[00:00] Hi [00:00] !ops | excess flooder! [00:00] excess flooder!: Help! Channel emergency! soren, lamont, mathiaz, Pici, Daviey, Tm_T or pmatulis [00:00] HELP About PPTP Server on Ubuntu 10.04 !!! [00:01] is there a way to know a specific server got unresponsive for some wierd reason and there is no option left except for reboot ? [00:01] kaushal: you need to reboot machine ? [00:01] yeah [00:02] since its not responsive [00:02] from terminal or SSH ? [00:02] can i know the reason for being unresponsive ness ? [00:02] you try with sudo reboot ? [00:02] i mean what caused the server to crash or not responsive [00:04] okay, i think i got jcastro's user dealt with by contacting that user another method ;P [00:04] HELP About PPTP Server on Ubuntu 10.04 !!! [00:05] Don't use PPTP, it's weakly encrypted. [00:05] twb: what to use better ? [00:05] ipsec/l2tp or openvpn [00:06] do you know any tutorial [00:06] ? [00:06] twb: hi [00:06] how to install and configure [00:06] Xelmep: not off the top of my head. Have you checked the Ubuntu Server Guide? [00:06] https://en.wikipedia.org/wiki/PPTP#Security_of_the_PPTP_protocol [00:06] is there a way to check "what caused the server to crash or not responsive" ? [00:06] not [00:07] i checked for pptp [00:07] Technically you could use EAP-TLS with PPTP, in which case it's no worse than using EAP-TLS for 802.11 WPA2-Enterprise, but I have never seen that in the wild [00:08] aha [00:08] ok [00:08] what is better openvpn or ipsex [00:08] ipsec [00:08] :) [00:08] But at that point you're rolling out an TLS cert hierarchy anyway, so you're better off using something else. [00:08] ok [00:08] thanks [00:08] Xelmep: well, ipsec is a standard part of IPv6, so IMO it's a better thing to learn long-term [00:09] i will try to install openvpn [00:09] Xelmep: for a simple point-to-point connection openvpn is very easy, for a hub-and-spoke layout it requires you to understand how TLS/x.509 work [00:10] ok [00:10] thank you, now i will try to install and configure [00:38] RoAkSoAx: How is it going? [00:48] Daviey: good good [00:48] Daviey: you? [00:48] lol === Ursinha_ is now known as Ursinha [00:54] RoAkSoAx: tired :) [00:54] RoAkSoAx: Were you able to look at ipxe? [00:55] Daviey: looking at it now [00:56] Daviey: though for what I'm reading in the documentation, it might be easier to use koan cause eoither way, we ould have to 1. flash the rom of the NIC, 2. chainload, and chainload means telling the DHCP to use the ipxe rom [00:57] RoAkSoAx: great, i'd love to hear what you think.. But i need to go afk now. Can you let me know when you are done, and i'll see it in scrollback? [00:57] RoAkSoAx: wait, flash to nic? [00:57] Nooo [00:57] You /can/ do that.. but it's a pure option. [00:58] Daviey: right, that's why I'm saying option 1: Flash NIC, option 2: PXE chainloading with ipxe [00:58] Currently, grub-ipxe plugs into grub like another kernel [00:58] So we chainload from grub. [00:58] Daviey: right, let me look at grub-ipxe then and we can catch up later or tomorrow morning [00:59] RoAkSoAx: Yeah, if you apt-get install grub-ipxe, you'll see the option in grub menu. [00:59] It'll give you a menu, but you can embed a script which hard codes (or other stuff) options [01:00] So we can gather from avahi where cobbler is, and inject it into the binary. [01:00] meaning that we are not dependant on dhcpd options. [01:01] The problem is, it's not clear to me if we can embed the script file without re-compiling. [01:03] Daviey: right, if this is only for reinstallations maybe koan is better option [01:03] Daviey: if this was for the cases on which we have a system without a OS yet, then is more useful [01:05] Daviey: koan (before the avahi stuff was removed from cobbler) was supposed to be able to automatically discover where the cobbler server is and boot from there [01:06] RoAkSoAx: noooo [01:06] How is it more useful with systems with no OS? [01:07] If the system has no OS, it has no grub - and therefore no grub-ipxe. [01:07] Daviey: exaclty [01:07] Daviey: that' was my point :) why do something that is already done by koan :) [01:07] uh? [01:08] Daviey: so, I install grub-ipxe, reboot, and the grub allows me to PXE boot, right? [01:08] RoAkSoAx: Right, that requires we ssh into the client.. and run the command? [01:08] RoAkSoAx: yup [01:09] Daviey: that means we need to install grub-ipxe and reboot whenever we want to reinstall [01:09] Daviey: so in a way, ssh into the client, and run the command, correct? [01:09] RoAkSoAx: Perhaps a phone call would help explain the point? This seems to be circular. [01:10] Daviey: yeah I think that would be best. But we can take it tomorrow if that's better for you since you were on your way out :) === pfibiger` is now known as pfibiger === twb` is now known as twb === bladernr_afk is now known as bladernr_ === bfreis_ is now known as bfreis [05:07] #dovecot is asleep [05:07] Anyone care to sanity check my dovecot change to enable a virtual (i.e. non-PAM) account? [05:07] http://paste.debian.net/147395/ [05:08] I'm a little bit scared to try it on the production system without any testing or peer review -- also I don't know offhand how to generate an SSHA [05:12] reading [05:13] twb: my dovecot is weak.. but it looks sane enough. [05:13] OK, thanks === bladernr_ is now known as bladernr_afk [05:47] hi all, i have a handful of vms that will be running apache2 and various websites, i want to put an accelerator "in front" to point to machines based on requested url -- is squid the way to go or is something like varnish "better" what else is there ? [05:48] pukeko: main problem with varnish is you need to install cc [05:49] I'd use squid always, but then I'm a *tad* biased [05:49] Also I wanted to do LDAP-backed HTTP auth to lock down access to most of my http backends, so I ended up using apache's mod_proxy :-/ [05:49] lifeless: squid's a pretty traditional (read: old fuddy-duddy) choice :-) [05:50] twb: heh [05:51] ..ok then squid is fine.. all i want it to do is direct traffic via url request, can i config it not to cache ? [05:53] or do i just tell it to not cache anything over some tiny kb ? [05:54] either of the abve [06:02] I also looked at nginx but the fact it has other baroque