[00:14] Plz help to create this a new server to filter this net work [00:14] How do I mount an external hardrive in ubuntu server edition? [00:15] I using ubuntu 10.10 [00:15] I install in hard drive internal [00:17] i need help [00:17] :S [02:15] Hey all, having some issues with a RAID array, wondering can anyone help.. [02:17] post the issue [02:19] I have a server with a 5 disk RAID 1 array that's resyncing. THe resync seems to have frozen though, the speed according to /proc/mdstat is 0k/sec, a cron rsync from another server has also frozen and this is causing the load on the RAIDed server to be huge, >5 with little or no CPU [02:19] Also, if I attempt to reboot, it stops and doesn't complete because it can't unmount the drives [02:43] Hi all! Ubuntu server 10.04.1 Followed official documentation to setup LDAP authentication. Now if i login as a LDAP user i got "i have no name!" in command prompt. Also passwd gives error on LDAP users. /etc/ldap.conf is world readable. Anyone? [02:46] <_Techie_> how to add a command to the end of runlevel 2, the ubuntu way? [04:25] Hey.. during the Ubuntu install, is there a place where the errors are logged? [04:25] that i can get a console and look? [04:26] sparc: vt4 [04:26] I'm on this out-of-band management console [04:26] Also /var/log/syslog [04:26] and i can't see the right half of the screen [04:26] hmm ok, i did see syslog [04:26] That is, prior to finishing the install [04:26] maybe i can find a way to make it wrap after 30 columns or something [04:26] export COLUMNS=30 maybe? [04:26] yeah, I get No root file system dr- [04:26] Please correct this from - [04:27] heh [04:27] Oh, you're THAT guy [04:27] oh :( [04:27] possibly [04:31] i'm not trying to be upsetting [04:33] Sure, I just don't feel like dealing with your problem [04:33] Someone else might [04:33] i didn't ask you specifically [04:33] but ok [04:33] have fun [04:37] sparc: /var/log/syslog/installer/ [04:38] fantastic, thanks i knew it was in here somewhere :) [04:38] i was looking under casper [04:38] good deal [04:45] Uh, casper's only present on the live CDs [04:46] i interest you now :) [04:56] sparc: it sounds like you installed using the desktop CD instead of the server CD [04:56] nope [04:57] thanks though [04:59] i'm just now testing my preseed config, and working on a system with poor console access [04:59] it's my fault, and i should be testing on another sytem [05:01] and thank you for listening to my issue yesterday [05:02] you probably don't often get thanks [05:02] maybe i should be less persistent [05:12] Don't worry about it [05:13] There are plenty of worse dickheads on freenode; if you annoy me enough I'll just ignore you, like I do them. [06:03] :) [06:20] i'm reluctant to ask, but it seems like partition in Kickstart is limitted to the first-lettered disk in a system [06:21] so if that ends up being a usb drive or something, then we're stuck with using partition recipes in preseed [06:22] which isn't the end of the world, of course [06:22] just the end of kickstart [06:22] gobbe :) [06:25] sparc: linux doesn't have "disk letters" [06:25] you don't have to talk down to me :( [06:25] it might be better if you put me on ignore [06:26] block devices do often have letters [06:31] cl [06:33] sparc: As in "A hard drive, brought to you by the letters d, e, v, s, d and a ? Why not rephrase your comment/question in terms that are either Linux-specific or (perhaps even better) OS-independent? Doing so might demonstrate willingness to learn... and yes, thats a known limitation of the Ubuntu "port" of Kickstart, see https://help.ubuntu.com/10.04/installation-guide/i386/automatic-install.html [06:34] i know i read that [06:35] did you also think that that wasn't linux-specific? [06:35] OK, so what exactly were you asking when you said "i'm reluctant to ask, but it seems like partition in Kickstart is limitted to the first-lettered disk in a system" ? There's no other apparent question or uncertainty in there to address, is there? [06:36] i was trying get other people's opinion about a limitation in the installtion? [06:36] do you think that's off-topic for the channel? [06:36] i could go, if you'd prefer [06:37] i'm pretty friendly, honestly [06:37] which