[01:15] hello all [01:15] i need help please [01:19] Hey, I am trying to install ubuntu server via USB using UNetbootin. When i get the the Unetbooting boot screen it does nothing when i click on install. I can only do default and then that freeze up when i get to the language selection screen. any ideas? [01:23] upon placing a new upstart job in /etc/init/, it vanished upon reboot. how can I prevent this? (or properly add an upstart job, if I'm doing it wrong) [01:27] i am trying create a ubuntu-server cd customized and i would like know where i found the script dialog ubuntu-installer 'Ubuntu Installer Main Menu' i need know that to change somethings.. do anyone know? [02:15] corpse: Try using usb-creator to make the image (or usb-creator-kde on 10.04). Those seem to work better. [02:38] RoyK: found my problem with the innodb plugin [02:38] had to add apparmor rules in to allow mysqld to mmap/dlopen the .so files [02:39] * chrismsnz ffuuu === zelda is now known as Zelda [07:20] Which FTP server would you recommend? [07:23] vsftpd. [07:23] 'one that works' ? :P [07:25] elnur: vsftpd [07:26] elnur: with chroot enabled [07:27] lifeless, yea :) [07:27] jpds, bc, thanks [07:28] * SpamapS has always been a proftpd man [07:29] but vs has a bit better record w/ security [07:29] and probably these days just as many great features. :) [07:30] i've tried pureftpd once and have no xp with vsftpd. anyone ever used pureftpd to be able to compare it to vsftpd? [07:31] SpamapS: On a high traffic server, with typical users, I haven't seen anything vsftpd didn't solve. [07:32] I'm about to fall out of the chair. Talk with you tomorrow, [07:33] Okay. Then goes vsftpd. Thank you, guys. [07:45] What is the difference with ftp and sftp? Why I get chrooted to my home dir with ftp but don't with sftp? [07:47] elnur: sftp happens over SSH. [07:48] jpds, huh. so ftp and sftp are handled with different services? [07:48] Yes. [07:48] elnur: You might want: http://www.debian-administration.org/articles/590 [07:49] there's also that magical FTPS that happens via SSL.. but thats always produced odd reactions when I suggest people use it. [07:49] mostly because client support is pretty hard to find [07:51] jpds, yea, i want that. thanks [08:20] bc: vsftpd is probably the most used on larger sites - take a look around :) === a is now known as notit === ogra_ is now known as ogra [10:22] New bug: #584862 in libvirt (main) "libvirt launches dnsmasq for non-dhcp networks." [Undecided,New] https://launchpad.net/bugs/584862 [10:41] I have a Ubuntu 9.10 server. It's a VPS running under OpenVZ. I'm trying to set up quota during the install process of ISPConfig 3, but I have a problem... / is not in fstab, and I'm supposed to add ",usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0" to the entry [11:07] FireCrotch: define your problem again? you should be able just edit fstab and add mentioned line to options [11:07] simplexio: my fstab doesn't have an entry for the root filesystem though [11:08] FireCrotch: ah.. mount should tell you needed ifrmation which to add into fsab [11:08] FireCrotch: but 'mount' command tell you that you have / mounted [11:09] yes, mount says that / is mounted... /dev/simfs on / type reiserfs (rw,usrquota,grpquota) [11:10] Note that that's before I've done anything related to quota (so anything relating to quota is a result of my webhost's config) [11:10] could it be that you need define root partition options in openvz virtualmachine settings or something [11:13] Yeah... I don't think I'll be able to use quota with this setup that they have. No big deal. But thanks for the help, simplexio [11:13] and if you can toy wth installation, so you sont loose anything if it isnt working anymore, you could just try add / to fsab [11:16] or try use mount -o remount,moreoptions / [11:34] wtf is simfs? [11:51] New bug: #584907 in postfix (main) "package postfix 2.7.0-1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 75" [Undecided,New] https://launchpad.net/bugs/584907 [11:56] New bug: #584910 in libvirt (main) "libvirt can't add host definitions to DHCP without network restart" [Undecided,New] https://launchpad.net/bugs/584910 === pgraner-afk is now known as pgraner === oubiwann is now known as oubiwann_ [12:59] hello [12:59] can anybody point me to a doc to allow name+mailbox@domain handling of mail on a dovecot/postfix ubuntu server ? [13:00] where name+mailbox@domain send your mail to name@domain and put it directly in the mailbox folder === oubiwann_ is now known as oubiwann [13:08] ploum_: i would use procmail or sieve for that [13:09] Daviey, I know that it is done in the dovecot-postfix package, that's why I'm asking (because installing this package breaks my config) [13:09] ploum_: pass then, sorry. [13:10] how would you do that with sieve ? Afaik, the mail will be refused by Postfix (name+mailbox is not a valid user) before being received by sieve [13:10] why do you think that? [13:25] ploum_: doesn't postfix have some rewrite magick like old sendmail? [13:27] he's yet to say why he thinks postfix will refuse such mail. [13:33] good morning. I've been getting an error when using apt-cacher-ng... "Signatures were invalid" [13:33] http://ubuntuforums.org/showthread.php?t=1491239 [13:34] any help would be greatly apprecaited. [13:36] Italian_Plumber, invalid signing key [13:36] yes [13:37] clear the file and put the right one in its place [13:37] which i assume you just download the file from either the ubuntu archive or a similiar mirror [13:38] where is the file? [13:40] Italian_Plumber, where did you setup apt-cacher-ng to put the debs? [13:40] I don't know. wherever the default is. I just typed "apt-get install apt-cacher-ng" [13:41] I think it's /var/cache/apt-cacher-ng [13:42] yeah [13:42] There's a file there "_impkeycache" [13:42] ubuntu/dists/hardy [13:42] is that it? [13:42] ubuntu/dists/hardy-updates it would seem [13:43] Here's what in the directory: [13:43] /http://pastebin.com/0R3xxYUV [13:43] sudo mv _impkeycache _impkeycache.old [13:44] then i would somehow resync or whatever with like an apt-get update and then check to see if it worked [13:44] awesome [13:44] I'll try that [13:44] I'm waiting right now for the client VM that I'm testing this with to finish resuming [13:46] well it might be a while because my speeds between here and that machine suddenly took a crap. :( [13:47] Unfortunately I still see the error. [13:48] http://pastebin.com/KRk8SJ5J [13:50] morning [13:50] Italian_Plumber: sudo apt-get update -o Acquire::http::No-Cache=true [13:51] jpds: thanks... should I do that on the client or server? [13:51] Cliet. [13:52] Unfortunately I still see the error [13:53] Here is the entire command: http://pastebin.com/46QPMciU [14:18] hello, is there a way to let users use my network printer without the need to install a driver for each client machine ? I tried cups, but it looks to me like just defining printers and let others use it via samba, am I correct ? [14:23] RoyK that's what I'm trying to find [14:23] but I don't even know the name of this feature ;-) [14:31] New bug: #584949 in vsftpd (main) "package vsftpd 2.2.2-3ubuntu6 failed to install/upgrade: subprocess installed post-installation script returned error exit status 3" [Undecided,New] https://launchpad.net/bugs/584949 [15:05] * ccheney will be headed to doctor in about an hour [15:16] kirkland: are you OK with your assignment on bug 584497? In other words, can I mark it in-progress? [15:16] Launchpad bug 584497 in landscape "The EC2 endpoint should be HTTPS by default" [High,Confirmed] https://launchpad.net/bugs/584497 [15:18] hggdh: leave triaged [15:18] hggdh: assigned to me is fine [15:22] kirkland: thanks === dendro-afk is now known as dendrobates [15:59] is the ubuntu server guide still available as a pdf? I can't seem to find it for 10.04, only older releases [16:06] On a shiny new server, I used tasksel to install the LAMP server software. While trying to set up Wordpress, going to locations like 'blah/install.php' my browser downloads the file, rather than displaying a dynamically generated page. This makes me think that PHP isn't installed correctly, but I assume that it should be. What's going wrong here? [16:06] linxeh_: working on getting it uploaded [16:07] linxeh_: bug 575771 has details if you're interested [16:07] Launchpad bug 575771 in ubuntu-docs "No PDF version available for 8.04 and 10.04 Server Guide" [High,Fix committed] https://launchpad.net/bugs/575771 [16:09] StrangeCharm: might double check that the php5 module is loaded... also try restarting apache [16:10] sommer, how do i check that? [16:11] ls /etc/apache2/mods-enabled/ ... should see php5.conf and php5.load symlinks [16:12] New bug: #364898 in vm-builder (universe) "python-vm-builder depends on kvm" [Wishlist,Triaged] https://launchpad.net/bugs/364898 [16:16] how can I disable framebuffer at boot? [16:16] it doesn't work because I have crappy vidcard, but it still tries and loads a bunch of modules [16:17] sommer, yep, seem to be there [16:17] StrangeCharm: I'd try sudo /etc/init.d/apache2 restart then... tasksel may have not restarted apache [16:20] sommer, nope, still getting the install.php file [16:23] sommer, any other ideas? [16:24] StrangeCharm: not sure... is there any errors in your /var/log/apache2/error.log? === Thell|Away is now known as Thell [16:28] sommer, nothing interesting. it does say "[notice] Apache/2.2.14 (Ubuntu) PHP/5.3.2-1ubuntu4.1 with Suhosin-Patch configured -- resuming normal operations" after a restart though, which bodes well for the presence of php [16:32] StrangeCharm: yep, might be something with the configuration then... might check the settings in /etc/php5/apache2/php.ini [16:32] §w§w§wstr [16:34] sommer, i have no idea what i'm looking for [16:34] StrangeCharm: check that you have php5 files in /etc/apache2/mods-enabled/ and write file index.php containing phpinfo(); in www-root [16:34] that is easiest way to test php install [16:34] StrangeCharm: I agree with simplexio :-) was about to recommend that [16:36] for soereason php has been allways kinda hit and miss instal, attleast for me. it works or you just cant figure what you missed [16:37] sommer, simplexio it just renders that string [16:37] so basicly your config dosent recognioze that .php endin files are phph [16:37] StrangeCharm: so you have two php files in mods-enabled ? [16:37] StrangeCharm: you might also look at the Server Guide: https://help.ubuntu.com/10.04/serverguide/C/php5.html [16:38] StrangeCharm: and i assume that you have restarted apache after you installed php [16:38] simplexio, i have restarted apache [16:39] StrangeCharm: are your files possibly stored in a directory other than /var/www, but referenced by symlink? [16:39] and php5.conf and php5.load in mods-enabled [16:39] SpamapS, my files are certainly in /var/www [16:39] StrangeCharm: try putting a file in /var/www/test.php , have it just one line: 'lo. [16:40] StrangeCharm: if you point your browser at that, does it show? [16:40] StrangeCharm: and est string foe index.php is phpinfo(); [16:40] StrangeCharm: or rather, does it show a giant page of PHP information, or just the line as-is. [16:40] SpamapS: hey man. How goes it? [16:40] SuperLag: :) it goes well thanks. [16:40] SpamapS, that renders a whole load oh php info [16:40] I'm looking for a simple recipe for central auth on Ubuntu 10.04. I have found some pages talking about there being a blueprint for this targeted at Lucid; anyone know if this was implemented? [16:40] :) is little while when i wrote php [16:41] StrangeCharm: so its working [16:41] Seems to relate to mathiaz and Puppet. [16:41] simplexio, yes, that would be my broad conclusion [16:43] so, why am i downloading files from /var/www/wordpress/wp-admin ? [16:43] some times i feel that there should be somekind standart test to take, so you could allways point first at your score when asking questions [16:44] simplexio, why not write a wiki page for that? [16:44] StrangeCharm: i assume that they are .php, cant rememeber but does apache need .php files to be executable [16:44] simplexio, it is a .php file: it's called install.php [16:44] I need entropy on my server for gpg --gen-key. I can't use user input for it either... So.. I found ekeyd-egd-linux which seems