stuff like a pop3 proxy built in, that kinda spooked me [06:07] ok thanks for the help.. bbl [06:23] Hi [06:25] Does anyone here can help me on the Ubuntu server? I'm very new to Ubuntu. [06:34] auston: just ask :) [06:37] I have just installed Ubuntu server 11.10 and has configured 2 NIC. Going to use for DHCP and Firewall purpose. [06:38] But I wish to know step by step how to setup for DHCP first. [06:48] auston: https://help.ubuntu.com/10.04/serverguide/C/ [06:56] I just installed dhcp3-server but when accessing the .conf file it stated New File. Is that the correct way? [07:02] auston: even better, https://help.ubuntu.com/11.10/serverguide/C/ (updated for 11.10 changes) [07:02] I believe 11.10 uses isc-dhcp-server [07:13] I try to install using isc-dhcp-server for 11.10. [07:24] auston: what's the problem ? [07:37] having problem downloading isc-dhcp-server [08:00] hi all [08:05] auston: can you explain beyond " a problem" [08:11] Having difficulty with /etc/security/groups.conf not autoassigning certain groups (i.e. plugdev) to LDAP users on the client system, thoughts? [08:11] I have pam_group.so included just fine for PAM and everything, yet it does not work. [08:15] Now, I'm able to download and installed the isc-dhcp-server. Just now was due to the internet connection problem. [08:19] I got error msg when restart isc-dhcp-server. It says, "Configuration file errors encountered -- exiting. [08:21] I found the problem with the line missing ";" at the end of file === huats_ is now known as huats [08:40] is this ok for an quick and dirty /etc/hosts file 192.168.0.77 www.example.com,support.example.com,ftp.example.com ? [08:40] i mean are the commas allowed ? [08:42] why i always got problem to ifup with my eth1? It says, R.. file already exist! when I editting the network interface conf file. [08:53] Berge: ikke ISDN dialup, med andre ord? [08:53] return -OWRONGWINDOW; [08:54] hi all === smb` is now known as smb [08:56] !virt-manager [08:59] RoyK: Never realised you were Danish. [08:59] RoyK: Or is that also valid Norwegian? [09:01] hey guys i need some help [09:01] how can i get virt-manager to connect to the local host for xen [09:02] virt-manager -c xen:/// [09:02] I think. [09:02] * soren never really used Xen. [09:02] im setting that up via the GUI yet its saying it cannot connect [09:03] Er.. [09:03] Did you try the command I just gave you? [09:03] yes that worked but for some reason its not seeing the hypervisor [09:03] Ok. [09:03] which is already installed [09:03] IIRC, you need to enable... [09:03] er... [09:03] the http endpoint? [09:03] xend has been started [09:04] ? not understanding what you mean [09:04] Can you pastebin your xend config? Maybe that'll job my memory. [09:04] jog my memory, even. [09:05] soren: http://paste.ubuntu.com/753444/ [09:05] Ah, yes, there we go. [09:05] ADd this: [09:05] humm i did ifconfig [09:05] i have 2 bridges [09:05] (xend-unix-server yes) [09:05] ...and you should be golden. [09:05] (well, restart xend afterwards) [09:06] The xen:/// driver in libvirt talks to Xen through that interface. [09:07] which interface [09:07] the virbr [09:07] and do i need to have the bridge scripts uncommented still [09:07] guess not im connected :D [09:08] soren: is the a gui front end to create the vm's [09:08] No, not network interface. [09:08] I don't understand your last comment? [09:09] do i use the command line to create the guests or is there some gui based application to create virtual machines [09:09] I don't know what people do nowadays. [09:09] ok ty [09:10] soren: how is the xen libvirt intergration these days ? [09:10] ikonia: No clue. I don't use Xen if I can avoid it. [09:10] Which I can. [09:10] ha ha, same as me then [09:10] whats everyones aversion to xen what advantages does kvm have over xen [09:10] thats why I didn't know how the xen/libvirt intergration was [09:11] eagles0513875: make up your own mind, if you like xen, use it [09:12] ikonia: i know im just wondering what advantage for instance does kvm have over it [09:12] personal preference [09:12] ok [09:12] read the tech spec pages, they do things differently [09:16] eagles0513875: I find kvm to be much more architecturally sound than Xen. [09:16] eagles0513875: I also prefer its security model. [09:17] eagles0513875: And the fact that it doesn't require a non-Linux kernel to work. [09:17] humm will have to do some research into it :) [09:17] soren: what do you mean by non linux kernel? you mean one that has it compiled into it like xen used to need [09:17] soren, the last point is sort of mood since 3.x ;) [09:18] smb: exactly lol [09:18] smb: No, it's not. [09:18] yes it is soren [09:18] No. It. Is. Not. [09:19] I'm not talking about the patched Linux kernel. [09:19] ahh ok [09:19] but I would agree that it takes longer to get it working. it has a lot of knobs and twiddles all of which one can get wrong [09:19] I'm talking about the Xen hypervisor kernel. Which is entirely different. [09:19] The Xen hypervisor is an operating system kernel. It's based off of a microkernel called Nemesis. [09:19] soren, Ok, that is true [09:20] On top of that, you run your dom0. [09:20] I've got a virtual 10.4LTS Server running. After a reboot (a new kernel had previouly been installed) this morning it's unable to start. It doesn't matter which kernel I select. The only thing I see on a normal boot is a black screen and a _ . If I boot in recovery I can see it detect the disks attached but after running "Begin: Running /scripts/init-bottom ..." and returning "Done." [09:20] the machine freezes and does not seem to be alive. [09:20] And all of your domU's. [09:20] I having problem to bring up eth1 [09:26] auston: Go on.. [09:40] i managed to installed and configured isc-dhcp-server and able to obtain the ip from eth0 but unable to access internet. [09:41] dhcp/internet access are two totally different things [09:41] auston: what gateway and dns servers did you setup in your dhcp server to give to the clients [09:43] the local network ip is 192.168.76.0 gateway 192.168.76.23 [09:43] not what I asked [09:43] re-read what I asked [09:44] eth0 ip: 192.168.76.245, eth1 ip: 192.168.76.5 [09:44] again, not what I asked [09:44] gateway to the