isn't really being returned in kind [06:37] No, that's on topic, I think, but I'm not sure that's what you said/asked... [06:39] I think trying to suggest that "the first-lettered disk" is anything other than Windows-specific is... unlikely, shall we say. so I looked for a question being asked, since you said "I'm reluctant to ask"... and apparently I guessed incorrectly as to what the question was. [06:40] scsi disks get assigned letters, like, sd[a-z], as we all know. maybe i should be more specific, later on... [06:40] i'm not asking a question, just discussing [06:44] i've been automating installs and working on redhat since 2001 or so. so far this week, i'm attempting to test ubuntu on our hardware in our datacenter. [06:44] i thought it would be good to have a few discussions with people who also want to take ubuntu seriously [06:44] so i hope you don't think i'm wandering in, trying to replace vista for the first time on my desktopo [06:45] and asking impertinent questions [06:45] because that's not the case [06:46] OK. Prefixing something that you are not asking with "i'm reluctant to ask, but" may not have been clear. I think you're a bit late for most folks in the USA and a bit early for the Europeans, so not many folks are around right now. [06:46] that's okay, i'm not in any rush [06:47] I think if you need maximum flexibility in automated Debian or Ubuntu installs, you should probably avoid Kickstart entirely and use Debian-native approaches instead, but I'm not any kind of expert on that. [06:48] i agree, that sounds about like how it's turning out [06:48] the initial kickstart support seems like a good direction though [06:48] maybe in 11 it will look really nice [06:49] i was excited to hear about the upcoming Cobbler support too [06:49] it is very helpful to be able to provision quickly, and be able to present a gui to other admins to use [06:51] Yes. Looks like soren has been interested in Cobbler since 2008... https://wiki.ubuntu.com/CobblerSpec [06:53] kickstart has a "preseed" command for Debian preseed settings. [06:55] yeah, i've been specifying my preseed config as a boot option, but maybe putting some of them in the kickstart file will change behavior some [06:55] The Ubuntu implementation has a lot of limits. You only have %pre and %post sections for the most part and a few non-preseed settings like url, user, lang, timezone. [06:56] the partitioning was the nicest feature i saw in the kickstart section of the automated deployment guide [06:56] too [06:56] I don't see why you would use the kickstart support at all [06:56] me either! [06:56] hehe [06:57] AFAICT it just provides a strict subset of the preseed functionality, in a fugly .spec-like file format [06:57] it has a nice declarative syntax, which is initially appealing [06:57] sparc: preseed is inherently declarative [06:57] i guess that's subjective [06:57] I'm working on a huge kickstart script for Lucid. 99% of the effort is in the %post section. I don't mess with the partitioning as my targets are home users with a wide variety of drives and/or RAID and/or LUKS/dm-crypt. [06:58] cool [06:58] I think I used kickstart as the name reminded me of kixtart (login script processor by M$ Norway) from my Windows days. [06:58] haha [06:59] kickstart is a RH knock-off of Sun's jumpstart [07:00] Biggest annoyance is that the %post (and %pre) script sections is parsed twice before ending up in /var/spool/kickseed/parse/pre/0.script on the installer's filesystem. Every backslash has to be double-escaped (1:4 ratio). [07:01] it does work well, even tho RH gets a C for creativity [07:01] jhansonxi: ooh, yikes [07:02] I actually wrote a tool to check them as it is an easy mistake to make and it kept messing up installs. [07:02] aah what a great idea [07:02] maybe one day someone will make a lint for preseeds and kickstarts [07:03] so we know before we commit, if it's gonna blow up [07:03] hehe [07:03] i do wonder though, if there's some way to include shell from somewhere else in the %pre and %post [07:03] so you don't have to suffer from the escaping [07:04] maybe i give it a try and share here if it goes well [07:05] I also wrote a quote checking tool as mismatches were another problem: http://jhansonxi.blogspot.com/2010/12/quote-count-debugging-tool-for-shell.html [07:06] I'm signing off. It's 02:00 here in Michigan, USA and I need my beauty rest. [07:07] ok, thanks for talking about deploying ubuntu :) [07:07] have a good sleep [07:07] I'll be posting my script mess on my blog in a few days when I get some deployment testing finished. Bye. [07:07] your work with ubuntu ce looks fun too [07:08] hope that comes along [07:08] bye [07:10] Anyone ever set up their server to work with time machine? I've followed 1 tutorial it worked for a little while then stopped. [07:11] "time machine" the apple backup product? [07:11] Isn't that OS X-specific? [07:12] no it's not, time machine can use samba share as network storage [07:13] chrislabeard: i did try, but i wasn't successful :( [07:13] sparc: I got it to connect to the server and started backing up [07:14] mine wouldn't use the sparsefile bundle i made [07:14] hah way to go [07:14] and then stopped. now I can't connect to it, It asks for a password and username then just times out [07:14] http://www.kremalicious.com/2008/06/ubuntu-as-mac-file-server-and-time-machine-volume/ [07:26] I thought time machine was basically lvcreate --snapshot for HFS+, but whatever. === chrislabeard_ is now known as chrislabeard [08:28] Anyone know why my server isn't broadcasting the shared volume to my computer? I have avahi configured correctly I think. [09:02] jmarsden: 2007. I just didn't write the spec until 2008. [09:16] New bug: #702265 in samba (main) "Error with overlapping idmap uids and gids " [Undecided,New] https://launchpad.net/bugs/702265 === lau is now known as Guest81994 [10:41] hi === Guest81994 is now known as lau [11:30] hi, for some reason I can't kill apache2. it's preventing a "service apache2 restart" operation [11:30] service apache2 stop fails? with error? [11:32] oCean, no. but service apache2 start says that the port 80 is already in use, and then ps -fe | grep apache2 shows that it is still running. then an attempt of : for PID in `ps -fe | pgrep apache2`; do sudo kill $PID; done [11:32] IIRC there are several ways to restart apache2, and -- at least in the sysvinit script -- only one of them is provided [11:32] apache2 will not die [11:32] airtonix: pastebin "ps auxf" output [11:32] airtonix: try a kill -9 [11:33] http://pastebin.com/vLajmGAK [11:34] airtonix: that is not your entire tree. [11:34] why do you need to see it all ? [11:35] Because I'm the guru and you're the querent. [11:35] :s [11:36] airtonix: tried a 'sudo kill -9 1552' ? [11:36] oCean, yes with no result [11:36] apache still alive [11:38] ah, it's Ds, uninterruptable [11:39] is a server reboot an option? [11:39] unfortunately not at the moment [11:40] uninterruptable sleep means the process will not be woken up by signals. It can be only woken up by what it's waiting for [11:41] how do i determine what it's waiting for ? [11:44] very difficult to tell. In this case it's probably the defunct process, which will never return, since it's zombie already. A reboot is the way to go here [11:46] ah think i understand now [11:46] http://www.linuxquestions.org/questions/linux-general-1/kill-d-process-339750/ [11:46] it's because (stupidly) the /var/www is actually symlinked to a dir on a external usb storage device [11:47] and i vaguely remember it being disconnected during current session [11:50] airtonix: that is probably it, yes [11:56] I have to have a server built and ready to install by this weekend, and Im trying to decide whether to use the customers copy of windows 2000 server or ubuntu. The problem is, for this one server, about 20 ppl hammer MS Access databases all day, and samba seems to have had issues with Access in the past. Is that still the case? [12:01] compdoc: maybe try #samba also [12:01] I have - but its early [12:04] maybe I should find the samba mailing list [12:12] Does ANYONE actually understand how quota grace times work? [12:17] its documented [12:17] quite well [12:24] compdoc: let me know if you find anything out about Samba/Access [12:25] k [12:27] gobbe: I guess it is, at that. [12:28] Maybe I'm just grumpy because quota infrastructure feels so old-fashioned [12:34] Is it a bad idea to set quotas *in