like it might be useful but it seems under documented. Any ideas how I can either use this or do something else to get the entropy I need? [16:45] StrangeCharm: are you sure /var/www/wordpress is not a symlink? [16:45] StrangeCharm: try give it execue rights [16:46] SpamapS, i'm sure. i just downloaded a compressed file into /var/www, then decompressed it [16:46] simplexio, what's the command for that? [16:46] and im using shell over gprs so typo more than normal [16:46] StrangeCharm: hmm [16:46] chmod uag+x file [16:47] its little overkill, and you should drop rights that are too broad. if it works [16:48] uag ? [16:48] you mean a+x ? [16:49] Is that a Suhosin thing, php scripts have to be +x ? [16:50] dunno. its just my nesxt best quess [16:50] that did not help [16:50] hmm no [16:50] StrangeCharm: who owns the wordpress dir / files? [16:51] SpamapS, who should? [16:51] StrangeCharm: depends on your strategy for security, but probably root [16:51] well apache can read it, so t should not be problem.. [16:51] StrangeCharm: or "not www-data" [16:52] StrangeCharm: I missed the beginning of this.. is this the default Lucid setup? [16:52] kirkland, ping [16:52] SpamapS, i hope so. the os came reinstalled from my web-host [16:53] * ccheney back [16:54] StrangeCharm: any .htaccess files in there maybe screwing you up? [16:54] SpamapS, in where? [16:54] StrangeCharm: wordpress [16:55] wow... does every package need to set a locale? [16:55] SpamapS, don't think so [16:56] StrangeCharm: can you move that test.php into the wordpress dir and see if it works there? [16:57] smoser: yo [16:57] hi ppl, i have a HP Proliant ml350 G& server with two sata 500gb drives in raid 1+0 (hp hardware p410i controller) [16:57] there have been a spike and now my ubuntu 9.10 server does not boot anymore [16:57] (kernel panic vfs not syncing) [16:58] i can mount the partition, i did e2fsck, fixed a few inodes etc. [16:58] kirkland, did you open a blvueprint for lxc work ? [16:58] SpamapS, test.php works when its in the same directory as the file i'm trying to open [16:58] i even tested grub, [16:58] i thought you had done that, memory ill. [16:58] but i didn't see one when i lookd [16:58] but it keeps giving me that error, any hints? [16:58] smoser: hmm, i think it was bob blair (smoothstone) that did [16:59] different lxc [16:59] thies was per ensemble. [17:00] i thought you had told niemeyer that you would ompen one [17:00] SpamapS, which suggests that there's something deeper broken [17:00] (same lxc different context) [17:00] StrangeCharm: indeed [17:00] StrangeCharm: are there any errors in /var/log/syslog? [17:01] smoser: i filed one for the last uds on lxc, though [17:02] no problem. [17:02] smoser: oh, hmm [17:02] smoser: give me a sec [17:02] i was just not wanting to duplicate work if you'd done seomthing. [17:02] SpamapS, nothing in the last 20 mins, though i certainly tested later than that [17:02] smoser: yeah, i filed an ensemble one for Gustavoe [17:03] smoser: https://blueprints.edge.launchpad.net/ubuntu/+spec/server-maverick-ensemble [17:03] StrangeCharm: permissions maybe? What are the perms of the file that works, and the files that don't work? [17:03] SpamapS, how do i check permissions? [17:04] kirkland: heh.. btw, I'm full time byobu now on my personal servers.. and I keep typing 'clear' every time I ssh in.. force of habit.. [17:04] kirkland, ok. thanks. [17:04] SpamapS: heh, thanks ;-) file bugs as you see them [17:07] SpamapS, i don't know how to see the permissions on a file [17:08] ls -l === RoAk is now known as andreserl [17:10] SpamapS, install.php has -rwxr-xr-x 1 flamsmark www-data 6547, while test.php has -rw-r--r-- 1 flamsmark flamsmark 17 [17:10] not sure what that means, though [17:12] StrangeCharm: try doing 'chmod a-x install.php' [17:12] StrangeCharm: any different? [17:12] though I think realistically thats there because you did uag+x earlier [17:12] SpamapS, no change [17:12] so if that doesn't work.. I'm really confused [17:13] two files, in the same directory, with the same owner, and the same extension. php executes one, but not the other [17:13] StrangeCharm: yeah I think PHP is being told not to execute install.php somehow.. [17:13] how would it know? [17:13] thats why I suggested looking for .htaccess [17:14] SpamapS, can you think of any other methods by which that could be occurring? [17:15] try chgrp www-data test.php [17:16] yeah thats worth a shot didn't think of that.. [17:16] but it would be.. confusing [17:16] yeah, it doesn't make a lot of sense why it would work like that, but it could, and testing/eliminating that is important [17:17] :) [17:18] after running that command, test.php still renders [17:18] StrangeCharm: and still nothing in /var/log/apache2/error.log ? [17:20] SpamapS, no, looks not [17:22] StrangeCharm: there's a logical explanation of this, but the usual reasons don't seem to be causing it... [17:22] SpamapS, there is an explanation, because computers don't act of their own accord [17:22] sommer: ah ok thanks [17:23] np :) [17:23] SpamapS, is there any way that something knows that i created test.php, but downloaded the wordpress files? [17:23] StrangeCharm: no [17:24] StrangeCharm: You *could* try the ubuntu wordpress packages [17:24] just to see if they work instead of the downloaded version [17:24] mv wordpress wordpress-downloaded before that though [17:25] SpamapS, is there an ubuntu wordpress package? [17:25] Yes, but it'll put wordpress in /wordpress rather than in the main site [17:25] If I recall :) [17:26] oh, so there is [17:26] cloakable, well, i can always move it back [17:26] :) [17:28] SpamapS, the ubuntu package gives me http://pastebin.com/Hj1fgnYX [17:28] StrangeCharm: doh [17:28] there goes those recommends installs [17:28] ? [17:29] SpamapS, sorry, i don't understand [17:29] StrangeCharm: recommended, but not 100% *required* packags are installed by default [17:29] you probably don't need dovecot [17:29] StrangeCharm: add --no-install-recommends to your apt-get line [17:30] apologies if this is touching on religious ground - but is there any group calendar server software available in