[00:00] Firstly, I'm having trouble booting ubuntu cloud images in kvm/libvirt. I realize the default console is set to ttyS0. Even if I change this to tty0 etc., I don't get a getty prompt to log in. [00:01] How do I log into an ubuntu cloud image on kvm? [00:13] There isn't an sshd listening by default either. [00:18] heh... after all the datasourceec2.py warnings looking for a metadata service, it finally continues to boot and gets to a getty. [01:34] i want to put a gnome desktop on my ubuntu server and use it remotely [01:34] how can i do this? [01:36] install gnome? [01:36] no [01:37] i mean how would i use the desktop remotely? [01:37] to open software and things like that [01:37] instead of SSH, VNC to it [01:38] yes thats what i mean [01:40] vnc or nx or ........ [01:42] ok vnc does just this? [01:42] allows me to remotely control a desktop? [01:42] blenderman: they are different things.. i would just try them both and see what works for you === acrocity_ is now known as acrocity [01:43] ok now what's the advantage of using ubuntu server [01:43] instead of controling an ubuntu distro with gnome with vnc [02:02] cuz command line is better for a server operating system. less overhead, less packages [02:06] is there something about command line that enables a server to run constantly [02:06] is it had gnome would you not be able to let it run all the time [02:09] blenderman: ? [02:10] it? you mean, the computer? you can leave it on if you want.. you just leave it one.. the DE or lack of DE has nothing to do with it [02:11] i guess my question is [02:11] why are servers left to run all the time [02:11] while we're supposed to turn desktops off [02:12] at night [02:23] blenderman: you arent "supposed to".. but you can.. if you dont need to leave it on [02:23] folks leave servers on cause they need to be on, and available [02:26] Hello. I have a subversion setup served over an Apache host with a self-signed certificate. It was working fine until today (possibly after applying the latest updates.) Now I get this error upon using the command line client: "SSL handshake failed: Secure connection truncated" [02:26] any ideas? [02:30] violinappren: openssl s_client -connect server:port usually shows more useful error messages than other things [02:30] violinappren: also check the ssl server log === Rallias is now known as Rallias-Plugged [02:37] SpamapS: I get the following error in the logs: user root not found: /svn/project1/trunk [02:38] accessing the URL in the browser actually works, prompts me for HTTP auth over ssl [02:51] its a client-side issue, i installed libneon and symlinked libneon-gnutls.so.27 -> libneon.so.27 [02:51] works fine now === cruzer45 is now known as cruzer45_ === cruzer45_ is now known as cruzer45 [03:17] violinappren: wacky; was the client busted in an upgrade at the same time? or...? [03:18] sarnold: yes, both are precise machines [03:19] sarnold: seems gnutls packages are all going haywire, i get this error now while trying to git clone something from github: error: gnutls_handshake() failed: A TLS packet with unexpected length was received. [03:19] [03:21] violinappren: the recent openssl security update busted some specific modes when run on CPUs with AES-NI instructions. The security fix was reverted, and we waited for OpenSSL team to deploy _new_ fixes, and after a few days in the -proposed queue, we shipped those. [03:21] violinappren: I think I heard Debian's update might have missed the fix. It might be worth looking into it to see if you can find out specifically which combinations break :( [03:23] its a disaster that openssl packages aren't getting enough quality control even on an LTS release [03:25] violinappren: it's a pain to track down when it only trips bugs in one odd mode on specific CPUs :( [03:25] violinappren: too true, some more comprehensive tests would be keen. === Rallias-Plugged is now known as Rallias === security is now known as fire [08:43] bug 1160893 [08:43] Launchpad bug 1160893 in mongodb "Unchecked access to SpiderMonkey’s JavaScript nativeHelper function" [High,Triaged] https://launchpad.net/bugs/1160893 === histo_ is now known as histo [09:27] hi guys, I've an ubuntu server openvz container, which gets the ip from dhcp [09:27] but on boot