01:55] like I said crappy firmware [01:55] anzenketh: unplug it from the network, that should do it (for testing at least) [01:56] Problem is it is also the modem [01:56] anzenketh: (for testing at least) [01:56] anzenketh: do that and see what happens when you restart networking [01:57] Well why is the system which is set to a static ip taking a DHCP address [01:57] anzenketh: i already said "couple of problems there" [01:58] Ok [01:58] You mean over reboots anzenketh? [01:58] Sometimes it's static sometimes not kinda thing or? [02:00] No I mean I have in /etc/network/interfaces it set to static [02:00] But it is not keeping the IP I assigned it [02:01] oh well, i'm gone. good luck [02:02] So it just up and changes while the machine is up? [02:02] Yep [02:02] That shouldn't be possible unless you're restarting networking. [02:02] I am not the server is [02:02] Even so with that interfaces file it shouldn't be asking for an IP to begin with. [02:02] Very very odd. [02:02] How do you know it's changing? [02:03] Becouse messasges is braudcasted to the screen [02:03] What message? [02:04] as0t1: Disabled Privacy Extensions [02:04] Might want to look and or pastebin /var/log/messages [02:04] Might be something useful in there as to what is going on. [02:05] http://pastebin.com/DU5Hs8b0 [02:05] That is what is in messages when the issue happens [02:05] That's an OpenVPN adapter. [02:05] Which makes sence due to I have openvpn installed [02:06] Plot thickens then hehe [02:06] I don't have a lot of experience with it. [02:06] Could openvpn be causing this issue? [02:06] considering the nature of the software I'd say yes. [02:07] Have you configured it at all? [02:07] OpenVpn Yes [02:07] Just basic configuration [02:07] I'd try disabling the service if possible and rebooting. [02:07] See if the issue persists. [02:08] I will try that [02:08] Wish I new more about the repeatablity of the issue [02:08] Well that would be a good start as to narrowing it down ;) [02:10] What is this error message about no IPv6 Routers present? [02:13] Do you need IPv6? [02:13] If [02:13] not disable it. [02:13] Think it's one you can ignore regardless. [02:14] ok [02:14] and no I don't [02:14] anzenketh: I can't see why you're getting a DHCP client on eth1 [02:14] yeah me either. [02:15] I'm taking a stab in the dark that openvpn is doing something funky. [02:15] anzenketh: have you restarted the system since configuring static networking? [02:15] I could be way off. [02:15] anzenketh: it's JUST possible that a stale dhclient is still lying around and causing the strangeness [02:15] the actual computer [02:15] You know I don't know if I have [02:18] Well now time to fight a windows virus issue. [02:18] Thanks for your help will let you know how it goes [04:14] greetings, Got a quick question. I have a 320gb and a 1tb SATA hd that I am going to put server on. How would you set it up. going to have DNS, SAMBA, SSH, and LAMPS. Not sure what else yet but those are what I am thinking about. [04:26] What exactly do you mean? [04:26] How to partition the disks? [04:27] If so I'd probably use the 320 for the OS and maybe homes depending if you needs heaps of storage for user homes. [04:27] what would be the best way to divide up the disks [04:28] The 1TB for data storage [04:28] ok [04:28] 320GB is heaps though. [04:29] Unless you plan on having heaps of user data on there. [04:29] In which case maybe put /home on the 1tb. [04:30] going to set up so that it backs up a couple of laptops and for music storage [04:30] You'll probably want that going to the 1TB then. [04:30] :) [04:32] thanks. was wondering if anyone was watching the channel. was bout ready to go to the ubuntu channel and ask [04:36] No worries. [04:36] Does get quiet in here around this time. [04:39] Hi there everybody [04:39] A user directed me to FreeNAS and Openindiana last time I was here, asking about setting up a network share with windows users [04:39] I now have freenas running on vbox, and i was hoping for a little help getting it to work [04:40] does anybody here want to help with that even though it's not a ubuntu server thing? [04:40] or can somebody direct me to the right irc channel? [04:41] Never used FreeNAS myself sorry. [04:41] I assume it's got a nice web front end for configuring it. [04:42] yeah, i'm sure it does. i'm using it in vbox tho cand can't get to the nice web front [04:42] lol [04:42] what do you use for nas? [04:42] SAMBA / NFS and AFP. [04:42] Samba and NFS off my ubuntu server. [04:43] Network File System = nfs right [04:43] can windows see those shares? [04:43] Nope. [04:43] But it can see Samba share. [04:43] That's what you're after. [04:43] yeah, i've had crappy experience with samba [04:43] Depends. [04:43] I haven't had too many issues with it. [04:44] I wouldn't try and use it as a domain mater though in a Windows environment. [04:44] *master [04:44] what's a domain master [04:45] Keeps tracks of all shares on the network. [04:46] Ender: http://www.linuxtopia.org/online_books/network_administration_guides/samba_reference_guide/17_NetworkBrowsing_14.html [04:48] Even using freenas you'll be using samba. [04:48] It's the only open option afaik. [04:48] For windows file sharing on non-microsoft OS. [04:49] yeah that's what i thought [04:49] that's what i've heard anyway [04:49] and read [05:01] is it possible to be able to access a GUI on my ubuntu server [05:01] from a different computer [05:01] like with Remote Desktop [05:03] Ender: by default, Ubuntu Server does not have a GUI. We encourage users to administer their servers from the CLI. [05:04] Ender: as X11 is network-transparent, you can install GUI applications on the server and run them individually on your local machine (a la Citrix), or you can install a full desktop on the server, and then use it locally and/or export itover VNC. [05:05] There is an alpha-quality RDP server for Ubuntu/Linux, but I advise you not to bother with it. [05:06] ok [05:07] are vnc and rdp just different protocols to perform this task? [05:07] !vnc [05:07] VNC is a protocol for remote desktop. https://help.ubuntu.com/community/VNCOverSSH describes how to use it securely. It works best over fast connections, otherwise look at !FreeNX [05:07] !rdp [05:07] VNC/RFB is a lowest-common-denominator raster damage protocol. [05:08] RDP is a little better, it uses win32 primitives. It's Microsoft's clone of Citrix's architecture. [05:08] X11 is the native drawing system for Unix/Linux/Ubuntu. [05:09] does RDP stand for Raster Damage Protocol? [05:09] X11 is reasonably wire-efficient, but in practice modern GUIs use it mostly for raster painting, meaning that it ends up being slower than VNC (which supports lossy compression) unless you use a compressor like NX. [05:09] RDP = Remote Desktop Protocol [05:09] oic [05:11] so FreeNX is porting the X11 drawing directly [05:11] NeWS is the best design given modern hardware, but unfortunately Sun charged hefty royalties, and when NeXT bought Apple they killed off its network transparency. [05:11] whereas VNC allows X11 to draw it on the host machine, and then ports the visible image to the guest machine? [05:11] Nowadays it's being reinvented by web weenies on top of HTML Forms :-/ [05:11] Ender: right. [05:11] i see. [05:11] Ender: strictly, NX is just compressing the existing X11 drawing operations [05:11] and for now, vnc is kinda the best option [05:12] on non-windows machines [05:12] In a heterogeneous environment, VNC is your only choice [05:12] right? [05:12] oh. [05:12] Because RDP servers on Linux suck, and X11 clients (called "X servers") on Windows sucl [05:12] *suck [05:13] Of course, a CLI will be orders of magnitude faster than even VNC [05:13] right [05:14] i'm comfortable using the cli [05:14] i just wanted to try it [05:14] ok, but to abandon that now [05:14] and instead: [05:15] I'm trying to set up a Samba NAS on a ZFS pool in a heterogenous network environment. I'll need virus scanning for incoming files from windows boxes, and a firewall to block all internet traffic to one of the machines. [05:15] how do you suggest i do that? [05:16] Ender: I have used clamscan package [05:17] Ender: ZFS support on Linux is legally iffy [05:17] yeah, i know. it really upsets me. [05:17] Ender: IMO you are better off using ext4 and migrating to btrfs when it becomes production-ready, or using opensolaris in interim [05:17] i was thinking about using a virutal BSD os to handle the nas [05:18] (Except Oracle killed OpenSolaris.) [05:18] ugh, another upsetting thing. [05:18] Just a reminder not to buy into in proprietary solutions :-) [05:19] seriously!!! [05:19] it's so hard sometimes though, you know [05:19] people use Access to manage their orders, Quickbooks to handle finances, and Outlook for email [05:20] that deer-in-headlights look at the mere mention of a different system, even a FREE system... [05:20] anyway... [05:20] what i want is a software raid [05:20] and my understanding is that zfs incorporates a whole lot of really interesting raid functions by default [05:21] BUT i'm open to other suggestions [05:22] is it possible to get data redundancy in a software raid using Ubuntu Server and ext4? [05:24] Ender: You could use mdadm and whatever raid level you want...and format ext4. [05:24] !mdadm [05:24] twb: eh? opensolaris is opensource, not proprietary, and there's already openindiana to continue from where opensolaris left [05:25] can i run a good antivirus and firewall on openindiana [05:25] i would say that no [05:25] Ender: How many viruses targetting openindiana are you aware of? [05:26] he's doing fileserver with samba (for windows) [05:26] yeah. [05:26] so i need oi to scan for windows viruses [05:26] lol [05:26] i just find that funny [05:26] Ender: freebsd has antivirus, firewall and zfs [05:26] Ender: I use clamscan [05:26] for anti-virus on ubuntu and can be used on any unix os. [05:28] this needs to be a pretty hands-off system that auto-updates virus definitions [05:28] does clamscan have a function like that? [05:28] yes clamscand auto updates virus def [05:29] the package on ubuntu is actually called clamav [05:29] so what's the advantage of using ubuntu server instead of openindiana [05:29] You get suppport in #ubuntu-server :) [05:30] loll [05:30] yeah seriously [05:30] that's a huge factor [05:30] afk 15 [05:30] Ender: This is a support channel for Ubuntu Server, not an advocacy channel. Debate about "which OS is better" does not belong here. [05:35] jmarsden, i didn't mean to violate any rules, i'm just investigating the best way to accomplish my goal. in fact i will stick with ubuntu server and seek support form this channel. [05:36] That's fine... I just had visions of the channel deteriorating into why one OS is "better"... didn't happen, so all is well :) [06:01] ha! nip