[00:00] update* === rideh_ is now known as rideh [00:38] if i have installed ubuntu-desktop . how to make ubuntu/kubuntu boot and only go to console automatically. do not load gui? [00:40] sarnold: so, in theory, there's no easy way to install an old version of postgres, even if I have the .tar file? [00:41] always thought it was as easy to install old versions of things in linux as it is in windoes [00:41] Quest: create an override file for lightdm's upstart job: http://upstart.ubuntu.com/cookbook/#override-files [00:43] izanagisan: if you're starting from a source package, you've got to compile it. if you just want an old binary package, you can go find those (for example, https://launchpad.net/ubuntu/+source/postgresql-8.2 ) [00:44] izanagisan: but 8.1 looks too old, if you change that 8.2 to 8.1, you'll see what I mean.. [00:44] oh, ok. Must be easy to restore the 8.1 database into 8.2 [00:45] it's just that restoring a 8.1 database dump into 9.2 is damn near impossible [00:45] sarnold: thanks a lot : ) [00:46] heh, even the 8.1 package directory is empty: http://archive.ubuntu.com/ubuntu/pool/main/p/postgresql-8.1/ [00:46] .. I thought it'd be there. [00:52] sarnold: I checked the downloadable files, but I get .tar.gz files [00:52] so, I'd also have to compile them from source, I think? [00:52] as you can see, I come from UbuntuDesktop and have never compiled from source [00:52] so I'm not even sure of what's going on there [00:53] oh no! [00:53] found the .DEB [00:53] : ) [00:53] yay! [00:53] where was it? [00:53] I can apt-get install a .deb, right? [00:53] http://launchpadlibrarian.net/12789312/postgresql-8.2_8.2.7-1_ia64.deb [00:53] dpkg -i .. [00:53] thanks! [01:04] I get expectable dependency problems. How do I download packages? with aptitude?. For instance: "Package libkrb53 is not installed" [01:06] izanagisan: you could try installing those with apt-get install libkrb53 or whatever you prefer.. hopefully you won't need similarly ancient versions of the other packages [01:07] well, I did *apt-get -f install* [01:08] and that somehow solved those issues [01:08] I also had to download the postgresql-client-8.2 [01:08] so I did and now it would appear that I have postgresql 8.2! [01:09] hate when I don't quite understand what happens but oh well, I need to run that test [01:20] how do you people keep your ssh keys if you have to use several desktops for example and u need to access servers from all of these desktops? [01:20] i give each system its own SSH key [01:20] and each of them is able to SSH auth to the user on the server (never root directly) [01:20] TheLordOfTime: do u add passphrase as well? [01:21] absolutely [01:21] usually a different one for each system, but i'm hyper paranoid [01:21] TheLordOfTime: how do u keep track of all of these passphrases? [01:21] * TheLordOfTime can't tell you. [01:21] secret. [01:21] TheLordOfTime: iohhh CAMON [01:21] nah i kid, i keep an encrypted database with the passcodes/keys, if I forget. [01:22] TheLordOfTime: likee keepassx? [01:22] i don't like keepassx, but that's one method [01:22] TheLordOfTime: truecrypt? cryptfs? [01:22] * TheLordOfTime GPG encrypts a sqliteDB with the data on it, and then encrypts the folder the GPG-encrypted sqliteDB is in. [01:23] as i said, i'm hyper paranoid, don't use my methods for normal use [01:23] TheLordOfTime: i see... [01:23] but generally speaking I will create an SSH key for each system. [01:23] and place its public key data into the authorized_hosts file on the server those systems SSH to [01:24] (the only system that I do NOT have a ssh key passphrase on is one which requires two-factor authentication, biometric+smartcard, to login [01:24] but that's an implied hyper-secure system) [01:24] ThatOneRoadie: dafuck who are u working for - pentagon? [01:24] xD [01:24] oh btw [01:24] !language [01:24] Please watch your language and topic to help keep this channel family-friendly, polite, and professional. [01:24] ^ that [01:24] TheLordOfTime: Error: "that" is not a valid command. [01:25] SHUT UP BOT! [01:25] * TheLordOfTime kicks uvirtbot into /dev/null [01:25] ^ dpck [01:25] izanagisan: Error: "dpck" is not a valid command. [01:25] heheh [01:25] ^ dpkg [01:25] izanagisan: Error: "dpkg" is not a valid command. [01:26] TheLordOfTime: it's almost kinda be cool to have some usb drive-based keychain or something... [01:26] ah, I get it, it's not a valid command for the bot, thought it checked Ubuntu commands [01:26] izanagisan: :P [01:26] nah. [01:26] it's just being stupid. [01:27] ddsss: i don't work at the pentagon. but i'm also NOT saying where I work [01:27] because i'm not in the mood to talk about that [01:27] TheLordOfTime: yep. there are many otehr 3-letter agencies. [01:27] having said this, you can either (1) clone your ssh keys across all your systems so they're shared, or [01:27] (2) create an SSH key for each system accessing the remote system, and on the remote system add each key into the authorized_keys file [01:27] or some weird auth system other than that one, but to answer your question... :P [01:28] TheLordOfTime: yeah. I guess that's what I'm gonna do. [01:31] the complication is that if you often ssh from one 'desktop' to another 'desktop', compromising one key probably compromises all the keys. I think it's a bit false-sense-of-safety to have multiple ssh keys for multiple client machines, if you ssh among them regularly. if you never ssh -to- a certain machine, giving it a unique key then -does- feel useful to me. (say, your phone or tablet or travel-laptop..) [02:45] 4/exit [04:30] is it possible to create a bridge interface like br0 with no actual devices in it? [04:31] like an empty bridge staged for future use? [04:57] phunyguy: I think either the libvirt or lxc startup scripts create an empty bridge at boot.. [05:06] can someone please help me why flash not woring in ubuntu server> [05:07] im using 12.04 LTS [05:07] ive tries tons of methods just not working [05:07] manual install.. flash 11.. and the newer plug in [05:07] none of them work [05:12] is there a command i can run that searches for specific text in say, php files? [05:14] petey: grep -r "specific text" /path/to/top/directory/of/files [05:14] perfect! [05:14] thank you very much [05:52] I'm having a problem with vsftpd, internally on an ftp conection I'm able to see the directories that I specify, if I access it from an external address, I have to force it to use active mode, and it shows no directories [05:55] cuken: does your ftp server have an externally routable IP address? [05:56] cuken: How does your nat/firewall handle the data channel? [05:56] yes it does [05:56] I have it on a dynamic dns [05:56] accessible through a name, I'm able to make a connection [05:56] and pass my credentials through [05:56] :andol I changed the default port and it appears to be open [05:57] Still sounds like you have your control channel working, but the data channel being the problematic one. [05:58] ahh, ok so I need to open the data channel as well? [05:58] yeah [05:58] is it possible to have the inbound and data share the same port? [05:59] use sftp instead :) [05:59] Yeah, sftp alt. http/webdav is really preferable to ftp in almost any case. [06:00] cuken: http://wooledge.org:8000/FtpMustDie [06:00] ;-) [06:00] sheesh, I've uncovered a movement! I'm sorry! [06:00] that said, stuff like wordpress seems to like having ftp available :-P [06:00] I just wanted easy access to a share, cant we all just get along [06:01] sftp is easier too! :) [06:01] ok so use my ssh server with sftp [06:01] Exactly, nothing easy about ftp. [06:02] Due to your wisdom oh powerful ones, ftp server is dead [06:03] Thanks for the tip guys [06:03] yw :-) [06:03] woo :) [06:04] * Nafallo sings ♫ another one bites the dust ♪ [06:08] hrm. that URL I posted changed location... [06:08] http://mywiki.wooledge.org/FtpMustDie === security is now known as megha [06:36] I'm failing on this sftp setup lol [06:37] I'm getting a message about a non trusted key, and then after i click accept anyways, it looks like my server is blocking it [06:48] Hi, i just found a /dev/.tmp11-1 what is this? is it legitimate? [06:50] xro: i don't have one on my raring laptop.. check lsof to see if anything is using it? [06:52] Did I read correctly that the root directory of your CHRootDirectory in sftp needs to be owned by the same group as the user connecting? [06:54] cuken_: sshd_config says "all components of the pathname must be root-owned directories that are not writable by any other user or group" -- I'd guess that group ownership is only necessary if you want to grant read access via group permissions rather than to everyone [06:55] Hm. I'm not understanding that correctly then. I have a ZFS pool mounted at /storage [06:55] currently my regular account owns /storage, and all subfolders within it [06:56] I'm trying to delegate sftp access to my girlfriends account and mine so we can access it externally from our network, so I created a CHrootdirectory to be based off of storage, but external calls are getting blocked [06:56] did I just set it up completely wrong? [06:56] erm not blocked but denied by sftp [06:57] /storage would need to be root-owned and not writable by your