it doesn't have the default route, if I do dhclient eth0 it has the route [09:35] seems an issue with 12.10, 12.04 works fine [09:39] but running ifdown/ifup works fine === Nigel_ is now known as G [10:22] Hi all, any canonical folks here? [10:24] spurious, yes [10:25] Hi jamespage, is there a particular channel for AWS specific questions? [10:26] spurious, here is fine [10:28] I'm having trouble getting hold of the ubuntu AMIs through the AWS API. When I query using the Canonical ID provided on help.ubuntu.com I get a huge list of AMIs, not the 32 listed on the AWS marketplace, and those listed in the marketplace are not in the list [10:29] When I query with the specific AMI ID however, I do get the correct AMI, but there is no Canonical ID in the metadata [10:30] Hence, I guess, it not showing up in the query [10:30] Is there a preferred method for getting the Ubuntu AMIs? [10:31] And also what are the ~350 AMIs that do show up under the Canonical ID [10:38] jamespage, any idea on how I should proceed? [10:39] spurious, yes - try https://cloud-images.ubuntu.com [10:39] its organized by ubuntu release and will give you the correct AMI's for every ec2 region for the official ubuntu AMI's [10:40] spurious, https://cloud-images.ubuntu.com/releases/12.04.2/release/ for example [10:41] spurious, the data on site is is also machine consumable [10:42] jamespage, is there a specific nomenclature for the AMI metadata? I am sourcing AMIs over the AWS API througgh image repository queries [10:43] spurious, the cloud-utils package for ubuntu contains a 'ubuntu-cloudimg-query' tool which you can use to query this data [10:44] spurious, the machine readable data is here - https://cloud-images.ubuntu.com/query/ [10:44] jamespage, unfortunately I'm using the AWS Java SDK in an enterprise product running on RHEL, so I'm not sure I can use the cloud-ing query tool [10:45] spurious, there is more information on the cloud image query data here - https://help.ubuntu.com/community/UEC/Images [10:46] utlemming, smoser: do we have any cloud image tooling for folks not running ubuntu? === acrocity_ is now known as acrocity [10:47] spurious, actually thats just a bash script so you can probably run it on anything with bash [10:49] jamespage, one of my dev test boxes is an Ubuntu machine, if I pull down the cloud-utils package from the Canonical repo can I pull out the query sh and run it independently? [10:49] I think you probably can! [10:50] spurious, or you could grab it from here - http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/raring/cloud-utils/raring/files/head:/bin/ [10:50] jamespage, thanks, I'll give it a go. Do you know anything about the original question of why the 32 official AMIs on the Canonical marketplace page don't show up on the provider ID query? [10:51] spurious, unfortunately no I don't - utlemming might be able to help with that [10:54] jamespage, thanks! utlemming, any input you could offer would be greatly appreciated [10:55] spurious, np - utlemming is around ~3 hours from now probably (central US) [10:56] jamespage, that's fine, thanks. I'll leave this open and contact him around then. Cheers. [10:56] (or her, shouldn't jump to conclusions) === irv_ is now known as irv === EntropyWorks_ is now known as EntropyWorks [11:27] jamespage, do you happen to know if the query script is redistributable? [11:28] spurious, its GPL-3 so yes === _KaszpiR__ is now known as _KaszpiR_ [11:28] spurious, its part of https://code.launchpad.net/cloud-utils [11:28] jamespage, great, thanks === luminous_ is now known as luminous [12:04] spurious, so, you raise an interesting question about the "marketplace" amis. [12:04] we do not have machine formated data that gets you those ami-ids available. [12:04] utlemming can explain more what the difference between the non-marketplace and the marketplace amis are. [12:06] they are bit for bit equivalent to the non-marketplace ami. [12:13] spurious, i personally would appreciate it if you'd put your interest in using ubuntu-cloudimg-query on RH to https://bugzilla.redhat.com/show_bug.cgi?id=907756 (see comment 12) [12:15] Hi smoser, it seems that the marketplace amis are solely for use through the AWS marketplace GUI. [12:15] i didn't want to say that. [12:15] The