it in the bud eh [06:01] well done [06:01] ok [06:01] so...i have ubuntu server installed on a p4 system i rescued [06:01] how do i make a samba share [06:02] well first, a software raid [06:02] do i need to instal Server on the software raid? [06:03] Ender: https://help.ubuntu.com/community/Installation/SoftwareRAID [06:03] Is there some compelling reason why I shouldn't use the same public/private SSH keypair on all my computers? That is, the keypair I use to SSH into remote hosts. Or should I generate keys for each machine I use? [06:03] It's easier to do software raid during the install, but if you understand what is going on I am pretty sure you can do it afterwards instead. [06:03] hrm. i don't really understand what is going on. [06:03] so how do i learn? [06:03] just google software raid? [06:03] Ender: I would do it through the gui install...pretty easy to do through that link I sent you. [06:03] tohuw: Keep the private key on ONE host, and keep the public key on all the servers, that's usually fine. [06:05] jmarsden: sorry, i should be more clear. I have three computers I use. I remote into two servers. On this computer, I have a pub/priv keypair so I can ssh into both servers using that. Can I just copy this keypair to the other two client machines I use? [06:05] ok i'll read that website, thanks! [06:07] tohuw: You can put the same private key (you don't need the *pair* on any one computer) on multiple client machines, as long as all are equally secured against theft etc and you have a good passphrase. [06:07] ok [06:08] thats what i thought, thank you [06:08] tohuw: Many people do not put their private key on a laptop/notebook, for example, because there's a higher risk of it being stolen or lost. [06:10] i remote into my servers frequently from my laptops. that being said, they're also relatively secure, and i would revoke my key from everywhere its authorized if I thought it was compromised. such are the risks one takes [06:10] live a little, i say [06:15] do i need the samba packages to be able to READ samba shared or JUST to host them? [06:16] the samba package itself is the server. [06:16] Look at smbfs and smbclient for the client side of things, if I am remembering that right. [06:16] so if i've made a samba share on my server then a default install of ubuntu desktop on another computer in my network should be able to read the samba share with no additional packages? [06:18] smbclient is required on the client. [06:20] new question If I am going to set up a SAMBA server, do I want to set up the mount points off my home directory? like /home/samba [06:20] glenp: its up to you...I usually put them in /mnt/ you can really define them any where in the config file. [06:21] glenp: You can share any part of the Linux filesystem as a samba share... there is a [homes] section in smb.conf that makes it easy to share users home directories consistently if you have a multiuser machine [06:23] please confirm if the default run level for ubuntu is level 2 [06:23] draven_sol: I believe it is 3 [06:24] ok I have a 1tb and a 320gb 320gb is the os and all the stuff. The 1tb was going to be the /home directory and also set that up for samba and backup. Im trying to figure out how to divide the 1tb drive. the only thing that I have not setup is the /home partition [06:24] thesheff17, do you know a command i may run to test what level i'm currently in? [06:26] draven_sol: I see runlevel [06:27] thesheff17, thanks that's what i needed [06:31] glenp: depends if you are having user shares or just a couple shares....you could easily just do /home/samba if you are going to have 1 share. [06:34] I was going to make a share point so that people on the network could do like /home/samba/dir1 thru dir10 [06:58] I am looking to build a Fileserver using 8 2TB Drives and RAID6. Id like to use Ubuntus 10.04 mdadm. Am i better off to buy a good Raid Adapter or is mdadm ok? In addition, id be grateful for any Motherboard recommendations. [07:00] overrider: do you have several sata-bus? [07:00] overrider: it's ok if you put drives to several sata-bus, instead of one or two [07:02] gobbe: im looking to purchase a motherboard with 8 sata ports - or a good raid adapter. [07:03] overrider: raid6 is ok with mdadm, i wouldnt go with raid5 and large sata disks because of huge rebuild-time [07:03] overrider: but your plan sounds reasonable [07:04] gobbe: the question is whether mdadm and possibly lvm etc is more stable or a good raid adapter. [07:05] overrider: i would go with mdadm, separate raid-card is singlepoint of failure === airtonix_ is now known as airtonix === bgupta_ is now known as bgupta [07:27] does mdadm reduce performance === twister004_ is now known as twister004 [07:34] Ender: Compared to what? On most modern CPUs it is no noticeable, but if your CPUs are super busy number crunching you might notice a slight difference. Most 'normal' servers are I/O bound anyway, and so have spare CPU to use for mdadm. [07:35] ok and what about on desktop hardware [07:35] Ender: Also note that RAID1 needs less CPU than RAID5, if you want to minimize how much CPU mdadm uses. [07:35] old desktop hardware [07:35] like p4-era [07:36] P4 isn't all that old :) I ran software RAID on PIII servers a decade or more ago... as long as you are not expecting that desktop to also be doing something super-CPU-intensive my guess is that you are unlikely to see the difference. [07:36] I used a 400 MHz Pentium II for a home file server with two disks in software raid 1 and four disks in software raid 5 until last year. Worked perfectly. [07:37] wow sweet [07:37] ok that's very reassuring [07:37] i have four 80gb western digital caviars sitting here [07:38] lemme go see if ican plug 'em all in.... [07:38] RAID 10 here we come? :) [07:41] well damn [07:41] this is a dell minitower [07:41] no space for even a 2nd harddrive [07:41] [07:41] I'm happy with Raid Mirrors for my backups :) [07:42] Diff batches of disks obviously. [07:47] what's the best way to increase the life of my storage drives [07:47] i mean, the best way to set up a nas [07:47] like, should i run the nas on the same harddrive/raid array as my server's OS [07:47] or is it substantially better to run the OS from a separate physical drive [07:48] ? [07:57] i'd start by learning how to use lvms [08:00] airtonix, any particular suggestion on how to do that? === _TechAway_ is now known as _Techie_ [08:01] http://www.linuxconfig.org/Linux_lvm_-_Logical_Volume_Manager [08:01] http://en.wikipedia.org/wiki/Logical_Volume_Manager_(Linux) [08:01] https://wiki.edubuntu.org/Lvm [08:01] Ahem... maybe start with the Server Guide: https://help.ubuntu.com/10.04/serverguide/C/advanced-installation.html [08:01] Seeing as we are in #ubuntu-server :) [08:01] https://help.ubuntu.com/community/Installation/RAID1%2BLVM [08:01] ok point taken [08:02] i can use google just fine, thanks [08:02] i typically use google [08:02] but that last one is what you want [08:04] hi. anyone can show me the way of ssl. been struggling on this issue for a couple of week, though. ubuntu server 10.10, though. [08:04] the files: /etc/apache2/sites-available/abv.conf is at http://paste.ubuntu.com/564276/ and /etc/apache2/ports.conf is at http://paste.ubuntu.com/564274/ [08:05] ok good people of ubuntu-server, i am off to bed [08:05] goodnight! [08:05] and good hacking (: [08:05] nite Ender [08:06] zairo: Did you read the Server Guide about SSL in Apache2? https://help.ubuntu.com/10.04/serverguide/C/httpd.html [08:06] If you followed that, how far did you get and what happened? [08:07] jmarsden: let me check it out first. before this my ref are http://www.tc.umn.edu/~brams006/selfsign.html http://www.tc.umn.edu/~brams006/selfsign_ubuntu.html [08:08] zairo: It is usually safer to use official docs, not random stuff from elsewhere :) [08:09] jmarsden: gulp. i guess i've slapped from far, hu2 [08:12] :) The link to the Server Guide is in the /topic for this channel for very good reasons... it really does have quite a bit of useful info in it, and if that info is bad, we can file bugs and provide patches and get it corrected :) We can't do any of that for other web pages "out there somewhere"... [08:18] ok i read this also: https://help.ubuntu.com/10.04/serverguide/C/certificates-and-security.html [08:20] OK. So... try doing it and if you get an error, come back here and ask for help :) [08:22] i should :) my due for this is end of this week. [08:24] For just one web site, it should be only minutes of work, not days :) [08:31] jmarsden: i am a programmer cum network, though. quite new on server but been 2 years on Ubuntu desktop. [08:31] * jmarsden has been using Linux since 1992 :) [08:33] jmarsden: your phone also Linux, don't u? mine Nokia N900 [08:34] Android G1, but I need it to stay working, so I have not hacked it :) But, that's not really on topic in #ubuntu-server [08:36] ok i am at step 6 of https://help.ubuntu.com/10.04/serverguide/C/certificates-and-security.html. === twister004_ is now known as twister004 [08:40] zairo: so you have made a self-signed ca-cert and need a servercert signed by your ca? [08:41] xampart: yep already servercert, i think. [08:42] step 7 quite confusing [08:43] ok i create new file name mydomain and put it at/etc/ssl/certs/ [08:48] zairo: A new file? Or a copy of the correct part of /etc/ssl/newcerts/01.pem ? Also, the filename of the certificate is usually mydomain.crt not just mydomain [08:49] jmarsden i put extension .pem though [08:56] New bug: #715056 in dovecot (main) "invalid ssl-certificates in /etc/postfix/main.cf after security upgrade" [Undecided,New] https://launchpad.net/bugs/715056 [08:58] <\sh> I wonder why people still using RHEL or CentOS...the default install of CentOS (without tweaking manually) is a mess and gives me a headache. Why on earth someone needs Squid on a default install? *bhod* [08:59] \sh: At least it has SELinux by default. [09:01] <\sh> jpds: the only positive...but I wonder how many people are trained in writing selinux policies (is there a RHCE training for selinux? dunno) [09:01] jmarsden: ok finish step8. i restart server and got error :Action 'start' failed. [09:01] zairo: Usually this means you made a mistake in editing a config file somewhere. Read the logs under /var/log/apache to see what the problem is. [09:01] Um.... /var/log/apache2/ [09:02] Unable to read server certificate from file /etc/ssl/private/cakey.pem [09:02] OK, so check file permissions on that file carefully. [09:02] wrong cert file. what file i should use? [09:03] ok. should be 600 rite? [09:04] now Init: Private key not found [09:05] zairo: If apache2 needs to read it, it may need to be 0640 and group www-data [09:05] thanks. will do that now. [09:08] <_ruben> hmm .. been getting more and more kernel panics on my