user.. [06:59] well thats not good lol [06:59] i have all my applications that dump data into storage running on my regular account [07:05] cuken_: why chroot? [07:05] that was the tutorial iwas reading [07:05] didnt know I had an option for something else === smb` is now known as smb === roniez is now known as roniez[af] === highvolt1ge is now known as highvoltage [10:37] hi all, I've some how managed to brick a rabbitmq install, used for openstack. Anyone know how to fix, service won't install or start http://pastie.org/7660220 ??? [10:46] ive tried apt-get remove and purge follow by autoremove, still no luck [10:46] ubuntu 12.04 server === security is now known as megha [12:37] jamespage, I am preparing a security patch for tomcat6. I want to test it before I load it up. jdstrand said that you might had a look before at the testsuite provided and might be able to help me. [13:07] ckuerste: jamespage is at ODS in California at the moment. It finishes today - I think he's back early next week. [13:07] okay, thanks. [13:12] Hey. I'm wondering how much efford it is to set up VPSs on my own. I have a scenario where I want to run 4 VPS on one machine. Does anyone know how much work is involded setting this up and maintaing the main system afterwards? What's the best technique? KVM? [13:17] I guess you mean "VM"? [13:18] yes -.- [13:19] The question is, if this is doable with medium linux skills or how much I have to learn in order to run this host machine [13:19] qemu-kvm is probably best bet yes, many prefer to use it with libvirt to assist management [13:19] certainly doable, so long as you are not afraid to 1) google and 2) read [13:19] Okay, thanks for the suggestion mardraum. === wedgwood_away is now known as wedgwood [13:21] Probably a silly question, but here goes. I'm running Ubuntu Server 12.10 and I wish to create a few shared folders that are going to be accessed from a Windows system. I.e. I want to map the network drives. Previously I've done so with Samba, but I kind of missed having the ability to see space remaining and such. Would NFS be better in this scenario? [13:22] nfs client on windows? no. [13:22] also hi sarnold - maxb n' izanagisan :) [13:22] afternoon :-) [13:23] Thank you again for the help yesterday :) [13:23] mardraum is there any way for a linux share to display on the remote computer what kind of space is remaining? :) [13:23] a linux share? [13:23] you mean samba? [13:24] Whatever gets the job done, really. I'm just wondering if I can make a network drive on my ubuntu box that I can add on Windows, and see how much space has been used out of the available space total. [13:24] i.e. like a normal network drive. [13:25] Haven't gotten that to work with Samba. It just seems silly to check the server every time I need to see free space. [13:25] Apologies if I'm not explaining myself properly :D === me is now known as Guest21897 === Guest21897 is now known as dir [15:47] does godaddy not support grep? [15:47] i tried running grep on a godaddy server (which worked great on my VPS) but nothing showed up on the godaddy one === darrenjones is now known as DazJones [17:01] is there a way to tell apt to not try to execute things in /tmp/ ? i have it mounted noexec for a reason, but if I can set that working directory to be elsewhere, I can avoid problems. [17:04] are vservers as good as OpenVZ vps [17:05] Nothing definitive on if I can hot-add cpus in kvm on 12.04.2 server. Anyone know more on this? I have a vm with 2 current, 16 max cpus. Bumping current to 4 doesn't appear to affect the guest OS (also 12.04) [17:20] GrueMaster: I don't think that's currently possible. Try asking on #virt @ irc.oftc.net [17:20] k, thx. [17:24] http://www.linux-kvm.org/page/CPUHotPlug [17:35] Yea, I read that (and several other web pages). Nothing definitive, everything contradictory. [17:37] "A new cpu should appear on /sys/devices/system/cpu" - this is the part that appears to fail for me. virsh vcpuvount changes accordingly when I use virsh setvcpus <#> --live, but the guest doesn't get an acpi event. [17:43] What's the difference between a Cluster Instance and the Regular Ubuntu server on Amazon Web Services? [17:43] good afternoon [17:43] just tested with 13.04 - no new cpu found after that command [17:43] (for GrueMaster) [17:45] Hmm. Not sure if this is a bug or incomplete feature. [17:46] imho an incomplete feature is a bug :P === yofel_ is now known as yofel [17:56] Hey, I should be ok using an openvpn install script written for an OpenVZ vps on KVM V vps, right? [18:00] sw0rdfish: I'd expect the openvz-based script to create and modify network