lucid, that can be accessed over a network (ideally ssl) to clients such as MS outlook ? :o [17:30] actually probably already installed and just dovecots are broken. :( [17:30] linxeh_: its actually a point of much interest for the current release cycle [17:30] I've been looking at egroupware and zimbra, but they seem a bit heavy handed (using dovecot atm for the mail) [17:31] and with the apple caldav server being opensource etc, though not found any really supportable packages for that yet [17:31] SpamapS, it still tries and fails to install dovecot [17:31] linxeh_: I saw Kolab suggested but I think its kind of big too [17:31] StrangeCharm: apt-get remove dovecot-common [17:31] New bug: #585026 in clamav (main) "freshclam won't execute /etc/clamav/onupdateexecute.d scripts" [Undecided,New] https://launchpad.net/bugs/585026 [17:32] SpamapS, then try again? [17:32] this kind of thing https://wiki.edubuntu.org/CalendarServer [17:32] SpamapS: yes, I looked at kolab but it seemed like the wrong solution for me [17:33] SpamapS, now wordpress doesn't complain during the install [17:33] however, going to /wordpress downloads a file for me, rather than rendering a page [17:34] StrangeCharm: after install, did you restart apache2 ? [17:34] SpamapS, no, i'll do that now [17:35] SpamapS, i'm still downloading a file [17:36] New bug: #585027 in multipath-tools (main) "Race condition with dmsetup causes 'map already present' messages" [Undecided,New] https://launchpad.net/bugs/585027 [17:39] StrangeCharm: is the file the contents of index.php ? [17:40] SpamapS, roughly. it's a file called download.php, and it claims to be the wordpress master file that loads whatever other file is needed for the page [17:40] smoser ping [17:41] hggdh, yo [17:41] smoser: ok, I confess. I cannot get the bloody vmbuilder to build for ec2 [17:42] smoser: so... how do they do it? I installed python-vm-builder and -ec2 [17:42] i'm consused as to what you're after. you want to build a ec2 vm ? [17:43] * SpamapS wonders if the FilesMatch isn't working the way it should [17:45] smoser: yes, this was my intention... [17:45] StrangeCharm: try creating a file, /etc/apache2/conf.d/something.conf with this line: [17:45] AddType application/x-httpd-php .php [17:46] hggdh, i have not actually done this with vmbuilder 0.12 . the nightly builds use 0.11. the code that invokes it is at https://code.launchpad.net/~ubuntu-on-ec2/vmbuilder/automated-ec2-builds [17:46] hi all [17:46] * SpamapS is a bit confused at the use of FilesMatch which seems.. more confusing [17:46] hggdh, that said, why do you want to do that i wonder ? [17:46] I'm configuring ocfs2 file system [17:47] smoser: verifying a bug that states cannot build ec2 images ;-) [17:47] is there something insufficient with the existing images ? i find it generally easier (and faster) to start with that base image and modify it. [17:47] although I get no error I'm missing something [17:47] hggdh, ah. well, then. [17:47] :) [17:47] which, I guess, I *am* confirming [17:48] isn't supposed that any file change be replicated through all the servers? [17:48] SpamapS, do i need to activate it in nay way, or can i just restart apache? [17:49] I mean, if i create a new file in on machine of the cluster, it should be visible on the other node of the cluster, because we're talking about a shared storage. wright= [17:49] SpamapS, just restarting apache doesn't solve the problem [17:49] wright? [17:50] when I create a file on one of the nodes, I can't see it on the other node [17:50] I've got the same config on both [17:50] can someone help me? [17:51] is there any other channel where I should post? [17:51] hi guys [17:52] im having troubles with node registration in UEC. I cant register my node, when i run euca-describe-availability-zones verbose i get 0/0 [17:52] someone can help me? [17:53] additionally my nc doesn't have the nc.log, despite the eucalyptus-nc service is running [17:54] StrangeCharm: I'm at a loss. This really makes no sense at all. === oubiwann is now known as oubiwann_ === oubiwann_ is now known as oubiwann [17:57] tyska, you're on 10.04 ? [17:57] yeah [17:57] could you please open a bug ? ubuntu-bug eucalyptus-common [17:57] please answer yes to 'attach logs' [17:57] SpamapS, this is most frustrating. the problem persists after reinstalls [17:57] i had the UEC working with 9.10, then i decided update my UEC [17:58] this is something we've seen, but dont have a bug for. we need to address it. [17:58] SpamapS, that is, os reinstalls [17:58] ok [17:58] where i open a bug? [17:59] tyska: I had the same [17:59] deregister and register the cluster controller again fixed it for me [18:00] the cluster? [18:00] hmm [18:00] i do it with the node [18:00] deregister and register the node [18:00] i will try with cluster [18:03] StrangeCharm: may be time to open a bug if a brand new install can't run from the packaged wordpress [18:03] binBASH: your UEC is working with ubuntu 10.04? [18:03] tyska, run it on the CC [18:03] SpamapS, where should the bug go? wordpress? lamp? tasksel? php? [18:03] (the ubuntu-bug) [18:04] StrangeCharm: not sure [18:04] smoser: oh ok [18:04] and i will second that you might have success with deregister and register of node [18:04] but please open bug (first) [18:04] smoser: i've tried deregister and register the cluster, now im rebooting [18:06] StrangeCharm: take a look in /usr/share/doc/wordpress before you do anything else [18:06] StrangeCharm: start with README.Debian [18:06] tyska: Yup, it was [18:06] I switched to simple kvm virtualization though [18:06] don't need a cloud really [18:09] smoser: deregister and register cluster does not worked! =( [18:10] smoser: now i already did the ubuntu-bug that you said [18:11] smoser: it's very weird, because i run euca_conf --register-nodes NODE_IP, receive a success message [18:11] smoser: but when i run euca_conf --list-nodes i receive nothing [18:26] How do I set the landing directory on a chroot'ed sftp server? [18:28] tyska, i'm really sorry. i don't really have more suggestions atm [18:28] smoser: =( [18:28] smoser: ok, thanks anyway [18:29] thank you for opening the