router vms .. not good .. about time to do some investigations [09:08] \sh: sure, there is training for selinux on rhce and afaik squid would never be on by default, depends which package groups you are installing.... bluetooth by default for @base in RHEL5 sucked though [09:10] Init: Unable to read server certificate from file /etc/ssl/private/cakey.pem [09:10] zairo: Is there really a valid server certificate in that file? [09:11] i am not sure which is the correct file [09:12] If you followed the instructions, I think cakey.pem is the root CA key, not a server certificate... [09:13] <\sh> TREllis: I just installed the server layout from centos without tweaking the package list manually...and squid is installed by default..which looks really awkward [09:13] zairo: What makes you think cakey.pem should be a server cert? [09:13] It's name says it is a CA key, and the instructions show it being a CA key... so why are you telling Apache it is a server certificate? I'm confused... [09:14] i follow all the steps and named. is there other name for key file in that tutorial? [09:14] <\sh> TREllis: and yes..bluetooth should be avoided [09:15] sorry i am not really understand what i am doing. [09:15] i just follow the step. [09:16] if you "followed the steps" then your server cert should be a file called something like /etc/ssl/certs/mydomain.crt [09:16] So did you tell Apache that is where the server certificate is? [09:17] that is in which steps? [09:18] Incidentally, "not knowing what you are doing" is fairly bad, this is security related setup you are doing here... you need to know what you are doing, to keep your server secured. [09:19] i do. will try to understand after i manage to make it. [09:19] Back on the https configuration page, I see "The directories /etc/ssl/certs and /etc/ssl/private are the default locations. If you install the certificate and key in another directory make sure to change SSLCertificateFile and SSLCertificateKeyFile appropriately"... what did you do regarding those paths === FkCek is now known as FkCek|a [09:19] i leave it as default folder [09:20] Then why is Apache trying to read a *server* certificate from /etc/ssl/private/cakey.pem ? [09:21] pastebin your ssl site config file so I can read it, please. [09:22] ok wait a minute [09:24] zairo: OK, but it is 01:24AM here, I do need to go to bed sometime soon :) [09:24] http://paste.ubuntu.com/564318/ [09:25] Um. So.. line 5. Why did you tell Apache that there is a server certificate in the cakey.pem file???? [09:25] ic. may be u can show the way. i think i use the wrong file or something. [09:25] Look at the example in line 10... [09:25] ic. where is the correct file? [09:26] It uses a .crt file. You made one of those, right? Hopefully /etc/ssl/certs/mydomain.crt [09:26] line 10 is what i am doing before this [09:27] Where did you put the server certificate that you made? [09:27] In step 8. [09:27] That is where it is. [09:27] It is where you put it :) [09:27] in /etc/ssl/certs/ [09:28] OK, so then the SSLCertificateFile line needs to say that :) [09:28] that's for SSLCertificateKeyFile /etc/ssl/certs/api_undies_my.pem, rite? [09:29] I don't understand the question. [09:30] the error on certificateFile or CertificateKeyFile? [09:31] Line 5 is definitely wrong. I don't know about Line 6. [09:31] So change line 5 and retest. [09:31] Hi. I need to create a couple of dummy interfaces which I can connect to a bridge (br0). Need to test mac number assignments against a switch. Any recommendations? It's like a bunch of virtual machines having there own interface. [09:31] tun? [09:33] jmarsden: sorry for your time. may be i can solve it on my own. [09:33] thanks a lot bro :) [09:33] Blinkiz: It is late and I should sleep, but... do tun devices have MAC addresses?? If you need a MAC address, I'm not sure tun will get you one. [09:33] zairo: OK... goodnight :) [09:33] ok. nite n nice to meet u [09:34] jmarsden, tap? :) [09:35] aaa, tap seems nice [09:35] Blinkiz: Maybe... sounds plausible, at least :) [09:35] I really need to sleep, sorry :) [09:36] jmarsden, Sleep... It's 10am here :) [09:36] 01:36AM here :) [09:36] Am sitting here a work. :) [09:38] jmarsden: will continue tomorrow :) [09:39] adios [09:39] <_ruben> ok, this is getting crazy .. got like 4 kernel panics in an hour === DaBeast_ is now known as DaBeast === _Techie_ is now known as _TechAway_ [09:50] * dwatkins wonders if there's a lost manpage for parted which has a --show-exact-bytes option documented [09:50] dwatkins: Easier to read the source code than find a possibly nonexistent man page, I would suggest :) [09:51] jmarsden: heh, I guess I may as well use sfdisk, then ;) [09:52] I'm trying to document how a root mirror is defined, so I imagine I need to specify the exact partition size for the possibility one disk gets replaced and needs repartitioning and adding back into the mirror [09:57] dwatkins: Oh... well... doesn't the unit B command in parted do that? It's documented in the man page I have... is it missing from yours? [09:58] ooh, I was expecting a switch... [09:58] Always read the man page :) [09:58] hehe [09:59] also, read it in full, don't just skim looking for what you expect, as I've demonstrated :) [09:59] Indeed :) OK, now I really really need to go to bed... goodnight :) [09:59] ciao jmarsden - and thanks [09:59] You're welcome [10:12] hmmm [10:13] i'm trying to do this tutorial http://www.wtorrent-project.org/trac/wiki/DebianInstall [10:14] i'm at the point where i putty into the server, and apt-get the list of packages [10:14] but it says here it cant find some packages [10:16] http://pastebin.com/0pN8K6C2 [10:18] how can i fix this? (note, this is one of my first linux expierences :D ) [10:18] Sounds like it's an old tutorial, DaBeast [10:19] yup [10:19] are there any replacement packages? [10:19] there was more, but i fixed some (i think) [10:20] It appears g++ is now at version 4.4 [10:22] 4.5.1-7ubuntu2 is the current version of libstdc++ [10:22] the ++ makes searching with apt-cache difficult [10:23] yeah, i tryd that, and it came up with lots of irrelevant apps like VLC and codecs lol [10:23] That sounds like a regexp problem, what did you try to do to install it? [10:23] also, it may already be installed, check 'dpkg -l' output with grep [10:27] gotta check my putty logs, its out of bounds hehe [10:28] hello [10:29] i want to host the reccomended packages for ubuntu server locally, saving bandwidth in a corporate environment. can anyone describe how to do this [10:29] also to allow multiple installs without having to create more proxy exceptions and so on [10:32] DaBeast: I suggest using GNU Screen to be able to scroll back (increase the default scrollback to a about 10,000) [10:32] S1am: I believe you can use apt-mirror for this. [10:32] ok, will install that, thanks! [10:33] i'm kinda new to this, hehe [10:33] DaBeast: screen is excellent for situtations where you might want to disconnect but leave processes running on a remote system, but takes a little getting used-to [10:33] It's worth the learning-curve, imho [10:33] how would you recommend hosting this? [10:34] i see, is screen also a SSH client? [10:34] S1am: I've never used it, so I can't really give any reccomendations - what do you mean by hosting, though? [10:34] DaBeast: no, it provides the equivalent of a tabbed command line client but all with text [10:35] well, should on it be a server all of its own with all the repositories as well, or could it be hosted on a server with a bunch of other apps [10:35] nice [10:35] can the server part be by itself and then all of the storage elsewhere [10:35] *should it be on [10:37] S1am: I suspect it's the same as hosting any other kind of file repository, it depends on your requirements in terms of bandwidth. Serving out files tends to be mainly bound by the network, not the CPU, so you should be able to run other services, depends on the overall load of each. [10:38] Also, if you have multiple concurrent connections, the speed of disk access may be an issue, but I suspect network will still be your main bottleneck. [10:38] ok [10:39] S1am, there are a number of different ways to do what you're asking, apt-mirror is one of them [10:39] they each have advantages and disadvantages [10:39] apt-mirror downloads the packages right? [10:39] I use squid to proxy them myself [10:39] yes [10:40] S1am: Transparent Squid proxy. [10:40] at my work we have an inline bluecoat proxy [10:40] Hi. I need to test a dhcp server here by creating around 10 interfaces. I hope I can do this from only one computer. Meaning, 10 dummy interfaces with 10 unique MAC addresses [10:40] Any suggestions what am seeking for here? Some kind of bridge and tap interfaces maybe? [10:40] the best thing about a transparent squid proxy is that there is zero client configuration [10:41] Hmm no windows build, ah well, no biggie [10:41] bluecoat provides this as well [10:41] Blinkiz, nothing that complicated, configure interfaces as eth0:0, eth0:1, etc [10:42] qman__, Do they get different mac addresses this way? [10:43] not sure [10:43] qman__, another problem is that it does not seems like I can get a subinterface up without specifying a static IP address for it [10:44] how do you actually host the packages downloaded by apt-mirror [10:44] is there a way of installing a http server with the right configuration [10:45] and configuring the installer to download from that location [10:45] S1am, no special configuration is necessary, you just need the files in the right tree [10:45] the problem with that method though is that each client must be configured at install time to use your mirror [10:45] iyes [10:46] i'm ok with doing that though [10:46] i'm trying to make cookie-cutter installs [10:46] i work a at a mining company that has 2 mines and has just acquired a third [10:47] each mine has lots of PLCs and they are networked with switches, we want to be able to monitor everything using nagios installed on ubuntu-server [10:48] so i don't want to have to be downloading everything over and over- especially the remote sites [10:48] well, the web server itself doesn't need anything special configured, the files just need to be in the right tree and places, and I'm fairly sure apt-mirror does that part for you [10:48] you'd have to read the documentation for specifics, though [10:48] i don't understand how the installer knows how to actually download it though? [10:49] it's configured in apt [10:50] ok given that this is always going to be installed on a virtual server via an iso [10:50] can i just mount the iso and configure apt there [10:50] you'd have to configure apt and then rebuild the image [10:51] at least as far as I know [10:51] I haven't done it myself, just going by what I've read [10:51] what do you mean by