router, dns server to local server. [09:45] ok, so test each component [09:45] 1.) can you client do a dns lookup [09:45] your [09:47] managed to get local dns server ip [09:47] no [09:47] what is the point of that [09:48] you can't get to the internet [09:48] can you lookup internet addresses [09:49] the ubuntu server itself is able to access internet [09:49] again, not what I'm asking [09:49] why are you not answering the questions I am asking [09:50] I'm not too sure where to look for [09:51] right, so if you don't know say "I don't know how to do that" rather than just providing random information [09:51] on the client type "nslookup www.google.com" [09:52] ok, i got response from client: **** Can't find address for server... [09:53] ok, so there is your problem [09:53] or at least the start of your problems [09:55] I have setup 2 NICs on the ubuntu dhcp server. Eth0 connected to the router and eth1 connected to client. [09:55] <_ruben> and both in the same subnet at first glance, which tends to be asking for trouble [09:56] agreeed [09:56] But the client managed to get ip from dhcp server but not able to connected to the internet. Only the server itself can do. [09:56] Both server and client configured using the same subnet mask [09:57] auston: DNS !!! [09:58] auston: your client can't resolve addresses, that is reason 1 it can't connect to the internet - fix that [10:00] But can't even ping to local dns server using ip. [10:00] I give in [10:00] pinging a dns server does not mean it can use it as a name resolver [10:00] auston: your client cannot use the DNS server you have set as a resolver, fix that [10:01] Do I need to change the DNS setting from dhcp? [10:01] no [10:02] I'll say it again - the DNS server you have set the client to use, is not responding to public internet zone lookups from your client [10:02] you need to FIX the DNS server you have set to accept connections from your client and do public lookups [10:02] soren: norwegian :) [10:02] RoyK: Ok :) [10:03] auston: Why are both your interfaces on the same subnet? [10:05] As I'm not sure how to do that for both interfaces. Thought it can route the traffic from eth1 to eth0 interface. [10:06] Why would you want to route from one interface to the other if they're connected to the same network? [10:07] Because client conneted to eth1 interface needed to get ip from the server and this server will be acting as a firewall once the internet connection is done. [10:12] * RoyK somewhat guessed soren was danish by the name [10:13] auston: So both interfaces are in fact connected to the same network right now? [10:13] RoyK: I'm not very good at being under cover :) [10:13] hehe [10:27] question, /etc/hosts ... whats the syntax for "cnames" ? are they comma separated or can i put them on a new line ? [10:27] ie [10:28] No commas. [10:28] 192.168.0.7 www.example.com [10:28] Just put multiple names on the same line. [10:28] 192.168.0.7 www.example2.com [10:28] Separated by whitespace [10:28] No. [10:28] 192.168.0.7 www.example.com www.example2.com [10:28] thanks [10:28] Sure. [10:33] is it ok to stick a squid reverse in front of a CalDav server ? [11:12] i have downloaded image http://uec-images.ubuntu.com/oneiric/20111110/oneiric-server-cloudimg-amd64.tar.gz [11:12] what is default username/passwd for this to login ? [11:12] or is it restricted to login via key-pair [11:17] * koolhead11 pokes smoser [11:31] koolhead11, hello [11:31] use your key and log in with ubuntu@ [11:37] koolhead11: smoser is sleeping ;) [11:37] koolhead11: it's normally ubuntu/ubuntu afaik [11:39] koolhead11: if it's juju deployed then just your ssh id will suffice with ubuntu@ as uksysadmin said [11:41] New bug: #897632 in postfix (main) "package postfix 2.8.5-2~build0.11.04 failed to install/upgrade: подпроцесс установлен сценарий post-installation возвратил код ошибки 75" [Undecided,New] https://launchpad.net/bugs/897632 [11:51] Hi. What do I need to do in my firewall or ftp server to get active mode ftp working behind nat? [11:57] hola uksysadmin lynxman [11:58] I'm trying to create my own .deb package of nginx using checkinstall but it don't work, I'm using the sources of 'apt-get source nginx' and downloading the two new modules manually. It configure, using the ./configure with the same parameters of the official package plus new modules, compile (make) but the .deb generated (by the checkinstall) don't work. Anyone knows what I'm doing wrong? [11:59] capeta: try #ubuntu-motu better place with l33ts :D [12:00] lynxman: uksysadmin i remember yesterday able to login the instance with ubuntu/password [12:00] ok [12:01] koolhead11, I've never logged in using a password to the uec images - ubuntu/ubuntu ? [12:01] uksysadmin: i have curious issue with mine, juju needs zookeeper installed as 1st pkg once it gets started/initilized [12:01] then only juju ststus will work [12:02] does bootstrap not do that for you? [12:06] uksysadmin: bootstrap will do only if my network/nova has direct internet access [12:06] which i dont have [12:07] ah - yeah I'm battling with that since I've been playing [12:07] could do with being able to set proxies in the process [12:08] uksysadmin: you mean during bootstarp [12:08] ? [12:08] yeah - is that possible? [12:08] juju bootstrap --proxy= ? [12:08] :D [12:08] * uksysadmin is a n00b [12:08] ;-) [12:09] uksysadmin: me too [12:09] * uksysadmin is attending charm school on Friday [12:09] come to #juju am troubling hazmat there, and talking on same issue [12:09] uksysadmin: me2 :D [12:09] I noticed ... I'm there [12:15] New bug: #855711 in bacula (main) "Bacula director not listening " [Undecided,New] https://launchpad.net/bugs/855711 [12:20] http://uec-images.ubuntu.com/oneiric/20111110/oneiric-server-cloudimg-amd64.tar.gz this is the image i am using [12:21] the euca-get-console-log says this :- [12:22] http://paste.ubuntu.com/753592/ [12:26] lynxman: i can confirm that you cannot book cloud-images with any password, I am building an image from a Oneiric ISO and doing modifications :P [12:27] *log [12:43] koolhead11: aww shucks [12:44] lynxman: am happy, finally learning how to play with bundling my own ubuntu image :D [12:45] hola Daviey :) [13:06] lynxman: i see your tweet :P [13:09] koolhead11: wha? [13:10] lynxman: :D [13:10] beer [13:18] New bug: #897663 in image-store-proxy (universe) "don't connect to imagestore.canonical.com" [Undecided,New] https://launchpad.net/bugs/897663 [13:19] Anyone ever have an issue building cuda on 11.04? [13:22] koolhead11, there is no username and password combination [13:22] there is a user 'ubuntu', but you cannot login via password unless modificatoin is done. [13:22] please read https://help.ubuntu.com/community/UEC/Images#Ubuntu_Cloud_Guest_images_on_Local_Hypervisor_Natty_onward [13:24] lynxman, just for your info [13:24] ^ [13:29] smoser: thanks sir. creating one image for myself because i need to add proxy-info for apt before uploading to bucket [13:30] and will ahve to add cloud-init pkg too inside it [13:30] lynxman: i spent like good time trying hit and trial :P [13:30] koolhead11, so you're just wanting to add apt-proxy config ? [13:30] smoser: oh cool, ty :) [13:31] smoser: yes because my nova is running via proxy [13:31] i'm paticularly interested in this as I'm trying to make that "just work" and use local proxy or mirror if available. [13:31] and when am staring juju-bootstrap, 1st thing it does is tries to install zookeeper [13:32] koolhead11, i wonder, would it be acceptable to you to set up a dns name "ubuntu-proxy" in the default domain ? [13:32] this is not working yet, but that is one way i'm considering allowing that to "just work" [13:32] smoser: i doubt :( [13:32] no? [13:32] i will simply add /etc/apt proxy file [13:32] can you think of some other way that it could "just work" ? [13:33] and add the proxy server detail [13:33] koolhead11, right. and that will work fine. [13:33] smoser: :) [13:33] but you're not the only person that will hit this [13:33] and i want to make it easy for it to "jsut work" without modification of the images. [13:33] smoser: hehe. :D [13:34] so i'm looking for input from you [13:34] why wouldn't you be able to add a dns name ? [13:35] smoser: because that proxy server is my desktop machine running different subnet :( [13:36] btw hazmat has filled a bug on same [13:36] do you have a number? i dont see anything like that [13:36] i really dont think juju should do proxy configuration [13:37] smoser: 1 sec [13:37] smoser: https://launchpad.net/bugs/897645 [13:37] Launchpad bug 897645 in juju "juju should support an apt proxy for private clouds" [Undecided,Confirmed] [13:38] smoser: also wondering why z00keeper is also not installed on cloud image :P [13:39] koolhead11, the most immediate answer is "its not in main" [13:40] the second would be, that it is probably at this point still an unlikely package to be installed. [13:41] smoser: http://paste.ubuntu.com/753662/ [13:41] first thing juju does once i run juju status [13:41] :D [13:41] the only other issue with zookeeper is that it depends on java and thus its probably a 50M additional footprint installed (complete guess) [13:42] koolhead11, can you get the rest of that ... [13:42] i'm interested in "Need to get" and "After this operation, XXXX will be used" [13:43] smoser: pasting it [13:43] wait [13:43] :D [13:44] smoser: http://paste.ubuntu.com/753667/ [13:45] koolhead11, you shorted me... [13:45] there should be a line like "After this operation" ? [13:45] i can figure it out myself easily enoug, though. [13:46] good morning [13:46] smoser: http://paste.ubuntu.com/753480/ euca-get-console-output :P [13:46] hola zul [13:47] hi koolhead11 [13:53] zul: good moaning :) [13:55] koolhead11, i thought i had a bug for this. [13:55] but just opened one, bug 897688 [13:55] Launchpad bug 897688 in cloud-init "cloud-init should support apt-proxy and hostname based mirror selection" [Undecided,New] https://launchpad.net/bugs/897688 [13:57] smoser: apt.conf is what i used simply added proxy server and port [13:57] koolhead11, above, regarding zookeeper, "127 MB of additional disk space" [13:57] smoser: 127 0.o [13:57] so we'd grow the images from ~ 600M to 725, which is a 20% growth [13:57] yeah. [13:57] smoser: better leave it [13:57] so its not an easy thing. [13:57] let juju handle it :P [13:58] smoser: the cloud image is 205 MB i think :P [13:58] 205 compressed download. [13:58] 600~ filled filesystem. [13:58] smoser: o00h [13:59] the disk image you get is a qcow compressed sparse disk === bladernr_afk is now known as bladernr_ [13:59] cloud-publish-image takes good time :P [13:59] the cloud-init 'make-iso' stuff that i pointed you at is the start at making scripted modifications to the images easy to do. [13:59] ie, you could easily [14:00] * download image [14:00] * boot image with cdrom attached with userdata that did: [14:00] * install zookeeper and other juju stuff [14:00] * change local mirror [14:00] * shutdown [14:00] * publish modified image [14:01] i guess other things to do there would be something like: 'apt-get clean' and some other "clean up my mess" stuff, but the point being local modification is easy without having to even be root. [14:01] smoser: i used KVM and did exactly same, am now at last step :D [14:01] you did it via user-data ? [14:02] juju could easily publish doc on how to do that before uploading your own cloud image, and could do it all via one program wrapper. [14:03] smoser: i booted the machine via KVM and then used VNC viewer to connect it and do needed modification [14:05] yeah... i just don't like having to have you interact with it. [14:05] as then its un-reproducible [14:05] clearly here its fine, but a fully reproducible solution is much nicer. [14:06] thanks for your input by the way. this has been helpful. [14:07] !user-data [14:07] smoser: would you please point me to the user-data doc/method doc i should look into [14:08] https://help.ubuntu.com/community/UEC/Images#Ubuntu_Cloud_Guest_images_on_Local_Hypervisor_Natty_onward [14:08] you can probably figure it out from there. [14:08] do the bzr brandch of cloud-init as shown, then look at 'user-data' in cloud-init/doc/ovf [14:08] it is just cloud-init userdata. [14:09] just put a '#!