query script, however, works fantastically [12:16] and i'm not certain that there is not a benefit to canonical if you chose to use the marketplace amis. === xcrracer_ is now known as xcrracer [12:16] it may also imply some sort of suport, i'm not sure. but if you wanted support and you're using the official amis not listed in the marketplace, you can get it. [12:17] if they're bit-for-bit equiv it doesn't really matter that they are different, as long as they are overall queriable for users of the API [12:17] ah, I see what you mean, and indeed the amis in the marketplace do come in 4 versions, normal and then 3 llevels of support [12:18] which is fair enough, that's where the main business is [12:19] somser, the status of that request on rhel is CLOSED ERRATA [12:19] oops, smoser, sorry [12:19] spurious, yeah, as in its "fixed" they've included it (without the ubuntu tools) [12:19] but its fine to comment there. [12:19] just maybe no one will listen :) [12:20] ok, will do. Would love to see this on rhel, a fantastic tool. [12:20] spurious, fwiw, the data that it queries is https://maas.ubuntu.com/images/query/ [12:20] and we're working on making improved data available. [12:21] oops. wrong url [12:21] ubuntu-cloudimg-query hits [12:21] [DIR] raring/ 27-Oct-2012 17:57 - Ubuntu Server 13.04 (Raring Ringtail) daily builds [Current Development Version] [12:21] [TXT] released.latest.txt 25-Mar-2013 19:37 279 [12:21] GAH. [12:21] http://cloud-images.ubuntu.com/query/ [12:21] ^ that one. [12:22] and the new data format is available at http://cloud-images.ubuntu.com/eightprotons/ , but the location will likely move sometime soon. [12:22] http://cloud-images.ubuntu.com/eightprotons/streams/v1/com.ubuntu.cloud:released:aws.js is what you're interested in (aws data). [12:23] smoser, looks good, thanks! [12:25] utlemming, we could put simplestream data for the aws marketplace ids also. [12:25] at something like 'com.ubuntu.cloud:released:aws-marketplace.js' [12:29] smoser: jamick http://paste.ubuntu.com/5655036/ [12:29] this GPG error am getting for PPA/testing/grizzly [12:29] installing ubuntu-cloud-keysring [12:29] was of no help [12:31] Daviey, ^ [12:33] smoser: hello btw :P === chmac7 is now known as chmac [12:51] . [12:51] jamespage: ping [12:51] I cant seem to start my mysqlserver [12:52] when i try to connect to mysql: ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2) [12:56] var/run/mysqld/mysqld.sock doesnt exist [12:58] Hello Guys. i have problems with my ubuntu server. since a few days it is getting a very huge amount of "connections through firewall" and "ipconntrack". any idea how i can see whats happening? [12:59] Styler2go, sure, check the conntrack table [12:59] can you tell me how i can do that? [13:00] http://styler2go.de/munin/localdomain/localhost.localdomain/fw_forwarded_local-week.png this is the graph where i see the amount... [13:00] cat /proc/net/ip_conntrack [13:01] and what can i do with this table...? i mean.. is it ok that the amount is that high? [13:01] in that case, no idea [13:01] cause that just counts packets [13:01] tcpdump? [13:02] http://styler2go.de/munin/localdomain/localhost.localdomain/fw_forwarded_local.html http://styler2go.de/munin/localdomain/localhost.localdomain/fw_conntrack.html [13:02] how should we know what is high or low, it's your system, and only you know how it's setup and what is connected to it, and what is expected [13:03] i cant explain this huge amount... [13:03] they are all local connections, so they will show up in netstat [13:03] maybe your part of the *largest ddos in history* [13:04] lol [13:04] my server is running as great as everytime [13:04] and that means? [13:04] i dont think its an "attack"? [13:04] just cause your server is running fine, doesn't mean it isn't attacking someone :) [13:05] I didn't say you where getting attacked [13:05] hmm [13:05] how can i defend me? [13:05] and I didn't really mean it seriously [13:05] just the timing of it, with the timing of the latest ddos [13:05] i had some ddos attacks in the past, but i dont think this is one [13:06] I dunno why you won't bother taking anything I say, and look into it though [13:06] you had some ddos against you? or you where part of the source of a ddos? [13:06] i got ddos'd [13:07] i like saying ddos [13:07] * patdk-wk_ gets tripple dos [13:07] I perfer 3 d's, bigger is better right? [13:07] i thought the term was "Denial Of Service" [13:07] airtonix_, no [13:07] is ddos "Daves Denial of Sanity" ? [13:07] Distributed Denial of service [13:07] multible attack sources [13:08] hard to block [13:08] patdk-wk_: not really [13:08] generally thousands of source ip's, always changing [13:08] so you have no idea how to get more informations why its so huge? [13:08] well, only by taking down your ip or subnet [13:08] patdk-wk_: I often find switching the power off to be effective [13:08] Styler2go, I gave you MANY OPTIONS, you ignored them all [13:09] airtonix_, then you just created a dos [13:09] no [13:09] i used your commands [13:09] so you gave into their goal :) [13:09] well, you now have the info you need to look into it then [13:09] i spose you'd have to be interested first, which i was not [13:09] http://pastebin.com/b4HR0YVC for example === jrib1 is now known as jrib === jrib is now known as Guest35945 === Guest35945 is now known as jrib [13:11] maybe i really should restart the server... [13:11] looks like a lot of web hits from that .dip.t-d..... isp [13:14] hmm [13:14] but this is not the problem or? [13:17] I dunno [13:17] check your weblogs [13:17] determin if that is something you allow, or something you think is bad [13:17] it could be as simple as someone having a webpage open from your server, that has autorefresh on it [13:17] so it's just reloading a crapload === xcrracer_ is now known as xcrracer === xcrracer is now known as 1JTAAFY3Q === xcrracer__ is now known as xcrracer [13:26] jamespage: https://code.launchpad.net/~zulcss/python-ceilometerclient/1.0.0/+merge/155962 [13:31] smoser: wassup? [13:31] koolhead had a question above [13:31] http://paste.ubuntu.com/5655036/ [13:33] that isn't CA [13:34] good morning folks! [13:34] gpg --keyserver pgpkeys.mit.edu --recv-key 81DCD8423B6F61A6 ; gpg -a --export 81DCD8423B6F61A6 | sudo apt-key add - [13:34] is one way [13:35] jcastro_: helo [13:37] isn't "CA" ? [13:38] good ol jcastro underbar. [13:38] i love that guy. [13:38] ? [13:39] zul, pong [13:40] jamespage: can you add python-ceilometerclient to the ci [13:40] zul, how much beer is that worth? [13:40] jamespage: hmmmm....2 beers :) [13:41] jamespage: ill get them at the mirantis party for you [13:42] zul, w00t! [13:42] cheapsake [13:42] cheapskate [13:42] totally [13:45] hi guys, after ubuntu 12.04 I cannot copy paste something when using byobu since every n seconds I lost selection, is there a fix for this? [13:48] alex88: probably a better question for #ubuntu, as -server does not have a gui. [13:49] Pici: byobu is a gnu-screen wrapper that display a status bar and the window list [13:49] you think is a client-side terminal program? [13:49] alex88: are you using screen's selection method to copy and paste? or are you using a mouse? [13:50] Pici: mouse, and client is not ubuntu [13:50] http://bazaar.launchpad.net/~zekus-one/byobu/trunk/revision/1992 seems there is already a fix undergoing [13:50] Ah [13:50] sorry just found it now :) [13:50] I've not had that issue myself on any OS, thought it was a terminal/putty/whatever issue. [13:50] jamespage: one more beer https://code.launchpad.net/~zulcss/python-ceilometerclient/1.0.0/+merge/155962 [13:51] alex88: no problem, my fault for not understanding,. [13:51] no worries, thanks anyway :) [13:56] ok, byobu ppa fixed the issue ;) === wedgwood_away is now known as wedgwood [13:59] zul, +1 on the MP for ceilometer-client [14:00] jamespage: cool im just working on keystone now so ill have that for you in a couple of minutes [14:09] how to find out all the dependencies for a installed pkg ? [14:10] I have pkg ``foo'' installed.. how to find out all the other packages installed as dependecy pkg ? [14:15] I'm having an issue with my upstart script logic. I have a script that needs to be run at boot and at shutdown. The script takes args (start and stop) so I believe i need to have each command run in the pre-start/post-post. The upstart script works on boot but when i issue i shutdown its never run. Any ideas? http://pastebin.com/YsHLGafp [14:15] I've also tried stop on stopping network-services* [14:16] apt-cache