rebuild [10:51] there should be documentation on all of this, though [10:51] as in set up the boot sector again [10:51] no, build a custom iso image [10:52] with your special configurations [10:52] ok [10:52] do you have any links to documentation on this [10:52] building custom iso images? [10:53] dwatkins: thanks, i got it to downloading files now [10:53] cool DaBeast [10:53] I don't have any off hand [10:54] look into building custom live CDs and USB installers [10:54] that should point you in the right direction [10:54] k [10:54] also, look for apt-mirror documentation [10:54] since I'm sure you're not the first person who has wanted to do this [10:54] yes [10:55] thanks for the help [10:55] disk imaging might be easier, too [10:55] yeh up to a point [10:55] since basically all you have to do is remove /etc/udev/rules.d/70-persistent-net.rules, set the host name, and configure the network to get a unique install [10:56] a short script would handle it [10:57] if you're using DHCP that last part goes away, too [10:59] do you think disk imaging, in general, is more likely to corrupt an install over just installing everything naturally [10:59] in the case of desktop pcs [10:59] for a school, for instance [10:59] corrupt installs only happen with hardware failure [11:00] or broken operations/kernel crashes [11:00] a lot of schools have aging computers with 5 year old hard drives [11:00] and in the case of failing hard drives, corruption is equally likely with either method [11:01] the question is which is easer on them, disk imaging or installing naturally [11:01] you're more likely to notice the failure with imaging though [11:01] If I were administering a school, I'd consider having the disks in the PCs replaced with either internal USB flash drives (cheap ones, glued in carefully) or have them boot off the network. [11:01] since the disk imaging utility will tell you [11:01] whereas a normal install may just have silent errors that manifest later [11:01] true [11:02] and yes, if you can afford one or two servers you can count on, network booting is a great option [11:02] the case i am working on is for virtual servers [11:02] vmware environment [11:03] for virtual servers I tend to just copy the disk images, or clone them as required [11:03] If I could authenticate via winbind, I'd run all my virtual servers on vmware [11:04] the servers need to be configurable [11:04] in terms of what software is installed to start with [11:05] you can create a base image with no software/whatever you have on every server [11:05] and then install on top of that [11:05] or you can create a few baseline images to use [11:05] ok so to install on top i would use apt-mirror? [11:05] after configuring the installer as discussed above [11:06] no, that's different [11:06] how so [11:06] if you want a base server image for virtualization, you just install from disc, configure it the way you want, then save a copy [11:06] the stock disc [11:06] yes [11:06] then, you copy that image for all the new servers, already installed [11:06] already configured to use your mirror [11:07] ah i see [11:07] you just have to change the hostname and configure the network for them [11:07] which you could automate with a script [11:09] <_ruben> what would be the easiest way to grab the output of a kernel panic/oops? they tend to scroll off screen [11:09] _ruben, use shift+pageup [11:10] <_ruben> qman__: isn't gonna do much good when the system is frozen [11:10] well, if it's locked solid, not much you can do [11:10] if it's reproducable, you can configure the console to output via the serial port [11:11] yeah, and log with your client application [11:11] well, you can do that in any case, but if it's a reproducable problem that's useful, I mean [11:12] or configure your framebuffer to use a higher resolution, so it fits more on screen [11:12] * dwatkins wondres if the magic SysRq keys might be useful here [11:13] Do you have any details on the panic, _ruben? [11:13] if the keyboard lights are still blinking, sysrq will probably still work [11:13] also, is anything written to /var/log/messages at the time? [11:14] and /var/crash [11:14] It appears boot.log isn't kept between reboots [11:14] <_ruben> it's somewhere in the (ipv6) network stack, and might be triggered by ospfv3 mucking with the fib [11:15] <_ruben> it's a vm (esxi), wonder if i can get it to use a highres framebuffer for the console [11:17] might as well ask again now that some people are here [11:17] my file server is having an odd problem that mkdir takes a very long time [11:17] for all users, local and otherwise, and in all locations, qman__? [11:17] I checked it with tune2fs, and I'm not anywhere near running out of inodes [11:17] yes [11:18] well, all locations on the raid at least [11:18] I can check the rest of the system [11:18] have you tried running strace with the relevant timestamps to see what's taking the longest time? [11:18] it's only mkdir, files are fine [11:18] I have not [11:18] Sounds like a metadata issue, but that would probably pinpoint it a little. [11:18] Start with strace -fro mkdir $SOMEDIR [11:18] no wait [11:19] directories not on the raid are quick [11:19] Start with strace -fro mkdir.trace mkdir $SOMEDIR [11:19] sorry [11:19] so it's just that filesystem [11:19] ok [11:19] also, what kind of filesystem and how is it mounted [11:20] when I do it with strace it goes quick [11:20] ext3, mounted on /home [11:20] only option is relatime [11:21] that's really wierd that it would work better when you're running with strace, that suggests it's a buffering issue, perhaps [11:21] like it's waiting for the queue to be full before committing the mkdir [11:21] I have some stuff running in the background but it's half what my GUI desktop is doing [11:21] it shouldn't be having any significant performance impact, the load is under .1 [11:22] iowait is normal [11:22] I'd suggest looking at sar data in that case. [11:22] get stats every 5 seconds for a couple minutes and have a good look at all of them. [11:22] ok [11:26] thanks for the tips, at least I have something to look for now [11:27] you're most welcome, qman__ - I'd be very interested to know the cause when you find it. [12:26] whats the hard drive equivilant of ifconfig? (i want to view all available drives) [12:27] ls /dev/sd? [12:28] probably fdisk -l === airtonix is now known as jadakren [12:54] with fdisk it says invalid option '1' [12:56] "-l" [12:57] still fails :p [12:58] how about "sudo fdisk -l" [12:59] invalid option -- '1' :p [13:00] not 1, but l as in list [13:01] New bug: #715141 in ntp (main) "ntp.ubuntu.com does not have AAAA record" [Undecided,New] https://launchpad.net/bugs/715141 [13:02] haha lol [13:02] hard to see the difference xD [13:02] cut+paste, eh? [13:03] and -h -option usualle helps [13:03] use a proper font in your IRC client :) [13:03] :p [13:03] makes sence :p [13:10] can linux use fat32/ntfs partitions? [13:11] i want to hook up my external harddrive to it [13:29] when I sudo, my server keeps saying that it can't resolve the name of the server, dispite it's identified in /etc/hosts [13:29] m_tadeu_: Is it the same in /etc/hostname ? [13:30] Pici: sames name yes [13:32] in /etc/hosts is identified like this "127.0.1.1 myserver" right after the localhost...should this also be a 127.0.0.1? [13:34] morning === oubiwann is now known as oubiwann_ [13:44] still unable to resolve [13:45] what else can I do about it? [13:46] New bug: #715152 in ntp (main) "ntp wont run on IPv6 only host unless lo has 127.0.0.1 address" [Undecided,New] https://launchpad.net/bugs/715152 [13:52] Morning all [14:00] hi [14:00] m_tadeu_: sorry, got stuck doing something else. It can't hurt if you put in another line that has 127.0.0.1 like the 127.0.1.1 [14:03] Pici: it's ok...well I checked and the 127.0.1.1 is to resolve the name...apparentely sudo needs this...problem is it's there and set properly and it still complains about it [14:04] m_tadeu_: Does `hostname` report the same hostname that you have set in /etc/hostname ? [14:05] hallyn: howdy!! libvirt 0.8.7 will be in natty right? [14:05] Pici: in deed it does [14:06] m_tadeu_: curious... [14:11] Pici: I'm out of my environment here....not much of a sys admin :) and I run out of solutions pretty fast. but this is weird...can bind have anything to do with this? [14:13] m_tadeu_: It could, if your server couldn't resolve its own hostname for some reason due to a misconfiguration. [14:13] m_tadeu_: does host $(hostname) work? [14:14] Pici: "not found: 3(NXDOMAIN) [14:27] m_tadeu_: Theres your issue. You can't resolve your own hostname. [14:31] Pici: will uninstalling bind solve the problem? [14:33] anybody offhand know the command to stop/start freeradius in ubuntu? [14:34] sudo /etc/init.d/freeradius restart? [14:34] resno, why are you such a genius [14:34] stop being so smart [14:35] heh [14:35] * resno bows [14:35] I was trying radiusd, because radiusd is what most other distros use. [14:35] duur [14:35] I feel like each freeradius guide I read, I have to adapt what I'm reading to ubuntu-based commands for it. [14:35] Roasted_: am i wrong? i was taking a guess [14:35] no, you're right. [14:35] I'm just explaining why I goofed. [14:35] ah lol [14:36] it seems debian land loves to be different, since debian is the same way. [14:36] but centos/etc has the "official" guide. pfft. [14:36] hey resno here's a pop quiz question for you. [14:36] New bug: #461441 in exim4 (main) "package exim4-daemon-light 4.69-9ubuntu1 failed to install/upgrade: subprocess post-installation script returned error exit status 1" [Undecided,Won't fix] https://launchpad.net/bugs/461441 [14:36] my ubuntu install in VMWare loses external internet connection frequently. upon bootup as well as at random times. the fix? disable/renable network manager. [14:36] why. :) [14:38] Access denied [14:38] . lolz. [14:38] ive actually read that [14:38] because it's a server I'm probably going to take out NM and static the IP in the network file. [14:38] Roasted_: you got vmware running? what version of ubuntu? [14:38] but I don't want to just accept that and not "fix" it or understand why [14:38] 10.04.1 32 bit [14:39] it's on a static IP too. [14:39] it did it on DHCP though as well. [14:39] i use virtualbox now, i want to go to vmware [14:39] m_tadeu_: I'm not familiar enough with bind configuration to say either way. [14:39] hey Pici [14:39] I'm trying to SSH into the box now and it's saying access denied. [14:40] yet I totally know my password is like, right, or something. [14:40] you are denied. [14:40] :( [14:40] resno, what did you read about that issue? [14:40] have any issues getting vmware running? [14:40] I didn't get vmware running. [14:40] the network admin did. [14:41] lol oh [14:41] I'm just doing this project