/bin/sh' script there that does everything you want, then cleans up, then shuts down. [14:10] smoser: thanks again. [14:10] if you have questions, please let me know. this is somethign that "should work" but is very admitteldly not documented well. [14:33] Help please ~ [14:34] smoser: one help. i have my custom image booted, but am not getting anything from euca-get-console-output useful, as i got from the cloud image [14:34] ? [14:34] is there specific package i need to add in the image to get that [14:34] how did you make the custom image ? [14:34] just by modifying the original? [14:37] smoser: yes and adding proxy information [14:37] you should get console output still. [14:38] i suspect somethign went wrong and it did not start. [14:38] How to see local IP in Virtual Server [14:38] ? [14:42] smoser: i get instance ID and time i-0000002c 2011-11-29T14:40:37Z [14:42] :( [14:42] Xelmep: ? [14:42] i suspect somethign awry in openstack [14:43] RoyK: i vant to configure VPN server [14:43] smoser: but when am booting yours cloud image i get the console output :P [14:43] byt i don`t know how to see local IP [14:43] Xelmep: what vpn server? what hypervisor? what sort of system? [14:44] Ubuntu server [14:44] !vpn [14:44] For more information on vpn please refer to https://wiki.ubuntu.com/VPN [14:48] koolhead11, right. but nothing you did shoudl have changed that :) [14:48] oh... [14:48] did you upgrade grub by chance? [14:51] smoser: no. :P [14:52] let me upload original image and see if i get console output [14:52] koolhead11, i'm almost out of ideas. do you have access to the nodes ? [14:52] yes [14:53] manually added 100 mb zookeeper and pkg, still juju status giving same error === nande is now known as nandemonai [15:41] Where can I find a guide of dir changes from 8.04 to11.10? [15:43] who is running the meeting today? [15:43] It seems that apache config files have moved and I cant get drupal happy [15:43] ok not me [15:46] My IT department runts its own Dropbox server (where one can drop and retrieve files from). Does anyone know what software I could use to run my own? [15:52] * lynxman looks at the channel topic [15:52] * koolhead11 plans to go home [15:56] smoser: no Actions from last week? [15:57] i didn't think i did last week. [15:57] i though tyou did. [15:57] smoser: my bad, it was jamespage [15:58] jamespage: no actions from last week? [15:58] Daviey: nope [15:59] great :) [16:00] can someone help me with how to disable default server responses in apache2? [16:00] default and default-ssl are disabled, and all my sites-enabled have explicit servername directives... [16:01] other than that, vanilla installation === shadeslayer_ is now known as shadeslayer [16:24] zul: let me know when you want me to test keystone thing [16:31] high everybody [16:31] anyone can help on openldap with lucid? [16:31] i'm following the server guide [16:33] but there is a pb when creating the backend, error: "ldap_add: Other (e.g., implementation specific) error (80) [16:33] additional info: handler exited with 1" [16:34] well ggling led me to consider i'm loading the module twice [16:35] but when adding a frontend with a test user admin ldap pass is not recognized [16:38] New bug: #855030 in nova "Encountering sporadic AMQPChannelException" [Critical,Fix committed] https://launchpad.net/bugs/855030 [16:45] debug log: [16:45] slap_listener_activate(8): [16:45] >>> slap_listener(ldap:///) [16:45] connection_get(10): got connid=1000 [16:45] connection_read(10): checking for input on id=1000 [16:45] ber_get_next [16:45] ber_get_next: tag 0x30 len 34 contents: [16:45] op tag 0x60, time 1322584896 [16:45] ber_get_next [16:45] conn=1000 op=0 do_bind [16:45] ber_scanf fmt ({imt) ber: [16:45] ber_scanf fmt (m}) ber: [16:45] >>> dnPrettyNormal: [16:45] conn=1000 op=0 do_bind: invalid dn (cn=admin,mydomain) [16:45] send_ldap_result: conn=1000 op=0 p=3 [16:45] send_ldap_response: msgid=1 tag=97 err=34 [16:45] ber_flush2: 24 bytes to sd 10 [16:45] connection_get(10): got connid=1000 [16:45] connection_read(10): checking for input on id=1000 [16:45] ber_get_next [16:46] ber_get_next on fd 10 failed errno=0 (Success) [16:46] connection_close: conn=1000 sd=10 [16:46] how do I get the pretty server information screen on a non ubuntu-server system? What is the package called that shows ip, dish space etc. on the ubuntu-server command line? [16:46] !pastebin | tubuntu [16:46] tubuntu: For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imagebin.org/?page=add | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [16:47] ubottu: ok [16:47] You're welcome! But keep in mind I'm just a bot ;-) [16:49] slapd log: http://paste.ubuntu.com/753823/ [16:50] why dn is invalid? is it because the admin user is not added yet? === Ursinha is now known as Ursinha-lunch [16:55] Daviey: sorry, no progress on keystone mir. doko had some comments to say about it though. I hope to get to my part this week [16:59] how do I make cron test if a script is executable before running it? [17:00] jdstrand: thanks [17:01] Weisse: cron runs each line through a shell.. so [ ! -x /usr/bin/foo ] || /usr/bin/foo [17:07] jamespage: not seeing anything like the errors here, http://pb.daviey.com/5yon/ on precise preseeded installs? [17:08] zul: did you have a cobbler bracnh around that has new methods for storing hw inventory related things? [17:08] adam_g: i have an untested patch [17:09] zul: can i peak? [17:09] sure [17:09] Daviey: hrm - I've not done one today [17:09] other than the standard ISO tests which kinda preseed [17:10] adam_g: i was going to add support for it but: http://people.canonical.com/~chucks/cobbler/ [17:10] I don't think I'll ever make it to a server team meeting at 1600 UTC when we're on standard time. 0800 - 0830 is the heart of chaos in this house. :-P [17:11] zul: ah, cool thanks [17:11] adam_g: if you want to do it thats ok as well [17:12] zul: i might. i was going to start doing the client side stuff, wanted to get an idea of what needed to get added to the server for each bit. [17:12] Daviey: ping [17:12] RoAkSoAx: Hey sailor [17:12] Daviey: so I foudn this: http://pastebin.ubuntu.com/753850/ [17:12] adam_g: if you want to do the client side stuff i can do the server side stuff. i just a list of flags needed to add [17:12] Daviey: but been testing it, and doesn't seem to be working [17:13] RoAkSoAx: oh dude [17:13] That is amazing. Where did you find that? [17:13] Daviey: http://comments.gmane.org/gmane.network.etherboot.gpxe/1600 [17:13] http://lists.ipxe.org/pipermail/ipxe-devel/2011-April/000592.html [17:13] Great! [17:13] Daviey RoAkSoAx: there is a patch floating around that adds better gpxe support to cobbler than what we have [17:14] Daviey: but doesn't seem to be working though, been trying to manually pass arguments and stuff but shows syntax error [17:14] zul: dig it out if you have it to hand.. note we don't care at all about gpxe. [17:14] RoAkSoAx: interesting.. So i tried grub-ipxe, and i got a syntax error from a default install.. so something is wonky [17:15] We need to fix that regardless [17:15] Daviey: that's weird, I didn't get syntax error on default install [17:15] I wonder if my dhcp was pushing out bad config then? [17:15] maybe [17:18] Daviey: maybe [17:19] RoAkSoAx: if we can get this working, i am so overjoyed you would not believe [17:20] lol [17:38] adam_g: do you have a list that you are gong to send to the server? [17:39] Daviey: hrm. looks like you've got a pxe-kexec sponsor that's already uploaded the package to Debian... [17:41] Daviey: ...so you're in the NEW queue already [17:43] zul: not yet, no [17:43] adam_g: k cool send me the list when you do [17:45] ahs3: yeah \o/ [17:46] ahs3: I hope you didn't spend time looking it over? [17:46] ahs3: long as we're talking sponsors :) any more thoughts on netcf? [17:55] Daviey: fortunately, not yet. so cool. you're on your way. [17:56] hallyn: working on the upload right this minute, actually.... [18:01] ahs3: That is great! I did try to ping you a few times, but assumed you were away [18:01] thanks! [18:06] Daviey, so this is moderately bad. [18:06] where moderately is overly nice [18:07] do we have any option to other than release noting "cloud images do not boot in recommended cloud or on recommended hypervisor" [18:07] bug 897795 [18:07] Launchpad bug 897795 in linux "-virtual kernel missing rtl8139 drivers" [High,In progress] https://launchpad.net/bugs/897795 [18:08] I am having a problem setting up Ubuntu Server 11.10 x64 on an HP xw4400 workstation. Intel C2D, 3GB DDR3, Ati Radeon 5450, Corsair SSD. It installs perfectly, but won't boot after the install. It acts like it going to boot, but then the screen goes to sleep. When I try the recovery mode it goes to the screen, but it's like my keyboard is frozen, I can't do anything. Anyone have any idea [18:08] what could cause this? I haven't had much trouble with Ubuntu Server before. [18:08] oops, I meant DDR2 [18:10] smoser: you are right that is kind of bad...and kind being on the high scale [18:11] smoser: To put into context, this is Alpha 1 right? [18:11] yeah, people dont expect that to boot. [18:12] but silly me did. [18:12] is there any potential of getting that fixed though? [18:12] smoser: Would it make sense to push for a fix when A1 lands, and we defer cloud image release? === Ursinha-lunch is now known as Ursinha [18:13] smoser: I'm really not sure it's worth asking for a kernel respin for A1 at this stage. [18:13] Daviey, smoser: we already have to do a respin of the images due to bug 897680 [18:13] Launchpad bug 897680 in ubiquity "Precise Desktop 64Bit: libc6 fails to install if "install 3rd party software" is selected" [Critical,Fix released] https://launchpad.net/bugs/897680 [18:14] utlemming, wrong bug number ? what do i care about desktop install path for cloud images. [18:14] utlemming: erm, i mean a kernel build vs a respin [18:14] Daviey, personally i dont tihnk its unreasonable to ask for a respin. [18:14] https://launchpad.net/ubuntu/+source/linux [18:14] shows that a new kernel hit the archive 23 hours ago. [18:14] i found it didn't boot 1 hour ago. [18:14] ffs [18:14] thats pretty good [18:20] smoser: we're installing the bad version of libc6 -- I matched the bad version of libc6 -- but the replication step comment #5 fails to match the cloud-images [18:20] so no repsin needed [18:22] wait. surely we do not have libc-bin:i386 installed. [18:22] in an amd64 image [18:25] noaXess, we don't [18:25] damn auto correct -- n.o. we don't [18:26] Daviey: what was that django app that i use as an example for dashboard? [18:29] zul: one example: [18:29] mumble-django - Mumble-Server web interface [18:29] python-django-mumble - Mumble-Server config application for Django [18:29] Daviey: thanks [18:35] dashboard :) [18:40] http://security.ubuntu.com/ubuntu/pool/universe/i/icinga/ <-- Hey, a package I want! What do I need to have in sources.list to reflect this? [18:44] I need help setting up RAID on ubuntu server 11.10 [18:44] it is showing my fake raid drive, but it is displaying it as read only [18:52] Mhaddog: why bother with fakeraid? [18:53] Corey: you need to enable the "universe" repository [18:54] i eas doing soft RAID anf got no where... [18:55] actually I wanted soft raid... but apparently ubunut is not even seeying my drives... [18:57] Mhaddog: start by entering your computer's bios, removing any raid config that may be present, and then disabling fakeraid, foreva [18:57] Never mind about my booting problem, hitting Ctrl-Alt-F1 fixed it [18:57] Mhaddog: then install using sofware raid [18:58] I didi try that, I setup the sata as regular and got nothing, will ahci mode work? i didnt try it [18:58] let me reset it again and set it to ahci [18:58] Mhaddog: are you sure you ① removed configuration AND ② disabled it ? [18:59] remove configuration, u mean kill the raid arrangement?? [18:59] Mhaddog: of course [18:59] done both.... [19:00] ok I got 3 options, Ide mode, AHCI mode and RAID mode, what u suggest between IDE and AHCI ? [19:00] Mhaddog: then what happens during the install, does the installer see all disks? [19:00] I will let you know [19:02] ok, review in ahci mode and raid disable [19:02] going into the installer... [19:02] @ installer [19:02] let me get to the parition part [19:07] anyone here ever setup MPI? on ubuntu? [19:08] MPI? [19:11] I'm on the partition disks [19:11] got them..... [19:11] finally... [19:11] Mhaddog: good work [19:11] ok so I setup a soft raid.... [19:11] now [19:11] in a manner of speaking, yes [19:11] swap at teh begining correct? [19:11] anywhere will do [19:12] what level of raid are you thinking