bug. [18:30] tyska: Here I deregistered all nodes via euca_conf then deregistered the cluster controller, registered the cluster controller again and also readd the nodes. [18:30] smoser: you're welcome. i hope with it, the problem can be found. [18:30] but I didn't use webiface [18:30] binBASH: im trying everything [18:30] just euca_conf [18:30] binBASH: the sad is that i have all working with 9.10 [18:30] like mine ;) [18:31] You describe exactly what I have encountered [18:31] binBASH: but now i cant use UEC with ubuntu 10.04 =( [18:32] tyska: I was able to use it with what I described ;) [18:34] binBASH: i run euca_conf --deregister-cluster 192.168.1.1 [18:34] binBASH: receveid this message: SUCCESS: cluster '192.168.1.1' successfully deregistered. [18:34] binBASH: and the run euca_conf --list-clusters [18:35] what is the name of the proggy that allows you to send ssh commands to multiple servers at once [18:35] tyska_, that shows none ? [18:35] binBASH: and received: registered clusters: c3cluster2 192.168.1.1 [18:35] ah. ok. [18:35] and then you register node and nothing [18:35] right ? [18:35] o.O [18:35] smoser: yeah [18:35] if i register a node, i have success message but no nodes are showed with --list-nodes [18:36] do you have important data here ? [18:36] or images/instances ? [18:36] bogeyd6: ClusterSSH [18:36] smoser: no [18:36] pssh [18:36] binBASH, thanks [18:36] 19:30 < binBASH> tyska: Here I deregistered all nodes via euca_conf then deregistered the cluster controller, registered the cluster controller again and also readd the nodes. [18:37] i'd recommend trying: sudo eucalyptus stop CLEAN=1 && sudo eucalyptus start CLEAN=1 [18:37] wait [18:37] binBASH: but i cant deregister the cluster [18:37] for me it did work then [18:37] sudo stop eucalyptus CLEAN=1 && sudo start eucalyptus CLEAN=1 [18:37] binBASH: i run the command and the cluster is still there [18:37] and same with the eucalyptus-nc [18:37] smoser: i will try it [18:38] smoser: progress, --list-clusters showed nothing [18:38] :) [18:38] smoser: now i will run the same command in the nc [18:39] yeah: sudo stop eucalyptus-nc CLEAN=1 && sudo start eucalyptus-nc CLEAN=1 [18:43] smoser: there is one thing very weird [18:44] smoser: my cc.log there is nothing writed since may 20 [18:44] smoser: is that normal? [18:44] probably no. [18:44] that is the symptom we're seeing [18:45] in a fresh install cc.log will not even exist. [18:45] wait. [18:45] maybe i've seen this with nc.log [18:45] ugh. [18:46] New bug: #585067 in mysql-dfsg-5.1 (main) "package mysql-server-5.1 5.1.37-1ubuntu5 failed to install/upgrade: subprocess new pre-removal script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/585067 [18:46] How do I set the landing directory on a chroot'ed sftp server? === dendrobates is now known as dendro-afk [18:51] smoser: i think cc.log always exist, including it is update all the time [18:54] s [18:58] New bug: #452219 in samba "SMB Browser Printer Queues Not In Alphabetical Order" [Undecided,New] https://launchpad.net/bugs/452219 [18:59] ohnoes === dendro-afk is now known as dendrobates [19:02] heh === dendrobates is now known as dendro-afk [19:18] * ccheney bbl, headed to austin [19:20] so.. I'm working on this https://blueprints.launchpad.net/ubuntu/+spec/server-maverick-uds-web20-workloads spec [19:20] at the UDS session we presented conflicting package lists and concepts.. [19:21] at the session we all agreed that we should push for libmemcached based solutions.. [19:21] but none of the listed language bindings are libmemcached based.. [19:21] worse, none of the libmemcached based language bindings except php5 are even in debian/ubuntu... [19:23] Wondering if I should rethink having all of the language bindings in main.. given that they will be new packages.. :-P [19:23] I have a pool of web servers for which I need to keep media files (images, etc) in sync. Is there a way to automatically monitor a directory for changes, and then sync those changes with the server pool? [19:23] savid: use drbd [19:23] thats pretty low level [19:24] or something like shared storage example glusterfs [19:24] and not active/active last I checked [19:24] savid: gluster is an excellent choice [19:24] its in universe in lucid, and works great [19:24] I'm using glusterfs myself ;) [19:24] binBASH: how long have you been using it? [19:25] https://blueprints.launchpad.net/ubuntu/+spec/server-maverick-cloud-gluster .. I'm drafting that btw.. ;) [19:25] actually I think its about ready for review [19:25] Not long. Just from 3.x [19:25] Hopefully w/ Maverick we'll be able to build nodes in the cloud that automatically mount/export gluster volumes [19:25] SpamapS: Here I run a geoip based setup with gluster [19:25] binBASH: *nice* [19:26] binBASH: so your storage volumes are spread out over multiple data centers? [19:26] some varnish servers which loadbalance some webservers. Clients are sent to the varnish hosts via geoip dns [19:26] SpamapS: No [19:27] that would be pretty cool though. ;) [19:27] I have here only 100 Mbit gluster though :p [19:27] savid: the simplest way btw, if you just have 2 or 3 servers, is choose one as the write-master, and just rsync to the others periodically. [19:27] savid: but if you have millions and millions of files, that won't work too well [19:28] binBASH: are you using the packages from lucid or your own? [19:28] so the varnish cache the small files in local storage [19:28] SpamapS: gluster from lucid [19:28] SpamapS, the main requirement is to have it sync whenever there are changes [19:28] binBASH: its good to know its already getting usage. :) [19:28] the varnish I packaged myself, because the ones from lucid are a bit broken [19:28] savid: yeah, you probably want gluster [19:29] binBASH: do tell? [19:29] SpamapS: Well the ones at lucid (2.1.0) have some issues with esi. [19:29] What's the difference between glusterfs and gluster-platform? [19:29] the varnish developers recommended me to use 2.1.2 [19:30] savid: gluster-platform is a management node to help you setup gluster [19:32] hi guys, i am trying to set up pure-ftp on a server behind an internet router using