since it's linux based so it's kind of up my alley [14:41] trying to halp out, etc. [14:41] ive just seen network manager have various issues when setting things up etc [14:41] yeah. it's been behaving itself nicely since 10.04 tho [14:41] ah ok [14:42] I guess I need to brush up on how to set up static interface files, I haven't done one in a while. [14:46] Roasted_: do you consider yourself a sys admin? [14:46] no. I do desktop support, but they've been getting me more into server stuff. [14:47] but I'm also a linux fanboy, so when a linux project came up it got tossed to me to do. [14:47] I've worked with servers a lot, but never on a grand scale of an enterprise environment. [14:47] i think im going for a careear change, but i am not sure how best to get into it. [14:50] resno, what in particular? [14:50] leaving developement and get into sys admining [14:50] side question - if I'm setting a static IP without network manager, do I set the DNS entry in the resolv config or do I add it in network interfaces? [14:50] well, not leaving but just less focus [14:51] resno, well, anything in IT is a royal headache. forwarning you. [14:51] that depends [14:51] there's days I leave wondering why I come in here. But Is till enjoy it. [14:51] if you have resolvconf installed, in network/interfaces [14:51] if not, in resolv.conf [14:51] resolv.conf is in /etc/network right? [14:51] no [14:51] /etc/resolv.conf [14:51] Roasted_: true, there are issues and headaches everywhere. when it works its wonderful. [14:51] yeah I have /etc/resolv.conf [14:52] resno, yep. gotta love it. :P [14:52] you should always [14:52] but do you have /etc/resolvconf [14:52] New bug: #581997 in exim4 (main) "package exim4-daemon-heavy 4.69-5ubuntu2 failed to install/upgrade: subprocess pre-removal script returned error exit status 3" [Undecided,Won't fix] https://launchpad.net/bugs/581997 [14:52] within /etc/resolv I have an update-libc.d folder. [14:52] but thats it [14:52] er [14:52] yes, i do [14:52] sorry. ID-10-T moment [14:53] has an entry for generated by networkmanager. [14:53] back it up and edit those entries accordingly? [14:53] na [14:53] safe method would be [14:53] add dns entrys in /etc/network/interfaces [14:53] and also edit /etc/resolv.conf [14:53] is it nameserver or dns-nameserver? [14:53] for when I add it to net interface? [14:53] RoAkSoAx: yes [14:53] resolvconf will wipe out /etc/resolv.conf if it wants do, but the network/interfaces one would be used then [14:54] hm, so what's optimal here. Do I let resolvconf alone and just add DNS to net/interface? or do I add the DNS to both? [14:55] leave resolvconf alone [14:55] if resolv.conf doesn't say, autogenerated by xxxxx, edit it [14:56] It says #Generated by NetworkManager [14:56] hallyn: someone filed a bug requesting the update, so I just assigned it to you :) [14:57] you have a gui installed on that box? [14:57] hallyn: tgt is horribly outdated in natty you want me to do an update? [14:58] yes, it's a desktop installation virtually on vmware, but it will be acting as a light duty radius authentication server. [14:58] I'm remoted in to it now [14:58] network manager is acting up, so I want to disable it and use static via interface files. [15:00] patdk-wk, by chance do you know the proper dns entry to put in net/interface? I've seen nameserver, nameservers, and dns-nameserver [15:00] man interfaces [15:02] already been there. [15:02] I saw everything *but* nameserver. [15:02] MTU size, TTL time, etc. [15:02] odd [15:02] gateway, pointtopoint, metric, bcast, netmask, etc. [15:03] just not... dns/nameserver :P [15:03] Pici: it worked....I unsintalled bind and setup the dns on a router [15:04] speaking of which, if I'm not using network manager, will /etc/init.d/networking stop/start/restart still work or does that only work to NM? [15:04] dns-nameserver x.x.x.x x.x.x.x [15:04] m_tadeu: cool beans, sorry I couldn;t be of more help. [15:04] thanks bro [15:04] Roasted_: it works for unmanaged interfaces [15:05] I just read different guides and they all said soemthing different. [15:05] Pici: it's cool...you brought me the light to get there :) [15:06] will it be sufficient to just remove NM from startup apps? I hate to uninstall it just yet. [15:06] no, it will remove itself [15:06] now for the next problem...I have asterisk installed, but when I reboot, the service doesn't start [15:06] New bug: #488615 in exim4 (main) "package exim4-config 4.69-5ubuntu2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Incomplete] https://launchpad.net/bugs/488615 [15:06] New bug: #526389 in exim4 (main) "package exim4-config 4.69-11ubuntu4 failed to install/upgrade: useradd: cannot lock /etc/passwd; try again later." [Undecided,Incomplete] https://launchpad.net/bugs/526389 [15:06] New bug: #691162 in exim4 (main) "package exim4-config 4.71-3ubuntu1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Incomplete] https://launchpad.net/bugs/691162 [15:08] removing it from startup apps won't uninstall itself though... [15:12] connect: Netowrk is unreachable. [15:12] :/ [15:20] where can I check if a server is being run on boot or not? [15:22] service? [15:22] patdk-wk: service yes :) typo [15:25] still cant get external access. [15:25] nice. [15:25] nice nice nice. [15:25] heh? [15:25] m_tadeu, I haven't found a good way [15:26] generally everything in /etc/init and /etc/init.d is run, and some of them disable themselfs depending on settings in /etc/default [15:26] oh, nice. somehow interfaces/network is reverting back to normal when I reboot. I lose my entire eth0 entry. [15:26] why would I lose my entire entry. [15:26] New bug: #715209 in rabbitmq-server (main) "Please sync rabbitmq-server 2.3.1-1 from Debian unstable" [Undecided,New] https://launchpad.net/bugs/715209 [15:27] New bug: #378307 in exim4 (main) "local_scan modification and new package" [Undecided,Won't fix] https://launchpad.net/bugs/378307 [15:30] please confir me...ubuntu only runs on run level 2? [15:31] there are no *levels* [15:31] the levels are just there to *fool* old startup scripts [15:33] patdk-wk: well it looks like they're still being used...and mine is nunning on L2...and the rcX.d dirs have all the links to the services [15:36] like I said, compatability mode [15:39] unknown host: www.cnet.com [15:40] patdk-wk: got it... [15:40] except, I have an ip, and subnet, and everything else. [15:40] Hi all. I have a 10.04 server that I'd like to enter a bug report for. Unfortunately, I backported packages from 10.10 to correct the problem, and I can't run ubuntu-bug to actually report the bug against the correct version. [15:40] it just wiped my config. [15:41] I was in the middle of reading over my network interface config and it came up and said something has changed, do you want to reload your file?> I hit reload and the entire eth0 entry disappeared. [15:41] what in the world is causing this. [15:41] who is running the meeting today? [15:43] zul: me [15:43] JamesPage: k i just added an agenda item [15:43] zul: nice [15:45] Anyway, the qemu version in 10.04 has a pretty bad bug with virtio disks > 2TB in size. Massive file system corruption. [15:51] is there a tool to control the services, if they start or not? [16:00] why does my network interface file hate me. [16:00] m_tadeu: Not yet (last time I looked). It's on the todo-list. [16:05] alvin: thx...not good...somehow my asterisk server is not running at startup and I can't figure out why [16:07] how can i cd to /dev/sdb1 ? [16:07] since cd /dev/sdb1 doesnt work [16:07] DaBeast: mount /dev/sdb1 somewhere [16:07] I.E. [16:08] sudo mkdir /media/mountdir; ; mount -t /dev/sdb1 /media/mountdir [16:08] Would it make sense that having DHCP still enabled on a machine with a static IP could result in the /etc/network/interface file changing at completely random times back to default? [16:08] that's one way at least [16:08] er [16:08] sudo mount* [16:09] m_tadeu: Since Ubuntu switched to upstart (without the 'legacy mode'), a lot of stuff doesn't autostart. I'm having the same struggles, but with other software. Proper boot logging would be nice too. [16:09] hmmm, let me try that, thanks :p [16:10] so its sudo mount /media/mountdir; ; mount -t /dev/sdb1 /media/mountdir ? [16:10] m_tadeu: Asterix is probably missing an upstart-compatible boot script [16:10] DaBeast: well, that's the code, but you need to put a few things in there [16:10] why does it do btw, the mounting [16:10] DaBeast: (a) you need to decide what to name the mountdir [16:11] DaBeast: (b) you need to know the partition type, and i have an extra ; in there [16:11] alvin: is upstart that thing that you start a service with "service xxxx start"? [16:11] its ntfs, for windows compatibillity [16:11] its an external usb drive [16:12] How do I disable DHCP without uninstalling it? If I uninstall DHCP it'll take out ebox, whcih is something I NEED to stay installed. I just want to disable DHCP on eth0. [16:12] DaBeast: then try this: sudo mkdir /media/; sudo mount -t ntfs /dev/sdb1 /media/ [16:12] do i need to mount it every time the server reboots? [16:12] m_tadeu: Yes. (I think the 'service' part might be deprecated) [16:12] DaBeast: generally, yes. but if its a server, how often are you going to be rebooting it? [16:13] whats the ? it seems odd to me, since its my first linux experience :D [16:13] probably allot [16:13] DaBeast: you put whatever directory name you want there :P [16:13] i.e. [16:13] since it wont be online 24/7 [16:14] if i wanted to name the directory i'm mounting the drive to "evil", i'd do this command then: [16:14] alvin: "service xxx start" works fine..."start xxx" says it doensn't know the job [16:14] its an old laptop, i'm wanting to convert it to a seedbox :D [16:14] sudo mkdir /media/evil; sudo mount -t ntfs /dev/sdb1 /media/evil [16:14] Isn't it 'xxx start'? [16:14] DaBeast: lol, i dont recommend running a laptop as a server [16:14] at ANY time [16:14] m_tadeu, sounds like it was moved to upstart [16:14] why not? [16:14] DaBeast: they arent built for it [16:15] consider. [16:15] true, it has ben on for a couple of days before tough [16:15] i've got this laptop here that runs standard Ubuntu Desktop edition. [16:15] worked just fine [16:15] it overheats every other day if I run it without shutdown [16:15] now consider my desktop box. [16:15] custom built [16:15] m_tadeu: There is a /var/log/boot (don't use - breaks your system) and a /var/log/boot.log (but information in there is not incorrect, but misleading) [16:15] running as a server [16:16] m_tadeu: But if your daemon is in there, upstart at least attempts to start it [16:16] hmm [16:16] overheats every 3 weeks that i dont keep tabs on the thing [16:16] also [16:16] more power in the desktop than the laptop [16:16] and when I say desktop, substitute server in for