showpkg foo gives the info.. but not so clean === FieldySnuts is now known as Fieldy [14:20] zul, ceilometerclient setup in the lab [14:20] jamespage: cool thanks [14:23] tobin: if you look at /etc/init/networking.conf, the event will not be emitted in certain circumstances by the looks of it. Try looking at your /var/log/upstart/networking.log. [14:23] jodh: hrm okay [14:25] jodh oh this is because of the virtual networking devices [14:26] is there a way to tell apt to use somewhere other than /tmp/ as its temporary directory? === tsaavik_ is now known as tsaavik [14:36] jamespage: https://code.launchpad.net/~zulcss/keystone/rc2/+merge/155977 [14:37] how do I configure which services are started at boot? [14:47] Fieldy: update-rc.d and check the init scripts afaik [14:47] s/init/upstart/ [14:47] ok thanks i'll poke around [14:48] Fieldy: what are you trying to achieve? [14:48] alex88: i don't want several services i just installed to start up at boot yet, until i'm done configuring everything [14:48] Fieldy: then update-rc.d -f servicename remove [14:49] ah okay that's simple enough [14:49] if they're init scripts, for upstart you should find the init script in /etc/init and comment the line "start on " etc [14:49] yeah simple enough [14:50] i'll reboot and give it a shot [14:50] zul, looking [14:50] zul, also opening havana on precise in the lab [14:50] Keep in mind this update-rc.d method will only work for daemons which are in the /et/init.d ( old sysvinit scripts) and not upstart scripts which are in /etc/init [14:50] jamespage: ack [14:53] zul, I've taken two decision for the lab today; 1) we won't be using the ubuntu series specific branches any longer and 2) the PPA is just called havana [14:53] jamespage: thats fine with me [15:01] genii-around: ok, thanks. yep, one service (bind / bind9 / named) still starts up. yet i can't find anything in /etc/init/ named as i would expect. any ideas how I stop bind9 from starting at boot? === Sendoush_ is now known as Sendoushi [15:03] zul, +1 on keystone === egersdorfer_ is now known as egersdorfer [15:23] jamespage: cool thanks https://code.launchpad.net/~zulcss/python-novaclient/ftbfs/+merge/155992 [15:24] zul, are we all good for client releases for grizzly? [15:24] jamespage: ill go through them one more time this afternoon [15:25] What deployment tool do you suggest to get my website from windows 8 into my VPS server on ubuntu server? [15:41] jamespage: last one i swear today: http://people.canonical.com/~chucks/ca [15:45] would putting my custom iptables rules and various forwarding commands in /etc/network/if-up.d/zzz-firewalling-and-forwarding be a good way to have them applied on boot? [15:46] no, that would run per ifup of each interface [15:46] hm... so it would do it once for eth0 and then once for lo? [15:46] (in my case) [15:46] should, yes [15:47] patdk-wk_: okay, my goal is to get this one series of commands run at boot but after networking is up, is there a better way to accomplish this? [15:48] simple upstart script probably [15:48] BUM? [15:48] but there is an iptables startup package, to emulate the rhel way [15:49] patdk-wk_: would BUM automate this? [15:49] how should I know [15:50] dunno what BUM is [15:53] i'm not either [16:01] test [16:02] the Bott Up Manager [16:02] boot [16:02] * [16:06] zul: https://launchpadlibrarian.net/135520670/buildlog_ubuntu-precise-i386.python-cinderclient_1%3A1.0.2.34.g7369310%2Bgit201303281130~precise-0ubuntu1_FAILEDTOBUILD.txt.gz [16:25] Daviey: where is that from? [16:25] Daviey: oh havana ppa [16:34] Daviey: should be fixed in the next build [16:34] i created the following upstart script and placed it in /etc/init/iptables-and-forwarding : https://gist.github.com/anonymous/5264686 however start iptables-and-forwarding returns: start: Unknown job: iptables-and-forwarding I was following this example: http://upstart.ubuntu.com/getting-started.html .. any ideas? [16:40] figured it out. there's an error on that ubuntu.com page, it says "Jobs are defined in files placed in /etc/init, the name of the job is the filename under this directory without the .conf extension. They are plain text files and should not be executable." i renamed it with .conf at the end of the filename and it works. [16:41] or at