an alternative port but i am having problems connecting fromthe internet since it appears that pure-ftp sends the local ip back to the client which obviously doesn't work :-/ another problem is that if i use the standard port, for some strange reason pure-ftp doesn't chroot the users although correctly configured, using the alternative port it works!? [19:33] binBASH: is that a known bug or just something you ran into? [19:34] SpamapS: Like, I said the developer told me that is a known bug. [19:34] it's not a special under ubuntu bug :p [19:34] the varnish release has that bug [19:35] hmm.. I'm not sure exactly where my starting point is.. gluster-platform seems pretty low-level, I mean it's talking about making disks bootable, etc... I'm running my sites on virtual servers, so not sure if that's applicable or not [19:35] savid: Just install glusterfs-server [19:35] it's very easy to setup [19:36] k [19:36] savid: http://www.gluster.com/community/documentation/index.php/Storage_Server_Installation_and_Configuration [19:37] I setup mine, on 6 nodes within 10 minutes. [19:37] binBASH, So do I do one server, and a client for each node? [19:37] savid: If you want to use only the storage of one server. [19:38] if not you can run server on each node, and have the storage in raid 10 [19:38] then you have failover + redundancy [19:39] also more server nodes give more speed [19:39] ah, ok === dendro-afk is now known as dendrobates [19:58] smoser: can you see my logs and conf files and see if you find something wrong? You can found all here: http://forum.eucalyptus.com/forum/no-nclog#comment-11351 [19:58] tyska, did you open a bug ? [19:59] smoser: the command ubuntu-bug? [19:59] yeah [19:59] yeah i've opened. [19:59] number ? [20:00] i dont know [20:01] wait a minute, i will send you the link [20:01] the link is https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+filebug/egx0regnhFORcAJyCYITaE8R4Vg? [20:02] i cannot access because i dont have an account [20:06] oh. well, that doesn't open a bug then. [20:06] :) [20:06] you have to then fill out the bug report. [20:06] ubuntu-bug just collects information and directs you on where to go [20:13] i'd really appreciate it if you could create an account and file the bug. [20:14] smoser: hmm, then i will do this [20:15] and please link to this trehad in the bug report [20:15] smoser: i dont know why, but now i get it work [20:15] yeah :-( [20:15] smoser: finally get the node registered [20:15] please do open the bug and link to that thread. thats the only way we're actually going to get it fixed in lucid. [20:16] smoser: that's right, i will do this now === xfg is now known as zul [20:29] SpamapS: maybe get it into universe first and go from there [20:30] savid: Did you setup glusterfs now? [20:31] binBASH, nope, I went to lunch :-) [20:31] ahh :p [20:31] zul: yeah thats what I'm thinking too.. its far more important that we get the right things in than we get the wrong things into main.. so I'm wondering if we should just let the other ones stay in universe [20:31] binBASH, now I'm reading the docs to try and get a handle on how it works [20:31] lunch.. what a good idea [20:31] SpamapS: probably...i wonder how many people use it though [20:32] * binBASH had lunch some hours ago ;) [20:32] lunch -- it's the new breakfast [20:32] zul: memcached or libmemcached? [20:32] SpamapS: either or [20:32] zul: memcached is in every single web shop I've talked to for the last 3 years [20:32] zul: it has become ubiquitous [20:32] SpamapS: What a surprise :P [20:33] SpamapS: meh ;) [20:33] in fact, "use something like memcached" has become one of those LAMP best practices that you get tired of hearing about at LAMP meetups/conferences [20:33] too bad ubuntu has not s3fs yet [20:34] smoser: its done [20:34] smoser: https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+bug/585108 [20:34] Launchpad bug 585108 in eucalyptus "Problems with node registration" [Undecided,New] [20:34] thank you [20:34] google for "memcached appliance" ... at least 5 companies have created them. Once people are quitting their day jobs to create things around a technology.. it has achieved ubiquity. :) [20:34] smoser: you're welcome [20:34] like spam filtering in the early '00s [20:35] smoser: i have another doubt, now on the manage of the cloud [20:35] smoser: ive used a centos 5.4 image to create a instance [20:35] smoser: how can i know what is the user of this image? [20:35] you mean what user launched the image ? [20:36] smoser: no, what user i need to use to login into the instance via ssh [20:36] theres no magic [20:36] smoser: in the uec ubuntu images, i use ssh -i mykey.priv ubuntu@IP [20:36] if you didn't configure a user, then there is no user [20:37] smoser: but how i configure a user?? I've downloaded the image from the tab extras on web interface of the cloud [20:38] smoser: then i try to login with ssh -i mykey.priv IP, without user, then it ask for password [20:38] oh. [20:38] i'm sorry. [20:38] its probably root then. [20:38] if it came from eucalyputs [20:38] i've not used that much. [20:39] i'm generally unfamiliar with web UI. [20:39] i thought you were saying you created the centos image yourself. [20:41] smoser: ok, will try with root [21:00] smoser: are you there? [21:00] i'm here. [21:00] smoser: worked with root! =D [21:00] smoser: thanks [21:06] New bug: #585121 in bind9 (main) "rndc.key permission denied" [Undecided,New] https://launchpad.net/bugs/585121 [21:12] smoser: now im having problem with EBS volume, i attached it, create a partition with cfdisk, but i cant mount it [21:12] smoser: it saids i need to specify a filesystem. What is the default filesystem type for Linux? [21:17] tyska, mount -o offset=32256 [21:18] if that fails, try offset=16384 [21:19] ChmEarl: mount: you must specify the filesystem type [21:19] tyska, use what you had before but add the option [21:20] still fails with the option? then check fdisk on the volume, and get the cylinders and size [21:24] ChmEarl: i've used just adding the option [21:24] ChmEarl: the number of cylinders is 10240 [21:28] hello to all. i have big problems to install the