the word [16:16] well, i'm wanting to underclock the laptop so it gets less hot [16:16] still dont recommend a laptop as a server platform [16:16] its only for downloading some torrents and nzb's [16:17] OH GOD... torrents... o.o [16:17] * EvilPhoenix walks out upon hearing the ungodly word that was just spoken [16:17] yes, i'm a sinner xD [16:18] also, torrents arent nessesairly illegal, you know :p [16:18] m_tadeu: Does asterisk needs network connectivity in order to work? [16:18] m_tadeu: (to start, that is) [16:21] alvin: it does...because it does some checking for name servers and other hosts, etc [16:22] m_tadeu: That' bad news. That is in essence my problem too. Someone is working on improving this aspect. A service, or part of it that relies on the network being there will often fail. (NFS, libvirt,... and apparently also asterisk) [16:26] I hear systemd has the same troubles with NFS... Booting isn't funny anymore nowadays. (But fast as ligtning!) [16:26] alvin: well should that try be logged in /var/log/boot.log? and the asterisk log should have something too, I guess [16:27] m_tadeu: I think it should, yes. [16:27] alvin: well it's not...that's why I think it's not even trying to run it [16:27] m_tadeu: My remark about the log nog being reliable is because the log might be full or errors/warnings about something not working, while it sometimes does. Like mounting remote drives. [16:28] alvin: I see [16:28] m_tadeu: And after boot you can start it manually, without errors? [16:28] EvilPhoenix now hates me xD [16:28] huh? [16:28] ohi [16:28] sorry, i was actually trying to find an excuse to get coffee [16:28] :P [16:28] anyways... [16:28] lol [16:29] so where was I [16:29] oh right [16:29] anyways, i think my laptop would be good starters hardware [16:29] *shrugs* [16:29] do what you want, I gave you the command structure for the mount command [16:29] yes, thank you, kind sir :D [16:30] once you make the directory you're mounting to, though, you wont need to recreate it each time [16:30] but you WILL need to mount it each time [16:30] you restart [16:30] cant i put it in some kind of autoexec? [16:30] dabeast, sure, install dos [16:30] maybe fstab... but idk [16:30] patdk-wk: lol [16:30] rofl patdk-wk [16:31] hasn't the whole auto-x thing cause most all security issues for microsoft? [16:31] autorun usb, cdrom, floppy, emailattachments, ... [16:31] patdk-wk: indeed, but discuss that in ##windows or #ubuntu-offtopic [16:31] kay? [16:32] na, it's related to this [16:32] :P [16:32] why I wouldn't want an autorun on mount in my unix machines [16:32] he doesnt want autorun ya tard, he wants it to just MOUNT [16:32] cmagina: do you have any complaints to zul's plan to update tgt for lucid? [16:32] he wants the thing to mount automagically on boot [16:32] i mean like, for network settings youd have /etc/network/interfaces [16:32] oh heh [16:33] it confused me :) [16:33] hallyn: no for natty [16:33] ya, fstab :) [16:33] * EvilPhoenix cannot assist with fstab config :/ [16:33] also, what will happen when the drive isnt inserted? [16:33] it'll fail to mount :P [16:33] and probly spit an error into the logs somewhere [16:33] i tought it might keep spamming the terminal or somethign :p [16:34] zul: d'oh. [16:34] alvin: yes with no problem at all [16:34] it might, since I've never actually had an issue with mounting, since I have RAID setup on my server box :P [16:34] zul: in that case, pure awesomeness, thanks [16:34] na, normally only once [16:34] i havent had an fstab issue yet so meh [16:34] :P [16:34] :p [16:35] I have nfs fstab issues, I still can't solve [16:35] m_tadeu: In that case, it's a bug in either asterisk or upstart. I'd try filing with $ ubuntu-bug asterisk [16:36] is fstab a component of mount? [16:36] because fstab doesnt seem to exist alone [16:36] man fstab, /etc/fstab [16:36] mount uses fstab [16:36] basically like a config file [16:36] i see, thanks [16:37] except since you said you're a linux newb, i'd not go messing with fstab if I were you [16:37] unless you're guided through the process :P [16:37] you mess up fstab, you can seriously screw up your ability to boot :P [16:37] darn :p [16:39] EvilPhoenix: Can't get worse. The NFS part in my fstab is the same as the fstab of my collegue. My shares get mounted and his do not. We have the same errors during boot. (failed to resolve server, name or service not known, terminated with status 32) [16:40] He always has to do mount -a after boot [16:40] lol [16:40] add the name to the hosts file? [16:41] I just nfs mount with ip's [16:41] He's not here, but the machine is up. trying... [16:45] patdk-wk: Well, well... it works. What's the difference? [16:45] your dns server isn't working when he boots [16:45] or, his dns server [16:45] We have been searching for this a long time and I don't have to add the IP [16:46] what dns server you using? [16:46] and is it in your hosts file? [16:46] We're both using DHCP. [16:46] No, I don't have the server in the hosts file [16:46] what dhcp server? [16:46] bind. It's the same machine as the dhcp server [16:47] heh? [16:47] you have atleast 3 machines there? [16:47] Now his log doesn't complain about not finding the server. Mine still does [16:47] yours, his, and a dhcp/bind router? [16:48] patdk-wk: Yes, (and about 80 other machines) [16:48] your machine must be faster than his then [16:48] guess his is taking too long to get dhcp going and started [16:48] where yours fails a few times, but then gets dns working, and goes on [16:49] I think he's on a slower part of the network. His machine is a core duo. Mine is i5 [16:49] Before upstart, we could all boot and mount without trouble with the same fstab configuration. [16:49] yep [16:49] cause upstart changed it all [16:49] Same network, but with some slower machines. [16:49] before everything was in a nice neat order [16:50] EvilPhoenix: succes! :D [16:50] i edited the fstab, it works [16:50] long live wikipedia lol [16:51] now it's chaos. This is my boot.log .... http://pastebin.com/vT92eYde [16:52] One look at that file is always: omg! it's all going wrong! [16:52] (and that's the log when shares actually do get mounted) [16:52] it's the same issue [16:54] patdk-wk: Do you know an open bug about this? [16:55] I have searched and looked, and I think they are closed [16:55] atleast your issue is easy to fix [16:55] mine not so much, and stil not sure why [16:55] Well, workaround. [16:55] It's different? [16:55] well, if you didn't use dhcp, it would work without hosts, easily [16:56] (I did file some bugs about this, but they are set to resolved.) [16:56] I use ip addresses, not host names, so I don't have your issue [16:56] patdk-wk: I started using dhcp because it didn't work without it in the past... (resolved now) [16:56] but if my network card is down, boot hangs on nfs mounts [16:56] ah [16:58] There is this bug 275451 [16:58] Launchpad bug 275451 in sysvinit "nfs mounts specified in fstab is not mounted on boot. " [Undecided,New] https://launchpad.net/bugs/275451 [16:59] For some people it only works when using DHCP, for some it only works with static IP. [17:01] New bug: #715281 in postfix (main) "Upon update, /etc/postfix/main.cf stopped procmail processing" [Undecided,New] https://launchpad.net/bugs/715281 === ogra is now known as Guest60731 [17:16] When installing WebDav on apache. How do I get it to create seperate directories for each user that logs in? like https://webdav.example.com -> user a logs in -> data gets mapped to folder a [17:28] kirkland: ok so what exactly woulod be the purpose of adding the indicator for testdrive? [17:29] RoAkSoAx: if someone has TestDrive installed, it would tell them when a candidate ISO is ready for A1, A2, A3, B1, RC, or GA [17:30] kirkland: if it's just for that, we might as well have notifications (notify-osd) with configurable remainders or something [17:31] until I can find a better use for the indicator [17:31] i'll have it anyways [17:32] kirkland: now, how can I determine if there's candidate ISO ready for A1,etc,etc? [17:35] RoAkSoAx: great question; i'm not sure. hggdh: is there a way to tell, programmatically? [17:36] RoAkSoAx: i was thinking something along the lines of the way the envelop lights up green, when you have email in the inbox [17:36] New bug: #715294 in openipmi (main) "Loading ipmi_msghandler generates a kernel Oops" [Undecided,New] https://launchpad.net/bugs/715294 [17:36] doesn't notify-osd disappear after a while? Seems the messaging indicator is a better place for this. [17:36] kirkland: bingo! messaging indicator is t3h bomb. :) [17:39] SpamapS: yeah, i agree, notify-osd goes away [17:39] kirkland: SpamapS yeah messaging indicator then. The only problem would be to know how if the candidate ISO is available [17:39] RoAkSoAx: yeah, for that, let's talk to hggdh and the q/a team [17:39] s/how if/when is [17:39] hggdh: ping [17:39] RoAkSoAx: http://cdimage.ubuntu.com/releases/11.04/ [17:40] RoAkSoAx: i think that should tell you === ogra_ is now known as ogra__ [17:40] kirkland: no, there is no way to tell programatically if an ISO is a candidate... [17:40] RoAkSoAx: pong [17:41] kirkland: if they gives us a .manifest or .manifest-daily that would be the best [17:41] hggdh: "" [17:41] hggdh: hmm, well that seems to be something that we should remedy [17:41] hggdh: ^^ [17:41] RoAkSoAx: let's see if cjwatson has any ideas [17:41] kirkland: iso.qa.ubuntu.com will send out an email when an ISO is *officially* available for tests, but this is all [17:41] cjwatson: howdy; we'd like a programmatic way to determine when there are Alpha/Beta/RC candidate ISOs available === ogra__ is now known as ogra_ [17:42] hggdh: hmm, can we get that process to put up a flag somewhere wgettable, too? [17:42] RoAkSoAx: we need an out-of-band (I guess) process for that [17:42] kirkland: the best way right now is to scrape iso.qa.ubuntu.com [17:42] kirkland: this might be a way. [17:42] candidate images are put there [17:43] cjwatson: okay, thanks [17:43] I wouldn't want to add something on cdimage until we add some mechanism for us to programmatically push to iso.qa - otherwise things will get out of sync [17:43] one could wget iso.qa.u.c, and check the headers [17:43] cjwatson: fwiw, RoAkSoAx is going to add a Messaging Indicator to TestDrive, that will flag users of TestDrive when there are candidate images available [17:43] cjwatson: fair enough; scraping should suffice for a first cut [17:44] right, I scanned scrollback [17:44] it's a good idea [17:44] cjwatson: thanks [17:44] * kirkland is always nervous running ideas by cjwatson ;-) [17:44] * cjwatson is now known as ogre [17:44] * sbeattie recalls that stgraber add a manifest of some kind of official candidate images somewhere on iso.q.u.c [17:44] s/add/added/ [17:45] but there