interfaces in an openvz-specific way === Ursinha is now known as Ursinha-afk [18:31] hmmmmmmmmmmmmmmmmmmmmmmmmmmmm. === Ursinha-afk is now known as Ursinha [18:58] guys, I am kindda noob in terms of Ubuntu... [18:58] how can I download an old version of a package? [18:59] the server where we run redmine was on version 10.04. But today it got upgraded to 12.04. The upgrade went just fine, but now redmine (which is version 1.1.1 (pretty old) won't work) [19:00] one thing I noticed, reading it's documentation is that it requires rubygem-1.3.1 [19:00] and now we have rubygems-1.8.5 installed [19:00] (among other dependencies) [19:01] Is the only way to do that downgrade by pointing apt-get to use 10.04 repositories? [19:03] Kenjiro: you may wish to read up on package pinning; you can enable multiple repositories and pin specific packages to come from different repositories. though I wonder why you cannot just use a newer redmine that works with newer software? [19:05] sarnold: my coworker was trying to make a newer install, on another server. He was having a hard time to make it work [19:06] that's the problem when something is installed and configured by some who is no longer in the company... [19:06] and worse, he hasn't documented anything he did to make the stuff work back then :( [19:07] *someone [19:07] Kenjiro: oh man :( my condolances :( [19:07] yeah [19:17] good afternoon all... I need to know how to tell if my ubuntu lts (latest stable) has a working outbound email... how would I tell that? [19:18] Arrick: mail someone@somewhare.com < /dev/null [19:19] not currently installed. [19:19] how do i install it? [19:19] then install it [19:19] apt-get install mail-utils iirc [19:20] unable to locate [19:20] apt-get install mailx [19:24] ok, it was mailutils [19:24] no dash [19:30] ok [19:31] thanks... its not doing anything after I set it up with all defaults... [19:31] install postfix [19:32] iirc the default is another mta [19:32] did that, even reconfigured it according to http://tsengf.blogspot.com/2012/01/send-email-from-command-line-in-ubuntu.html [19:33] does mailq show anything? [19:33] got it to work, had to use satellite in setup [19:33] thanks. [19:39] hrm... how do i change the name that it is sending as to noreply@ ? [19:46] RoyK, I see th following at a site regarding sending as a generic account... [19:46] >>/etc/postfix/generic echo "http@example.com admin@example.com" >> /etc/postfix/main.cf echo "smtp_generic_maps = hash:/etc/postfix/generic" [19:46] postmap /etc/postfix/generic [19:46] service postfix restart [19:46] but when I run the first part, it says "no such file or folder found" [20:39] hi all. Good evening-morning-night [20:39] found a contradiction on IBM's official guide on RAID [20:40] "A larger stripe size produces higher read performance. If your computer regularly performs random read requests, choose a smaller stripe size. The default is 64 Kbytes" === wedgwood is now known as wedgwood_away === wedgwood_away is now known as wedgwood [20:41] what is it then? if larger stripe size produces higher read performance, then should I not choose a larger stripe size if my disc array will be used mostly for reading? [20:43] (full disclosure: I'm installing RAID 5 on 3 physical drives that will hold a small-ish PostgreSQL database) [20:49] good localtime(); [20:49] can anyone tell me why I am getting this from a ubuntu server with apache2?? Incorrect access detected, this server may be accessed only through "http://10.3.20.9" address, sorry. [20:49] Please notify server administrator. [20:50] izanagisan: for a small postgresql db on hardware raid, use a smallish blocksize [20:50] 64kB seems like a good choice [20:54] I have a server with a Pentium D 2.8GHz. Is it possible to set cpu-scaling on it? [20:54] running ubuntu 12.04 [20:58] RoyK: thanks a lot! testing with 8KB, will then use 64KB [21:19] izanagisan: I just use sw raid these days [21:20] easier to manage when the shit hits the fan [21:20] 8k blocks seems rather low [21:26] izanagisan: is that the same box you were installing yesterday? [21:47] markthomas: indeed. Still testing [21:47] yesterday I configured RAID 1 on 2 of the drives, stripe size of 128 KB [21:54] izanagisan: Have you checked the postgres documentation for recommendations? === Ursinha is now known as Ursinha-afk [23:01] markthomas: searched for official stances but found only forum posts [23:01] seems like the agreement is that 64 KB will be optimal for my needs [23:06] izanagisan: sounds like a plan. === wedgwood is now known as wedgwood_away [23:50] Somebody here could help me with deluge on ubuntu server?