new ubuntu-server LTS edition on a Proliant ML530 G2 Server. it say me allways it can not find the Network Interface. Need to return to Network Recognisation Step to load some Kernel Modules but this dont help [21:29] sound like a hardware recognisation failure. have aborted the install process now and landed in the terminal modus. [21:29] tyska, use fdisk -l -u dev (and get the sector size and start) [21:29] can anybody help me with this Problem [21:30] xperia: do you know what chipset it uses? [21:30] tyska, in the instance, when you get an ebs volume, you get an empty disk [21:30] already solved [21:30] so you have to partition it (cfdisk / fdisk / sfdisk ... ) [21:30] ok. [21:30] yeah, mkfs.ext3 or mkfs.ext4 [21:30] yeah [21:30] mkfs.ext3 solved my problem [21:30] tks anyway [21:30] or if you want to see if you can lose your data mkfs.btrfs [21:30] :) [21:31] smoser: how can i install LAMP in my EBS? i found a page teaching how to install but using debian [21:32] i can't really help you with how to install centos lamp stack, sorry, its just not my expertise ;) [21:32] Spamaps: have saved the install debug log but to see what exact the problem is but the information are shown too fast. tryed cat with | less but less dont exist on the cdrom [21:32] but if you wanted a generally functiona lampstack, i would suggest getting the uec image from uec-images.ubuntu.com/releases/lucid [21:32] then 'apt-get update' [21:32] and tasksel [21:33] select "LAMP Server" [21:33] and i dont know how to find out the type of the network card. maybe dmesg ? [21:34] lspci [21:34] dmesg or lspci should both work [21:34] with lspci it's less scrolling ;) [21:34] okay will try that but what can i doo to see the output only page by page. normally i use less for that [21:35] you can try more [21:35] okay lspci has outputed some information [21:36] Host Bridge: Broadcom CMIC-HE ( Rev 22) [21:37] smoser: not centos, i wanna install LAMP on ubuntu [21:37] ISA Bridge: Broadcom CSB5 Southbridge (rev 93) [21:37] well, follow above. [21:37] smoser: but dont know how to do this on a EBS volume [21:37] i dont know what that means "on an EBS volume" [21:37] binBASH, I'm still confused about how I should be setting up glusterfs. Am I supposed to set up NFS on all my servers? [21:38] smoser: look this http://open.eucalyptus.com/wiki/install-service-ebs-volume [21:39] savid: No ;) [21:39] smoser: instances are transient, if i install LAMP on a instance of ubuntu image, when the instance terminate, or when the cloud machines are turned off, i lost all my files and configuration [21:39] xperia: You'll want one with Ethernet in the line [21:39] savid: You just run the client to mount the share [21:39] cloakable: strange thing is it dont exist such a line in the output but i am able to ping the server [21:39] xperia: hmmmm [21:40] odd [21:40] what does ifconfig say? [21:40] binBASH, oh I see -- so the server sets up the share and the client mounts the share [21:40] and the ethernet lamps do light up when i plugin the ethernet cable. one moment i will see what ifconfig output [21:40] savid: Yup [21:41] cloakable: ifconfig does not exist :-( [21:41] any other suggestion [21:41] tyska, the general process shown there will probably work. [21:42] smoser: yeah, but i will install debian squeeze. do you know some way to install lucid instead? [21:42] s/squeeze/lucid/ [21:42] okay with dmesg | more i have found now some lines that are related to the network [21:43] aha [21:44] TCP Cubic registered [21:45] Net registered protocol family 10 [21:45] hmm [21:46] lo: disabled Privicy extensions [21:46] Net registered protocol family 17 [21:46] Using IPI no shortcut mode [21:47] that is more or less all [21:48] xperia: there's no ethernet adapter listed in lspci, that would explain why the system can't find the driver... are there any "unknown", or similar, lines ? [21:48] oh my [21:49] will look again but i did not find any line that would have words like "eth" or net [21:49] xperia: is that a very old server? [21:49] it has a 3ghz xeon Processor 6 GB Ram [21:49] and a lot of Harddisk space [21:49] buyed it over ebay for 60 USD :-) [21:49] it works great. had windows on it [21:50] xperia: I'm asking because HP's site has no drivers past RHEL4 [21:51] i have read on the ubuntu forums that other people have succesfull installed ubuntu [21:51] xperia: but, according to specs, it should have an e1000 or tg3 .. which are pretty typical adapters for the time [21:51] Maybe its disabled in the BIOS? [21:51] on the "Proliant ML530 G2" server [21:51] hmmm okay will check out the Bios then === dendrobates is now known as dendro-afk [21:55] SpamapS: Bingo ! you was Right ! Compaq NC3163 Fast Ethernet NIC deactivated [21:56] xperia: ^5 [21:58] binBASH, so, do I even really need a glusterfs client? If all I'm doing is keeping files in sync accross the pool, shouldn't that work as long as I have the just the glusterfs server installed on all servers? [22:01] savid: no [22:01] savid: you need the client to mount the exported filesystem [22:02] I'm still confused then. [22:02] savid: it doesn't "keep the files in sync" [22:02] The directories that contain the files already exist [22:03] savid: it is a clustered filesystem, the data is spread out over all servers [22:03] SpamapS, ok, then maybe glusterfs is not what I need [22:03] oh [22:03] I see. [22:03] savid: so there is a single filesystem that appears the same on all servers [22:05] ldapsearch: not compiled with TLS support .. ldap from ubuntu lucid... do i need compile ldap to get TLS support? i don't understand why ldap from ubuntu-server 10.04 has no support for [22:05] SpamapS, Ok, so I would have a client and server on every server? [22:06] savid: if you so choose.. you can, of course, have more clients than storage servers. [22:06] savid: or more storage servers than clients [22:06] So, I guess I need to make sure the "volume" that I set up on each server is set up using the same parameters? [22:06] ie, when I do glusterfs-volgen [22:06] savid: you need to read the manual. :) [22:06] savid: like, the whole