is an issue there -- right now, for example, iso.qa is open for alpha2 tests... [17:45] heh, I generally like the idea of development-process indicators, we just haven't done many yet [17:45] hggdh: well, we can clean up the metadata [17:45] iso.qa is definitely the canonical place right now though [17:45] cjwatson: I agree. [17:46] hggdh: I've hidden alpha-2 [17:46] (this may make it hard to test the indicator app though! maybe stgraber or somebody can help you bring up a test instance?) [17:48] i guess the easies way to go would be to have a manifes at iso.qa [17:48] yes, this may be too restrictive. Or leaving it open (so that the bug references can be looked at) and adding a message/metadata "tests open/closed"? [17:48] RoAkSoAx: and how do you find out if the manifest is current? [17:50] hggdh: if the manifest is not empty :) [17:52] hggdh kirkland ok but first, if an image is in iso.qa, would that be the same image as in cdimage.u.c ? [17:53] it would be the same, yes [17:54] perhaps the easiest way would be to add metadata to the page when the tests are opened [17:55] hggdh: otherway would be something like having a .manifest with "alpha2 desktop" "alpha2 server" etc etc. when the images are available. when they are unavailable, then cleanup the manifest and leave it empty [17:56] hggdh: so that whoever tests the ISO with testdrive, he/she can just use the daily image in cdimage.u.c [17:56] hggdh: but the manifest will only server for the notification purpose [17:56] s/server for/serve [17:57] RoAkSoAx: I bite, it is workable -- cjwatson, agree? [17:57] and who will do it? [17:59] I have to say, ubuntu is easily the best distro to use with EC2. Thanks anybody in here who is involved in that :) [17:59] hggdh: yeah cause be pretty much the same approach as the .manifest-daily on cdimage.u.c. If the image is available, make it available, if not, don't. So in the tracker, if the ISO image is available in the tracker, make it available, so that testdrive can make a notification [18:00] and then, testdrive will poll that tracker manifest every X seconds/minutes/hours and make the proper notification [18:00] hggdh: if it's autogenerated by the ISO tracker, sure [18:01] but I don't know that codebase [18:01] cjwatson,sbeattie : what's up ? [18:01] stgraber: welcome, sir! [18:01] stgraber: we are discussing how we can automate announcements of a new ISO candidate [18:02] stgraber: IIRC, you added a path to iso.q.u.c that published the paths of the current isos under testing. [18:02] s/path to/path on/ [18:02] patch? [18:02] stgraber: I don't recall what that path was (the dl-iso script doesn't make use of it) [18:04] stgraber: (by paths, I mean it published the urls on cdimages to the isos) [18:05] stgraber: do you recall the path on iso.q.u.c where that got published? [18:06] someone already has used virtualbox-4.0 on lucid ? [18:07] Does anybody know if Ubuntu runs on an HP DL385G7 Server - AMD Opteron 6172 processor? [18:07] sbeattie: good question, I remember I implemented that but I don't remember the URL, let me have a quick look. [18:08] sbeattie: /dllist apparently [18:09] yep, seems to work, though there's nothing on iso.qa.ubuntu.com yet [18:11] stgraber: 404s for me, hrm. [18:11] sbeattie: iso.qa.ubuntu.com/qatracker/dllist [18:11] stgraber: ah, cool! [18:12] not sure what the format is though, if I were to re-implement that now, I'd do it with a json interface like I did for my other Drupal project (Edubuntu weblive) [18:12] hggdh: okay, there's the URL for testdrive to monitor for candidate isos/images: http://iso.qa.ubuntu.com/qatracker/dllist [18:12] RoAkSoAx: ^ [18:13] stgraber: IIRC it's just a straight list of URLs, no JSON or anything like that. [18:13] I understand that if the dllist is empty, there is no test currently scheduled [18:13] correct? [18:13] sbeattie: yeah, I remember it's plain text, not sure if that's just URL or if it also exports the name of the build and version (though that can be guessed from the URL) [18:13] hggdh: yep [18:14] RoAkSoAx: sounds easily implementable on testdrive [18:15] hggdh: cool then. That will do [18:16] thanks === NG_ is now known as ng_ [18:46] evening [18:46] SpamapS: i uploaded the FTBFS for handlersocket [18:47] When installing WebDav on apache. How do I get it to create seperate directories for each user that logs in? like https://webdav.example.com -> user a logs in -> data gets mapped to his/her folder on the server. [18:51] zul: muy bien, gracias. :) [18:52] SpamapS: bien sur === ng_ is now known as NG_ [19:04] at random times, ebox (zentyal) magically stops responding when Im on the web interface of it through my ubuntu server, but when I go to my ubuntu desktop edition server it works fine. Then if I go back to the interface on my laptop, its magically working. What gives? [19:33] lol === ogra_ is now known as ogra [19:51] smoser: btw will start looking into the ebs root snapshots thingy after I eat lunch, in the next half hour or so [19:52] hey kirkland! I see that you have reviewed a proposed likewise-open SRU for maverick and lucid back in December. It's marked as "Needs fixing". Any idea if anyone is looking after that? [19:52] EtienneG: sorry, no :-/ [19:52] kirkland, as reference, that's related to https://bugs.launchpad.net/ubuntu/+source/likewise-open/+bug/534629 [19:52] Launchpad bug 534629 in likewise-open "AssumeDefaultDomain does not work" [Medium,Fix committed] [19:53] kirkland, no, as in "nobody is looking after it", or no as in "I do not know if someone is looking at it"? :) [19:53] EtienneG: "i don't know" [19:53] EtienneG: but "i doubt anyone is" [19:53] EtienneG: let me check the bug [19:54] kirkland, ah, ok! get it [19:54] err, *got it [19:55] EtienneG: hmm, where's the patch? [19:55] EtienneG: what bug is in "needs-fixing" state? [19:56] kirkland, not sure, that's bug #534629 [19:56] Launchpad bug 534629 in likewise-open "AssumeDefaultDomain does not work" [Medium,Fix committed] https://launchpad.net/bugs/534629 [19:57] EtienneG: ah, i see [19:57] kirkland, seems like you reviewed a branch on December 8th, it's marked "Needs fixing" [19:57] EtienneG: i reviewed the patch as part of my patch pilot day [19:58] kirkland, I am not 100% up-to-date on developer practice. IIUC, the "Needs Fixing" tag mean the patch needs to be cleaned up, right? [19:58] so I guess I have to poke the guy who did the original patch [19:58] EtienneG: https://code.launchpad.net/~ssalley/ubuntu/maverick/likewise-open/likewise-open.fix627272/+merge/38741 [19:58] EtienneG: there's my review comments [19:59] EtienneG: quoting myself, "I'll be happy to sponsor this as soon as (1) is trivially fixed in your branch, and as soon as each bug is updated per (2). Then, the package will go into the -proposed queue, and we'll need you or someone else to go through each of those 9 bugs and work their way through the reproduce instructions, noting if the new package fixes the known bugs and does cause regression." [20:00] EtienneG: the diff is HUGE for an SRU, Diff: 1427 lines (+1234/-32) 14 files modified [20:00] EtienneG: and it claims to fix 9 bugs [20:00] kirkland, ok, got it [20:01] kirkland, thing is: there's a lot of bug to fix in likewise-open [20:01] EtienneG: heh [20:01] EtienneG: yeah [20:02] EtienneG: it might make more sense to just get a newer version in lucid-backports [20:02] kirkland, but then lucid-backport come with no commitment of maintenance [20:02] kirkland, I will ponder the question [20:03] EtienneG: true [20:04] EtienneG: alright, well, if you or someone will update each of the bugs fixed by this branch with an SRU statement in the top, i will sponsor the upload to -proposed [20:04] EtienneG: where it will need a fair amount of testing [20:04] kirkland, I think that can be arranged. I will see if I can delegate that to one of my minion :) [20:04] EtienneG: coolio [20:04] EtienneG: poke me when done [20:08] kirkland: btw!! jsut remembered... can you please release a new PowerNap when you have a chance? [20:08] RoAkSoAx: sure [20:09] kirkland: and let's discuss the second stage thing for powersave later this week ;) [20:10] im thinking of trying out ubuntu server for the first time, but i wanted to better understand the packaging lifecycle before I choose LTS or not [20:10] does LTS have a fixed package set (versions) that just get security backports? [20:10] or do the actual package versions get update between LTS releases [20:11] lirakis: no release updates the major version numbers [20:11] lirakis, version are frozen at release time, and all releases (including LTS) only get security fixes [20:11] got it. [20:12] and .. are there migrations between releases for a box that has already been installed? or do you pretty much have to wipe out the box and restart [20:13] because, I would like the shorter release cycle of non-LTS ... but if it just means that every 6 months i have to reinstall ... thats not really worth it. [20:14] fortunately though, i guess... 10.04 was cut not too long ago, so its probably got very recent set of package versions [20:14] Is there some where I can check what versions of specific packages are for a release, without having it installed? [20:14] like... i need to have libnl > 1.1 and some other bits for building a wireless access point [20:14] lirakis: packages.ubuntu.com can be a goldmine for that kinda info [20:15] thanks again - i will check it ot [20:15] *out [20:15] do you guys generally see more people running LTS, or non-LTS [20:15] just want to know which has a larger user & support base [20:15] lirakis: LTS [20:16] I can only speak for myself, but LTS [20:16] at least imo is the ebst [20:16] by defitiontion support will be strongest for LTS [20:16] "larger user" base? who knows really [20:17] 6 months is great for a desktop where I constantly want more toys. but it's far too fast for servers. that'd give me 3-4 months of stable before I'm already testing upgrades again [20:17] lirakis: of course you can upgrade to the next release [20:17] lirakis, same here. I would stick with LTS unless you have a very good reason not to [20:17] well i know LTS is supported for a longer period of time ... so I guess it has more time to accumulate users.... but if it was significantly less popular, it wouldnt necessarily "by definition" have more suport base .. from a community perspective [20:17] okay ... sounds like LTS is the way to go ... and fortunately its a recent release [20:17] who knows/cares what the community does? [20:18] resno, ... so I can research other peoples mistakes, and solutions [20:18] :D [20:18] lirakis: ah... [20:18] im bound to make the same ones [20:18] but barring package versions .. ill get LTS [20:18] in my opinion LTS is the best way to go. [20:18] thanks for the info guys [20:18] i havnet even updated my desktops yet [20:19] lirakis, one of the thing to watch out for is hardware support. lucid has kernel 2.6.32. As long as it support your hardware, your golden. Otherwise, you will either need to go for a "regular" release, or use the lts backported kernel [20:19] EtienneG, thanks for the heads up - I am not running any hardware newer than ... 