advance* for all UIDs 1000 through 9999, even though I currently only have a couple dozen users? [12:34] The aquota.users only grew from 10kB to 400kB, which is hopeful [12:45] any reason why you would think it is bad ? (adduser will run /usr/local/sbin/adduser.local when it finishes creating a user account) [12:48] airtonix: well in a naïve example, the kernel might reread that file for every write operation [12:48] My users live in LDAP, on another host, so adduser.local won't work. [12:49] (Unless I give the LDAP server RPC access (e.g. ssh) to the NFS server, which I'm loathe to do.) [13:31] Daviey: see https://code.launchpad.net/~termie/nova/db_migration/+merge/46073 [13:31] hm, wrong channel [14:10] i am trying to automate remote package upgrades via a fabric task such as [14:11] run('TERM=linux aptitude -q -y safe-upgrade') [14:11] I added the TERM=linux in order to avoid some debconf error [14:11] but I don't exactly understand it , can anyone explain his purpose ? [14:17] lau: what is TERM *before* you override it? [14:25] hi twb , according to fabric run("echo $TERM") , TERM was "dumd" before I override it [14:26] I would like to understand the relation between TERM and debconf , I am not sure I get the point [14:26] Oh, "fabric" is some kind of automation thing [14:26] It should be "dumb", not "dumd" [14:26] yes dumb [14:26] how can i copy /var/lib/backuppc/.ssh/ to /var/lib/backuppc.orig/.ssh/ ? [14:26] You should fix it by setting DEBIAN_FRONTEND to noninteractive (don't prompt at all) or readline. [14:27] you mean with DEBIAN_FRONTEND rather then with TERM=Linux ? [14:27] Yes. [14:27] Ugh, fabric is yet another NIH'd RPC framework. [14:28] ok I saw DEBIAN_FRONTEND in man 7 debconf [14:28] lau: it is definitely NOT correct to claim your terminal is linux when it only supports dumb. [14:29] cap_00: with -r flag [14:29] cap_00: cp -R or cp -a or even 'man cp' [14:30] for some reason that's only copying the known_hosts, but i got it, i used rsync -av [14:30] i wish there was just one copy command :S [14:31] that's the beauty of unix - there's always more ways to do things :) [14:31] yep [14:31] ok,,, new wish... i wish they all reacted the same way then [14:31] uuh, my order from shop.canonical.com is arrived! need to go and pick it up from post :-) [14:32] with DEBIAN_FRONTEND=noninterractive is dpkg impacted to ? I mean if a conf file changed does dpkg still ask to keep it or override with maintainer version ? [14:32] cap_00: what you are trying to achieve? [14:33] i had two external backuppc drives, i created the .ssh key for rsync backuppc, and i needed to put that same .ssh folder on the other backuppc drive [14:34] for some reason cp -r only copied the known_hosts and not the id_rsa or id_rsa.pub [14:34] but rsync -av did [14:48] how did you run it [14:49] it should copy all files [14:53] twb: don't know if you answered my last question do you ? [14:53] rsync is great [14:54] lau: DEBIAN_FRONTEND applies to debconf, which is invoked indirectly via dpkg. [14:54] lau: debconf is the thing that asks all the questions [14:54] lau: I don't know if it affects dpkg's "replace local conffile with new maintainer copy?" question.. I've never thought of that before... [14:55] lau: #debian or #debian-mentors on OFTC might know. [15:21] So it's my girlfriend / partner's birthday. The only present she is asking for is to have access to all her data in a common pathname across our network, regardless of which machine she is sitting at. I understand that openldap can accomplish this. Is this correct? What exactly is this feature called? (OK, so she actually also wants to paint our living room, so I guess that's two birthday wishes. But I think I can accomplish [15:21] this one.) [15:22] New bug: #702404 in spamassassin (main) "cannot change scores in /etc/spamassassin/local.cf/local" [Undecided,New] https://launchpad.net/bugs/702404 [15:23] jmyles, roaming profiles is one way [15:29] jmarsden: what client OS? [15:30] jmarsden: sorry - wrong nick - jmyles seems to have left, and autocomplete chose you :P [15:34] RoyK: I'm here - I had to reboot my gateway. What did i miss? [15:38] i'm looking at doing the same thing shortly, i have windows only clients [15:39] jmyles: what client OS are you using? [15:39] does this http://dpaste.com/313319/ mean anything bad. (from auth.log) [15:39] Royk: always ubuntu [15:40] then a shared nfs storage on a server should work fine [15:40] either static mapping or with automount [15:41] cap_00: for windows clients, see http://wiki.samba.org/index.php/Samba_&_Windows_Profiles [15:43] Royk: Sounds great! My only question is this: I see that it's easy to use NFS to 'share' a directory on the server and then have her manually mount it on whatever client, but how can I make it mount by virtue of her logging in? I need ldap for that, right? (Is that what you mean by 'automount')? [15:44] automount is a service for automatically mounting a filesystem when a dir is accessed [15:44] ldap (or nis) is needed for a shared user and group database [15:45] in a small environment, having the same (numeric) userids on local accounts should suffice [15:45] I'm trying to install dsniff on a lab machine, but it requires libdb4.6. Ubuntu tries to download/install libdb4.6_4.6.21-16_i386 but receives a 404, browsing the repository I see there's only a libdb4.6_4.6.21-17_i386, package lists are updated, how can I force it to install the other version? [15:45] say, user roy has uid 110, create user roy on another machine with the same uid, and nfs will allow access to that file [15:46] uid-based permissions are really simple [15:46] and quite slack :P [15:46] guhcampos: try another repository - probably something wrong there [15:47] RoyK, it's one of the official ones, it shoule be more reliable =( [15:47] guhcampos: well, it should, but system failures happen to _all_ systems, official or not [15:48] well I agree [15:48] that's just weird [15:48] Royk: Great, thank you very much. I will endeavor now. :-) [15:49] I thought that as libdb4.6 is quite old maybe the maintainer just neglected it [15:50] guhcampos: also, if you can browse and find the file, just download it and install it with dpkg [15:50] does this http://dpaste.com/313319/ mean anything bad. (from auth.log) [15:52] resno: It just looks like what happens there when apache starts up [15:52] genii-around: ah lol [15:53] well, i see it happening daily.. i hadnt done that [15:59] resno: iirc logrotate will restart apache for you :P [15:59] interesting [15:59] gotta do that to make apache release the old log file [16:00] how long is apache down? [16:00] otherwise it'll keep on writing to the rotated one, even with a new name [16:00] few seconds im guessing [16:00] resno: probably not even measurable [16:00] an apache2ctl graceful is run iirc [16:00] and that won't be noticable [16:00] oh, even better ok [16:00] for sure [16:00] much better then running a restart or something [16:01] thanks RoyK === unreal_ is now known as unreal [16:03] robbiew: hi! [16:03] robbiew: how are you doing? [16:04] mathiaz: o/ [16:04] SpamapS: o/ [16:05] SpamapS: how is Dallas doing? [16:05] robbiew: I've finally figured out to transfer the ubuntuserver blog - should I transfer it to you? [16:05] SpamapS: ^^? [16:05] mathiaz: quite well... though karaoke last night may have slowed us down a bit today. ;) [16:06] mathiaz: I believe we're transitioning it to an atlantic.net instance [16:07] SpamapS: ok - so wordpress.com won't be used anymore? [16:08] mathiaz: that seems to be the idea yes [16:08] mathiaz: yeah...you can transfer it, but we are moving away from wordpress.com....most likely this week [16:08] robbiew: ok [16:14] Hello, May some one knows, why iftop don't shows anything? [16:15] interface are correct (ifconfig shows that there is activity on venet0), but iftop -i venet0 don't show any thing. A bit strange? [16:21] New bug: #702433 in openssh (universe) "Graphical SSH passphrase prompt does not come up in front/foreground of guake terminal (focus)" [Undecided,New] https://launchpad.net/bugs/702433 [16:45] anybody knows if theres a way to make multipath use alias directive as mapper name (alias yellow -> /dev/mapper/yellow instead of /dev/mapper/mpath0) ? /var/lib/multipath/bindings says not to edit it by hand [16:47] lambda_x, this can be done via a multipath.conf file [16:48] ignarps: yup im setting multipath { wwn ... \n alias ... } but it still sets it as mpath0 [16:48] ignarps: even when i do multipath -F, rm /var/lib/multipath/bindings, multipath -v2, it gets mpath0 name [16:50] can you post your multipath.conf file ? [16:50] do you have alias defined for a multipath ? [16:55] ignarps: woops, it just started to work, I assume [16:56] it happened because it was mapped previously as mpath0 [16:56] at least it is working [16:58] I still got one more question - volumes from FC array are visible as /dev/sda /dev/sdb etc, those are available under /dev/mapper/yellow now, when I try to use LVM - for instance vgdisplay I get /dev/sda: read failed after 0 of 4096 at 0: I/O error, but vgdisplay should not use /dev/sda [16:58] it is just a warning, it reports correctly, for different volume (in this case /dev/sdf1) but those warnings are a bit annoying [17:06] New bug: #702472 in bind9 (main) "package bind9 1:9.7.1.dfsg.P2-2ubuntu0.1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/702472 [17:12] hey folks, I've got a windows server 2008 active directory domain with ubuntu clients joined using likewise-open. Is there any way I can block traffic on all ports on an interface unless the remote host is a member of the domain? Trying to configure good security for my cloud servers that I won't constantly have to change when I add a new server. [17:32] psyferre: 802.1x? [17:35] mathiaz: good morning, can you accept zul on uec-testing-scripts-dev? [17:38] RoyK: interesting... I hadn't thought about something like that. I'd been trying to find some sort of firewall package that would support it. [17:38] it's not really a firewall thing [17:38] yeah [17:39] That's why my googling was coming up pretty empty :) [17:40] http://www.stevens.edu/itwiki/w/index.php/Linux_802.1x [17:40] So, provided I can learn enough about this to understand what I'm doing... I could setup my domain controller to be a RADIUS authentication server and then set my machines up to accept only traffic from authenticated sources? [17:40] that was my thouht [17:40] I've never setup such a system, though [17:40] nice [17:41] thought, even [17:41] this keyboard isn't as sensitive as the old one :P [17:41] * RoyK just got a new i7-base 17" MBP :D [17:41] I hadn't really thought about it, but I guess I could also just make the DC the gateway for a VPN, couldn't I? And join the machines to that? Or would that be more overhead that I don't need.... [17:41] based, even [17:41] RoyK: :) [17:41] damn [17:42] niiiiiiiiice [17:42] RoyK: our company is moving to apple (mac laptops and iphones), so i picked up 11.6" macbook air for my new machine :) [17:42] It's just too fast for your fingers [17:42] cute :) [17:43] I'd say using radius or ldap auth for the vpn should be doable, but using the DC as a gateway won't be my first choice [17:43] mickysoft tend to say those things should be dedicated, for some reason [17:43] For security reasons? Or performance reasons? [17:43] i understand [17:43] psyferre: just for support reasons [17:43] yep [17:44] dedicating expensive hardware for a DC is outright stupid imho, but I guess MS says so to avoid problems [17:44] and DC's aren't even heavy cpu/memory/network usage [17:44] not at all [17:44] that's really stupid [17:44] Exactly... I've been trying to think of other uses for the server [17:44] perhaps mickysoft has deals with dell/hp/ibm etc :P [17:44] i try to force people to think that opensource might provide all tools needed [17:44] It's a cloud server, so I'm not wasting hardware on it, but stilll... [17:44] it's so easy to just pick windows [17:45] gobbe: AD doesn't run well on linux, you know [17:45] Yeah... I tried and tried to get a domain controller on ubuntu with zentyal [17:45] it was a nightmware [17:45] and AD is a hell of a tool for managing a bunch of windows clients [17:45] nightmare, even [17:46] RoyK: AD in many companies are just for user database [17:46] psyferre: there are some stuff in the works for samba4, but I doubt it'll stabilise this month, perhps not this year [17:46] RoyK: and there are quite many opensource ones for that :) [17:46] gobbe: if you have 200 windows clients, you want AD, beleive me [17:46] RoyK: that's true. but if you have 200 linux/osx machines, you dont [17:47] nope [17:47] RoyK: yeah, I