thing. :) [22:07] http://www.gluster.com/community/documentation/index.php/Storage_Server_Installation_and_Configuration [22:07] at least chapters 1 and 2 [22:08] Yeah, I did read through that.... [22:08] oooh I think I see now [22:08] savid: and then this one too: http://www.gluster.com/community/documentation/index.php/Client_Installation_and_Configuration [22:08] New bug: #585053 in ntp (main) "no user interface to configure whether time synchronisation is always made on establishing the network connection." [Low,Triaged] https://launchpad.net/bugs/585053 [22:08] I run the daemon on each server, and then run volgen once and that sets up the volume that's distributed across all servers? [22:08] savid: right [22:09] * savid 's light bulb finally goes off [22:09] savid: you can put the "client volume file" on all the servers, and then when you point clients at it (mount -t glusterfs server_ip /mnt/clustered_data) it will "just work" === dendro-afk is now known as dendrobates [22:11] one thing I don't know is how to handle adding/subtracting servers [22:14] question. does anybody know where to download the the cd rom image "SmartStart V5.5" [22:14] have set the Bios Password and full forget it now [22:14] for the server :-( [22:17] xperia: http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&cc=us&prodNameId=3285467&prodTypeId=15351&prodSeriesId=316545&swLang=8&taskId=135&swEnvOID=2025 [22:17] * SpamapS should definitely have bit.ly'd that doh [22:19] hp website is not SEO [22:19] smoser: do you know how can i use VNC to connect in the instances? [22:22] !samba [22:22] Samba is the way to cooperate with Windows environments. Links with more info: https://wiki.ubuntu.com/MountWindowsSharesPermanently and https://help.ubuntu.com/10.04/serverguide/C/windows-networking.html - Samba can be administered via the web with SWAT. [22:22] lenios: Search Engine Optimized? HP's website is better, it is SGWSO (Super Genius Web Surfer Optimized!) [22:24] SpamapS: woow you are great man. need to learn how to search better in google. just downloading it at the moment. very slow however [22:24] bookmarked the site just in the browser. looks like need to flash the bios tooo [22:25] xperia: that bookmark probably won't work after a while [22:25] HP's website is weird that way [22:25] oh, you bookmarked it, you could have just remembered the url [22:25] google "Proliant ML350 G2" was enough [22:25] lenios: I dare you to remember the *hostname* of that url tomorrow. ;-) [22:25] hmm i have searched for SmartStart V5.5 [22:26] xperia: with HP, its all centered around your model number [22:27] maybe it's easier to remember the ip [22:28] hope to remember the password again for the server. i wanted it to write it down but thinked ehh this thing i can remeber in some hours later without any problems and now exactly i cant remeber that thing [22:33] tyska, you cannot without hacking [22:33] or enabling vnc server in the guest [22:33] (ie, installing vnc server and running one) [22:38] hmm, I don't see the command glusterfs-volgen in ubuntu === Thell is now known as Thell|Away [22:58] is there output to the serial port during boot? [23:00] bondiblueos9: you'd need to setup a serial console in grub to do that. [23:02] bondiblueos9: you can add it to the commandline for the kernel.. console=ttyS0,115200n8 .. make sense? [23:02] what sort of interface would it give me? I mean, would it be a full terminal? would I be able to see all the services starting up (or failing to start up)? if it boots but ssh or the network is down, will I have a full terminal? [23:02] bondiblueos9: if you set the console to the serial port, then instead of the keybaord/monitor, you get all the boot messages on the serial port. [23:03] ok; do I set that in grub.conf? can I have both ther serial port and the monitor? [23:07] So... An Ubuntu client can connect to an Ubuntu Samba server - A Windows 7 client can connect to it and see shares - but can't connect to any of the shares. Any ideas what I'm missing here? [23:09] smoser: if i install vnc server on the instance and run it, should i get connect directly or i have to do something else? [23:15] SpamapS, is the glusterfs-volgen command supposed to be included with gluster-server ? [23:17] of course, now that I went and got the montior, cleared off space to set it down, and hooked it up; ssh is working fine [23:19] savid: dpkg -S `which glusterfs-volgen` [23:19] glusterfs-client: /usr/bin/glusterfs-volgen [23:20] savid: technically you don't need -volgen to run a server.. so it makes sense it would be in the client package [23:26] SpamapS, I have glusterfs-client installed... odd. [23:27] no /usr/bin/glusterfs-volgen [23:31] savid: on lucid? [23:31] savid: dpkg -L glusterfs-client .. [23:31] .. not included ;) [23:35] apache is telling me that it can't reliably determine my server's fqdn, even though i have a servername set in my vhost file. what's going wrong? [23:35] http://dpaste.com/198734/ [23:35] SpamapS, I'm on 9.10 [23:36] StrangeCharm: run 'hostname -f' .. if there are no .'s in there.. you don't have a real server fqdn. [23:36] savid: *ahh* [23:36] StrangeCharm, maybe /etc/hosts is wrong [23:36] SpamapS, that command returns "hostname: Name or service not known" [23:37] savid: consider going to 10.04 .. seeing as it will be supported longer, and includes glusterfs 3.0, which is what you really want at this point. [23:37] lenios, i'm not sure what to do with /etc/hosts [23:37] SpamapS, hmm.. is it difficult to do a dist upgrade? will anything get messed up? [23:38] you should have your hostname is there associated to your ip [23:38] savid: it *should* work fine, but it might break stuff.. so if the boxes are already production.. I'd be cautious. [23:39] in* [23:39] hmm [23:39] lenios, if there are multiple subdomains, should they be listed independently? [23:39] savid: the usual "do a full backup first" disclaimer should be enough to keep you safe. ;) [23:40] savid: also backporting wouldn't be all that hard. [23:41] only the domain, i guess [23:41] lenios, what if multiple subdomains map to other hosts? [23:42] i didn't go that far, use a search engine on how to configure /etc/hosts [23:42] fair enough