2-3 years old. [20:19] so I should be fine with 2.6.32. [20:20] centos 5 is running 2.6.18 ... and libnl was an issue with a kernel that old [20:22] this is my first foray into ubuntu based servers ... so again, i appreciate the info, thanks! [20:22] lirakis: what did you use before? [20:22] resno, centos for the last 5 years or so [20:23] lirakis, you are welcome. be sure to check the ubuntu server guide. It's nowhere complete, but it does have some good info [20:23] and ... gentoo previously [20:23] lirakis: heh, im just testing out centos now :) [20:23] zul: using the test rig? [20:23] resno, centos 5.5 is so old ... and 6 is too long in the waiting [20:23] hggdh: just cempedak [20:23] gimme a couple of minutes [20:24] at least i have some options with ubuntu for whether or not i want a more up to date, but less stable system ... centos doesnt have that [20:24] OK. Please do not reboot now, I will reset the systems to Hardy, and (then) put cempedak back into default PXE [20:24] so you will not get booted to Hardy also ;-) [20:24] EtienneG, I bought the official ubuntu server book just to have around .. but i will use online docs for sure [20:25] So, I've noticed that the placement of files for apache in ubuntu is different from what I'm used to (rhel based systems) which file would one place the DirectoryIndex in on ubuntu? [20:26] k4k: are you talking about index.html et al. ? [20:26] lirakis: i will say this, support here friendly then in #centos [20:26] guntbert: yes [20:26] resno, thats a nice thing :D [20:26] I'm used to that being in the /etc/httpd/conf/httpd.conf file [20:26] i also run ubuntu on all my laptops ... its a god send after using gentoo for so long [20:26] hggdh:its all yours [20:26] ... but i dont really know how it works behind the scenes so much [20:27] so .. i thought using it as a server might give me the opportunity to learn more about htat [20:27] *that [20:27] ive never had to "administer" a debian based system really ... so .. i figured i might as well learn since im using it for all my desktops now [20:27] lirakis: heh, its funny we are going opposite ways. me into centos,. you away [20:27] yeah [20:27] k4k: look into /etc/apache2/sites-available and sites-enabled [20:27] although i hold debain near and dear to my heart [20:28] guntbert: thanks I've been looking there, I'm assuming that's where they go, but there were no lines yet specified for "DirectoryIndex" in there so I was not sure [20:28] untill ubuntu .. i just didnt find debian a "reasonable" os to run. very dogmatic etc. [20:28] I'll try it and see what happens [20:29] ok .. LTS torrent is d/ling ... ill wipe out my box tonight and get started on it [20:29] im sure ill be around :D [20:29] k4k: seems the best way - I have no installed apache right now :-) === red2kic is now known as red2kic_ [20:37] zul: it is safe, again, to reboot cempedak [20:37] hggdh: cool thanks [20:46] smoser: ping [20:46] here [20:46] smoser: where's the best documentation for someone who wants to customize the Ubuntu EC2 AMI? [20:47] there isn't really good documentation :-( [20:47] but, for an ebs image, use 'create-image'. [20:47] for an instance store private image use 'ec2-bundle-vol' [20:48] with either of those routes, you really should not make things public, just because there is a high possibility of leaking things (your credentials and such accidently) [20:49] if you want to do it "right", then download image from http://uec-images.ubutnu.com and mount loopback, chroot, do things, ... [20:49] smoser: he definitely wants to keep them private [20:50] smoser: okay, so the loopback/chroot is the "correct" method? [20:50] then euca-bundle-vol or create-image if you can be private. [20:51] loopback and chroot via some revision controlled script, yeah, is what i'd really suggest. [21:02] OK I've been going around in circles here. I have apache site-available/default set to /var/www like normal. I have index.php set as a DirectoryIndex for that site. I have a2enmod php5 says it's enabled and I have my index.php file in the document root. However, it still wants me to download the php file, not show it to me. [21:03] k4k: can www-data execute the file? [21:03] www-data? [21:03] k4k: The user that apache normally runs as. [21:04] good question...I will be so ashamed if that's all it is... [21:05] does there need to be the execute flag on the file or can it just be rw-? [21:06] ...meh, just answered that for myself, didn't make a difference [21:17] k4k, you'll need "AddType application/x-httpd-php .php" somewhere [21:17] like /etc/apache2/httpd.conf [21:18] were can i download a xen based ubuntu=server [21:19] lenios: would that be apache2.conf or in the sites-enabled/IOB file I'm using? [21:21] apache2.conf [21:22] lenios: ok I did that right below line "DefaultType" and it still won't work, I'm now getting a blank "index of /" page [21:23] when getting http://host/index.php ? [21:23] no when going to http://host [21:24] doing /index.php after it get file not found which means something isn't looking in the right place, but I've got all the files pointed at the directory with this php file in it [21:25] in what directory is the index.php file? [21:25] can anyone help-me with nagios? [21:25] (full path) [21:25] /var/www/outboard/ [21:25] http://host/outboard/index.php ? [21:25] and I have DocumentRoot set to that same path and set as well [21:25] oh [21:26] does it work with an html file? [21:26] let me check [21:27] how can i add hosts to nagios? [21:27] lenios: nope, so it's definitely looking at the wrong directory me thinks [21:27] i think so too [21:28] what else could be telling it where to look though? [21:33] omg I got it [21:33] how stupid [21:35] for future reference, if anyone cares. Moved /outboard to /opt/outboard and re-pointed sites-available/IOB to point there instead and it worked...had to have been permissions related [21:50] k4k: Uh.. You kept saying it was in /var/www/outboard/? [21:58] kirkland: something like this? http://me.roaksoax.com/Screenshot.png [21:58] SpamapS: ^^ [22:02] random question; how can i chmod my file to look like this? -rwxr-xr-x 1 root root 3000 2007-03-06 22:40 /etc/init.d/rtorrent [22:02] i did sudo chown -R 0755 /etc/init.d/rtorrent [22:02] but now its -rw-r--r-- 1 755 root 4394 2011-02-08 22:23 /etc/init.d/rtorrent :p :s [22:03] why not a chmod? [22:03] whoops, pasted a second smiley :p [22:03] dunno, is there a difference? [22:03] man chmod [22:03] man chown [22:03] k [22:03] RoAkSoAx: thats "bloody brilliant" [22:04] how do i close the manual? lol [22:04] q [22:04] thx [22:06] got it, thanks [22:06] SpamapS: hehe now the hard part >P make it work rather than just display it lol! [22:09] does anybody know if there's a GUI for setting up freeradius on ubuntu? [22:30] kirkland, regarding the likewise-open bugs we discussed earlier, if I want them to be SRU'ed in both lucid-proposed and maverick-proposed, I first need to confirm in each respective release, and then use "Nominate for series" to nominate them to lucid-updates and maverick-updates. Correct? [22:33] EtienneG: correct [22:34] RoAkSoAx, thx, mang! [22:36] no probs ;) [22:36] RoAkSoAx, if I am not mistaken, it is the SRU team that add the verification-needed tag, right? Or should I do it myself when I add the SRU blurb to the bug description? [22:38] EtienneG: correct! the SRU team is the one in charge of that [22:38] cool beans, thx again [22:38] Anyone know what LSB information is? [22:38] update-rc.d: warning: /etc/init.d/rtorrent missing LSB information [22:39] EtienneG: so you should just get it fixed, get it uploaded or someone to sponsor the upload. Then the SRU team will put it in -proposed and add verification-needed for people to verify it actually fixes the bug. [22:40] RoAkSoAx, got it! [23:11] RoyK: hi [23:13] how's people? [23:17] <_Steve_P-T_> Hello everyone, Not sure if I'm in the correct place to ask about a vsftp problem? [23:17] _Steve_P-T_: try us [23:17] _Steve_P-T_: we hate FTP though [23:18] <_Steve_P-T_> I will ask it anyway, How you you share more folders in users other than there homes [23:19] <_Steve_P-T_> I was told to use the mount -- bind /folder /userhome/folder is that correct? [23:19] _Steve_P-T_: can you repeat the question? [23:19] _Steve_P-T_: I don't understand the question [23:20] <_Steve_P-T_> sorry, I will try again.. [23:21] <_Steve_P-T_> How do I share floders with ftp users, SVFTP locks user into their homes. But I would like to give acess to other folders [23:22] <_Steve_P-T_> is the best way to mount --bind /folder /userhome/floder ? [23:22] _Steve_P-T_: right off the bat everyone here would suggest SFTP === _TechAway_ is now known as _Techie_ [23:23] _Steve_P-T_: but for VSFTP it's probably in it's config, or it could possibly be as easy as sudo chown folder [23:23] <_Steve_P-T_> ok thanks very much, will do some reading about SFTP [23:23] _Steve_P-T_: or changing directory permissions so that all users can view those folders. [23:24] _Steve_P-T_: what are you running? [23:24] _Steve_P-T_: I have a guiless ubuntu server [23:25] <_Steve_P-T_> I have running ubuntu server 10.10 with vsftp works fine [23:25] _Steve_P-T_: Then SFTP is already installed basically [23:26] _Steve_P-T_: I use putty to remotely manage my server from my windows 7 netbook. [23:26] _Steve_P-T_: and then I use FileZilla to connect via SFTP to the server. [23:26] <_Steve_P-T_> I also connect with PUTTY frm W7, mediatomb works fine. server is headless. [23:27] <_Steve_P-T_> I have an archive which I need to share with a number of users whcih is spread over a number of folfers [23:28] _Steve_P-T_: in that case, connecting with a SFTP client is as easy as putting in your username and password that you use for putty [23:29] _Steve_P-T_: what ftp client do you use? I use FileZilla, which has a "File" -> "Site Manager" - > SFTP server connection area [23:29] <_Steve_P-T_> I think the term I was looking for was virtua folders, can SFTP creat those , Filezilla is used by the users [23:33] _Steve_P-T_: I believe so [23:34] _Steve_P-T_: to create a virtual folder, or link, I think that's lk /wherever/wherever/ [23:34] _Steve_P-T_: I warn you though, I'm a newb with linux. >.> [23:34] _Steve_P-T_: I was astouneded at how EASY it was for me to connect with SFTP [23:34] _Steve_P-T_: and how easy it was for me to get started, I was fearing a huge ordeal [23:35] <_Steve_P-T_> yes I know what you mean, === _Techie_ is now known as _TechAway_