saw that and got pretty wistful [17:47] but hoever still many companies build ad and try to put machines there :) [17:48] Thanks for the advice, guys. I really appreciate it. Enjoy the new laptops :) [17:48] :) [17:49] i don't have my new yet [18:07] New bug: #694059 in qemu-kvm "qemu fatal cp15 message report and image creation block" [Undecided,New] https://launchpad.net/bugs/694059 === luis__lopez is now known as luis_lopez [18:54] hggdh: done! [18:54] mathiaz: thank you [19:27] why does ipmitool showing all the dimms are not readable ? [19:29] dmidecode shows they are installed.. do I need to modprobe any kernel for that? [19:33] headphone not working in laptop using ubuntu 10.10 [19:37] headphone not working in laptop using ubuntu 10.10 [19:37] mogaj: wrong channel [19:45] !apt-fix [19:46] If an APT front-end crashed and your database is locked, try this in a !terminal: « sudo fuser -vki /var/lib/dpkg/lock;sudo dpkg --configure -a » [20:38] hi, after updating my server today (apt-get update) service apache2 start does not work anymore, the reason is obvious when looking at /etc/init.d/apache*: [20:38] -rwxr-xr-x 1 root root 0 2011-01-09 06:25 /etc/init.d/apache2 [20:38] -rwxr-xr-x 1 root root 6157 2010-11-18 22:18 /etc/init.d/apache2.1 [20:38] what has happened there? and how to fix it? [20:55] why does the mailx command work from the terminal of my vps, but I cant send via Thunderbird?? Any help? I've been on the IRC for linode all day but I think they've been busy with other things. I can receive via Thunderbird. Also, I used this tutorial for setting up my vps: http://library.linode.com/email/postfix/dovecot-mysql-ubuntu-10.10-maverick [20:55] webroasters: perhaps your local MTA has a problem? [20:55] what do you mean [20:57] i know what MTA is, but local? [20:59] exim or postfix [21:00] postfix [21:00] check the logs [21:00] mailq will list queued emails [21:01] http://pastebin.com/cH7XJhdb ..oh, the mail queue is empty [21:01] that's from mail.log [21:05] mathiaz: thanks [21:05] zul: you're welcome! [21:12] kirkland: when you have the time can you sync this plz? bug #702593 [21:12] Launchpad bug 702593 in corosync "Please sync corosync 1.2.1-4 (main) from Debian unstable (main)" [Wishlist,Confirmed] https://launchpad.net/bugs/702593 [21:16] another question. I'll come back to the other one. How does one have 2 ssl certs on the same server? Is there something I'm missing from the virtualhosts file? I have *:443 set on both vhost files. [21:18] webroasters: You need each one to be on a different IP address, if they will both use port 443. [21:18] is there a way to do with virtualhosts? [21:19] No, one port on one IP == one SSL certificate. SSL handshake happens before any info about which virtual host is being requested is transferred, so you can't have 'virtual HTTPS hosts' sharing a port. [21:20] You can put one of them on port 444 and use https://example.com:444 if you want. [21:21] is there a way to cover up :444 with mod_rewrite?? [21:24] I don't think so. You can play with SSL certs with alternate names, if you are making your own certs, so you use one cert that handles both domain names... see http://www.crsr.net/Notes/Apache-HTTPS-virtual-host.html for some ideas on that. [21:25] webroasters: I think #httpd is a better place to ask for that [21:25] and btw, virtualhosts with HTTP 1.1 rarely works well [21:25] yeah, no worries [21:25] thx guys [21:26] Daviey: http://www.rabbitmq.com/pacemaker.html [21:26] since the initial HTTP GET must be sent securely, it's no chance for apache to find which one to choose [21:27] RoAkSoAx, thanks [21:27] I appreciate it. The other question i had was; is there a way to figure out why i can't send email with thunderbird? [21:28] Do the current plans for Natty include having MariaDB packaged? [21:28] nfi - I use postgres :P [21:30] maxolasersquad: use the rmadison command to find out: rmadison -s natty maria says: maria | 1.3.5-2 | natty/universe | source, amd64, i386 [21:30] So looks like it is already there. [21:31] jmarsden|work: Thanks. Good to hear. [21:31] You're welcome. [21:44] Daviey: want to come talk about uds? [21:51] maxolasersquad: Looks like maria is not the same thing as mariadb... my mistake. [21:51]