of? [19:12] 5 [19:12] how many disks do you have? [19:12] ok it said that I only have one partition.... [19:12] 3 [19:12] alright [19:12] I havent setup a raid 5, I always do a raid 1 or 0 in centos... [19:12] so, I'm a little lost in this installer.... [19:13] do I need to parition them first before add them to the raid MD device? [19:13] no [19:14] ok so it is only seeing one disk [19:14] you said it saw all 3 [19:14] Mhaddog - are you sure you want to do software RAID 5, it may slow things down and if one disks fail, another could fail during the rebuild in which case you lose everything :/ [19:14] i need the redundancy, [19:15] it is not a database server, just will be a file sharing/data dump one [19:15] hmm, RAID 1 might be better, RAID 6 if you have another disk or just copy everything off daily to another drive [19:16] I though of raid 6, but disks are expensinve at this time, so I got two new ones and an a old boxed one.... [19:17] idd, are you sure you need RAID, are you really changing that much that often? Would a nightly manual backup be enough, that will also protect you against accidental deletion and make recovery easier if something goes wrong (no need to try and mount a sw RAID array in recovery mode) [19:18] yes, about 5K audio files per day [19:18] i will put a 1tb drive as an internal backup.... when I can get it out of my asterisk box [19:19] fair enough sounds reasonable [19:21] we are talking about a good 15 to 20 gb per day more or less [19:21] so I think raid 5 is the most convenient one... [19:22] a little more speed than plain mirroring [19:22] and a little added redundancy [19:23] I think writes will be slower because of parity calculations in sw but reads will be faster [19:23] ok this is strange, unles I make a parititon on the dirve, it does not let me select it for the RAID group (md device) [19:24] Daviey: so just to be clear, you did look at the etherboot+ipxe packages, and saw no problem? I should just uplaod them as soon as soft freeze ends? [19:25] Mhaddog - this might help (it gives you the general idea of how it works) : http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/3/html/System_Administration_Guide/ch-software-raid.html [19:27] thx I know that one.... I think the fake raid left paritions in there... I'm dd on teh drives... [19:27] I cannot create ext partitions.... [19:27] thx for your help [19:27] hallyn: I did have a look, but please let me have another [19:27] OTP right now [19:29] np :) [19:32] Daviey: ok, no hurry (until monday :) just checking to make sure i understood right. (which i didn't :) [19:33] hallyn: I only sniffed it, not confident enough to upload it myself :) [19:33] hallyn: but hey, if you did it - i'm sure it's great [19:34] Daviey: you weren't watching cjwatson's msgs to me yesterday apparently. [19:35] hallyn: we live and learn, right? :) [19:36] If that is the worst thing you've ever done, you are in a better situation than me :) [19:36] Daviey: so we hope. If I could make headway in these bugs i'm looking at, i'd spend somet time looking for dput wrappers to automatically show me debdiff against current archive version, etc. [19:37] hallyn: So i don't sign my uploads until i am happy with them to be uploaded. [19:38] hallyn: mkdir tmp ; cd tmp ; pull-lp-souce $PACKAGE ; cd .. ; debdiff tmp/*.dsc *.dsc | less [19:40] Daviey: yeah, i'll write my own if i want to, but i recall discussions from people about the wrappers they'd already written, and i figure they're checking for mistakes i haven't even dreamed of yet. [19:41] Daviey: I *was* happy with it :) problem was my changelog convinced me i was in the right tree, but i hadn't freshly downloaded one. which is why a script that always fetches it fresh and does a debdiff is in my future. [19:42] cool [20:10] I got lost [20:16] Daviey: i can see gluster pkg for 11.10 http://download.gluster.com/pub/gluster/glusterfs/LATEST/Ubuntu/11.10/ [20:18] koolhead17: right, it's been in Ubuntu for ages, including a hardy backport. [20:19] so when installing gluster is said it means apt-get install gluster [20:19] got it [20:23] koolhead17: glusterfs is in the ubuntu archives [20:24] probably not the latest and greatest though [20:26] * koolhead17 trying to find out who is doing pkging for gluster at http://download.gluster.com/pub/gluster/glusterfs/LATEST/Ubuntu/11.10/ [20:26] hello [20:26] i m on TTY mode on ubuntu server [20:27] ubuntu lucid [20:27] how to select ? [20:29] anyone there ? === WaVeR` is now known as WaVeR [20:49] I need help about counter strike server [20:51] Xelmep: what sort of help? [20:54] nice - MPI running nicely - soon 136 cores in a cluster... [20:56] Pici: I WANT IN MY VPS TO Install and run cs server [20:56] can you help me how to install and run [20:56] then to connect slients from other pc-s [20:56] Xelmep: go to the control panel, click on the Install-and-run CS Server button. Done [20:57] i have access only with SSH and Webmin [20:58] jMCg: i have access only with SSH and Webmin [21:00] Xelmep: the linux steps here should help, as well as the external links: https://developer.valvesoftware.com/wiki/Dedicated_server [21:02] thanks Pici [21:02] i will try to do it now [21:02] tell me after install i will need to create stages to server [21:02] ? [21:03] or i will need to connect then to create [21:03] ? [21:03] Xelmep: stages? [21:05] yes [21:06] stasis [21:06] Xelmep: I don't know what you mean. [21:07] pici: if you want to play a game you need to create game and select stasis [21:07] understand me [21:07] ? [21:09] Xelmep: If you are asking how to administrate the game once you have gotten it running, then you'll need to look at Valve's documentation. [21:39] jdstrand: i uploaded a fix python-keystoneclient [21:45] zul: ack [21:48] robbiew: ping http://lists.linuxfoundation.org/pipermail/virtualization/2011-November/018995.html [21:49] Daviey: ^^^ [21:51] zul: interesting [21:52] so they have it for A15 (which I don't care about atm)...but also ARMv7? [21:53] robbiew: i dunno maybe its a project for someone to try [21:53] * robbiew posts to #ubuntu-arm [21:55] zul: nice find [21:57] Daviey: rbasak is back next week...maybe something we can throw at him [21:58] keep him busy ;) [22:00] robbiew: works for me :) [22:01] robbiew: You'll provide some ARM hardware with virt extensions? [22:01] absolutely....top of my list [22:01] lol [22:05] I'm having some issues setting up a mail server, is this the right channel to look for help? [22:05] chudapati09_: yes! [22:06] zul: are you familiar with http://xen.org/products/xen_arm.html [22:06] Samsung's efforts? [22:06] robbiew: no im not [22:06] Daviey: okay wonderfully. I'm building a web server to host multiple websites. I got everything working except the mail server. I'm using linode to host my vps and I've been following their tutorials. Here is the link I've been following, http://library.linode.com/email/postfix/dovecot-mysql-ubuntu-10.04-lucid. I got up to the part where it tells me to restart dovecot, so I tried "service dovecot restart". But then I ge [22:07] chudapati09_: you were cut off after "But then i ge" [22:07] chudapati09_: try, sudo /etc/init.d/dovecot restart [22:08] Daviey: But then I get this "restart: Unknown instance:". i'm doing all this as root. I tried to start dovecot and it worked. but then I did "service dovecot status", and I get this "dovecot stop/waiting" [22:10] Daviey: when I try to get the status for it I get "dovecot stop/waiting" [22:10] chudapati09_: then try 'start' :) [22:10] rather than restart [22:11] Daviey: Same thing [22:11] happens. [22:16] zul: are you around? [22:16] zul: seems i don't have upload perms to qemu-kvm-spice. Would you mind pushing http://people.canonical.com/~serge/qemu-spice.debdiff ? [22:16] hallyn: almost eoding [22:17] ok [22:17] RoAkSoAx, ^ ? pretty please? :) [22:17] hallyn: sure [22:18] RoAkSoAx, thanks! [22:21] chudapati09_: did you try /etc/init.d/dovecot restart ? [22:21] Hi, I'd like to have the server stats that are displayed in my ubuntu server box on log in, also displayed in my desktop box. Am I able to just copy /etc/motd/ contents to the desktop from the server? [22:23] ewook: I get this message, "dovecot start/running, process 4664", I'm assuming that's a good sign. But when I do "/etc/init.d/dovecot status", I get "dovecot stop/waiting" [22:23] oh I see it's a symbolic link.. [22:36] hallyn: done [22:37] RoAkSoAx, thx [22:37] welcome ;) [23:03] I've got a master/spare db system where the master is contacted on a fixed ip address that's brought up on whatever machine is the master at that point. Last night, 1/2 hour after switching the ip addresses, I got an arpwatch notification that it had gone back. There were (new) established connections to the master ip, but the ip wasn't in ifconfig. I've seen it happen once more since then. I'm really not sure why I'm getting the arps are happeni [23:04] the (old) master is 8.04, spare is 10.04 [23:08] your question got cut off on " the arps are happeni" [23:08] I'm really not sure why I'm getting the arps are happening, or why the connections are being accepted on the spare. [23:08] what do you use for migrating the ip addresses etc? [23:09] ifup, ifdown, this time manually [23:09] do you use a cluster suite with RAs (like pacemaker) [23:09] nothing automatic [23:10] so ip addr show does NOT show the address on the slavenode BUT there ARE active connections to that address on the slave node? [23:11] yes, for slave == old master [23:11] as reported by netstat -an. [23:12] could you please verify with "ip addr show"? [23:12] http://pastie.org/2941346 [23:13] http://pastie.org/2941352 [23:13] that one with ip addr show [23:14] is ifconfig lying? [23:15] could you also do a "ip link show" please [23:15] http://pastie.org/2941366 [23:17] i meant, ip link show up [23:17] sorry [23:18] http://pastie.org/2941371 [23:19] well everything should be fine, but to be sure you could try deleting the address from the interface completely [23:19] but I agree there should'nt be established connections to that address on that host [23:19] shouldn'T [23:20] ifconqq [23:22] I've never seen this before. [23:22] it's not wanting to delete the address, it's just silently failing [23:23] how do you try to delete the address? [23:23] sudo ifconfig eth1 del 10.0.0.25 netmask 255.255.255.0 [23:24] with and w/o the netmask [23:25] try sudo ip addr del 10.0.0.25/24 dev eth1 [23:25] maybe this is more verbose [23:26] is ip a replacement for ifconfig? [23:26] that does seem to work [23:26] ifconfig is "kind of obsolete" in the sense that it uses an obsolete kernel interface to the networking code [23:27] (it uses interface aliases eth0:1 etc. for "emulating" multiple ip addresses etc.) [23:27] ah [23:27] iproute2 (that's the package ip comes with) is more modern and has way more options [23:28] (a lot of it is advanced stuff a lot of people never have to deal with like policy routing) [23:28] wiredfool1: did this kill the "ghost" sessions? [23:29] I actually killed them off a while ago, I'm just trying to figure out why right now so it doesn't happen again just after I go to bed tonight. [23:30] well by deleting the address this should ensure that. but it would be interesting why you had connections running over an interface that's in down state [23:31] maybe it's related to ifconfig not working correctly. [23:31] I don't know. The config on _this_ machine has been stable for O(years) [23:32] the other of the pair was upgraded to lucid last week [23:32] this was the first failover since then [23:33] only thing I can think is that there's some kvm stuff in there that's not being used [23:33] but it's been that way for a while [23:35] but it happened with the virbridge disabled as well [23:36] robbiew, jinkeys, https://blueprints.launchpad.net/ubuntu/+spec/servercloud-p-libvirt is still not showing up in status.ubuntu.com? [23:41] I like that even though ubuntu-server can be run as a hypervisor with no need for a desktop, but i also like that I can install a light weight desktop environment such as xfce4 [23:41] so i can manage my server in a gui as well as a command line...its the only type 1 hypervisor I know of besides kvm that can be managed that way [23:48] is there any management gui for the hypervisor in ubnuntu that works on windows? I have my admin station running windows do to some requirements that my outfit has [23:59] hallyn / ahs3: i spy netcf in Debian NEW queue [23:59] \o/