least, it could be written a little better. it's easy to read it as telling you to not have the file named .conf === io is now known as IdleOne [17:27] I recently upgraded a cluster of machines from 11.04 to 12.04, and some of them are showing a weird "phantom load". [17:28] The LA is quite high, but there's no significant CPU or IO activity, and there's plenty of free RAM. [17:28] Has anyone else seen this? [17:28] sigh. this crosscompile stuff becomes tedious. [17:28] * hallyn off to be afk for lunch [17:47] zul: those ca updates lgtm [17:47] jamespage: cool [17:51] jamespage: alrighty uploaded === Jikai is now known as Jikan [18:32] is it usual for samba to be so fricken flakey? i have 2 servers virtually itentical smb.conf only difference is the netbios name and the path of the shares...1 works the other doesn't.... [18:41] koyetsu: I've had good luck with samba<->samba things, but windows clients seem very flaky. (even windows<->windows seems trouble.) [18:45] well i seem to have worked it out....but samba irritates me [18:48] ok now on to more important issues. i'm working on setting up an automated ubuntu. i've created my ks.cfg and to answer the partitioning questions i created a preseed file. but i still get the very last question basically: here is the partition table do you want to write it yes/no any idea how to an answer passed to it? [18:49] this is my current preseed http://paste.ubuntu.com/5654282/ [18:50] i am using ldap auth, and i've set up some systems with the encrypted home dir [18:51] i was wondering if i changed the ldap password if i would still be able to mount the users homedir [19:01] ok if no one can help me on that, then one other issue is happening in my install. i'm pxe booting and pointing the install at http://local.server/ubuntu the ubuntu folder is the mounted iso for 12.04 server. during the install it pops up and says that /dists/precise/restricted/binary-amd64/Packages was corrupt. but if i continue i never seem to have any problem. [19:05] koyetsu: I think there's only a handful of packages in the restricted tree; mostly nvidia, fglrx, a few others [19:05] koyetsu: if you can't figure out why it is corrupted or don't care to bother you could remove the restricted from your sources.list and delete the file.. [19:06] sarnold: i'm just trying to make this a no key press install. and like i said this is the mounted iso. shouldn't be a corruption. i've redownloaded etc. and it still says it [19:07] koyetsu: oh hrm. [19:07] thinking about just mirroring the distro off ubuntu but not sure what to get or the easiest way to do it. i definately dont want a complete mirror just 12.04 server 64 bit === matsubara_ is now known as matsubara [19:25] anyone know anything about apt-mirror? i'm trying to severly limit what it downloads....i have to cut from 92gb to 56gb trying to make it mirrow a specific distro for local installs [19:34] oh encrypted homedirs didn't work for ldap users [19:34] meh [19:53] i have a machine with two ip addresses. i restarted the networking service, and the interface, it still remains. any ideas? [19:54] i can't ping one of both addresses though [19:54] but ip a s shows two ips [20:00] woohoo, qemu-linux-sparc working as ss10. [20:00] it's almost as fast as my sparcstation 1+ was === LargePrime is now known as Guest82744 === Ursinha-afk is now known as Ursula === pgraner is now known as pgraner-afk === Pricey_ is now known as Pricey [21:47] Has anyone been successful in setting up mode5 bonding via the method outlined at https://help.ubuntu.com/community/UbuntuBonding? [21:47] Howdy folks, I'm trying to preseed a netboot installer, and having a hard time figuring out which initrd.gz is actually being used. My tftp/apache directory looks like this http://archive.ubuntu.com/ubuntu/dists/quantal/main/installer-amd64/current/images/netboot/ (i just mirrored that page). I can run installs just fine, but would like to take it a step further. By all accounts ubuntu-installer/amd64/initrd.gz should be what is loaded, but e === kentb is now known as kentb-oout [22:30] okay forget my previous question, I had a second copy of the netboot folder in the wrong place, and was editing that... no wonder stuff didn't work :p === wedgwood is now known as wedgwood_away