[00:05] how can i backup my mysql db's without a root mysql user? [00:26] billybigrigger: backing up your mysql db doesn't require a root user [00:36] What might cause apache2 to not be able to "reliably determine the server's FQDN"? My FQDN is resolving/working fine. hostname -f shows the correct resolving ubuntu01.example.lan.. [00:36] I don't understand why apache couldn't find it if hostname -f can. I would prefer to avoid adding the ServerName directive to my apache2.conf.. [00:39] unless the .len tld is resolvable ... that's not a FQDN [00:40] however, you might want to set servername in the apache config [00:40] s/len/lan/ [00:40] the .lan is resolvable though. I use it for everything.. [00:40] ok, so set servername [00:41] There is a working bind9 server up, configured to serve example.lan/192.168.156.0/24 [00:41] Yeah that worked fine, but I shouldn't have to.. [00:41] I don't know why you think you shouldn't have to [00:42] Well it's a symtom of something else being broken I think. I've never had to do that before with 8.04 [00:42] it's possible that apache is sanity checking the hostname against internet tlds to determine if it's a FQDN [00:42] hrmm.. maybe. That's still new behaviour [00:43] Oh well I guess I can just deal with having ANOTHER file I have to edit to set up servers.. [00:43] new doesn't necessarily mean a bug [00:43] Just already have like 30+ configs I need to localize for these installs, Iike to keep that down as much as I can [00:43] Oh well thanks for the input giovani, it's appreciated. [00:44] Do you know if there would be any problems with setting ServerName to localhost? [00:44] ..rather than the FQDN [00:45] localhost should be fine [00:45] Well one way to find out.. Thanx [00:53] Ah very nice, seems to work like a charm. [01:07] hey guys [01:08] I am running ubuntu server in a Virtualbox VM for development....I created a small VM with LAMP [01:09] i have two ethernet connections for ubuntu server 9.04 in a VirtualBox VM. Even if I set both connections to NAT, one is defined as eth0 and the other as vibr0. I can't seem to get vibr0 to work in any configuration. Is this a limitation of ubuntu server? It seems to give eth0 to the main ethernet and vibr to all others. [01:09] I expected to see eth0 and eth1 [01:11] I also tried to setup the vibr0 in the /etc/network/interfaces as auto vibr0 iface vibr0 inet dhcp and that failed to work as well....anyone else see this behavior? [01:12] is there a way to make vibr0 become eth1? [01:16] sorry...been typing it wrong in chat....make that virbr0 [01:18] i checked it was right in the /etc/network/interfaces [01:19] looks like if eth0 is turned off that virbr0 will not work :( [01:26] virbr is probably "VIRtual BRidge" [01:27] i am trying a reinstall with the emulated nics being different [01:31] see if I can get it to see both eth0 and eth1 [01:38] that is so bizarre [01:39] this sounds like a virtualbox problem [01:39] try asking them [01:39] why would does the install ask for the main nic [01:39] huh? [01:39] it wants to know which nic to use ... obviously [01:39] for internet access [01:40] it could set them both at inet dhcp and you would still get to the same internet [01:40] but there's no advantage to that [01:40] well I guess one of them would fail [01:40] that complicates things for no gain during an install [01:41] but both Nics have different MAC addresses....wonder why its deciding to have one eth0 and the virtual bridge? [01:41] http://www.riccardoriva.com/archives/693 I see here that you can have multiple nics...wonder why this isn't working [01:42] of course you can have multiple nics [01:44] does ubuntu-server uses just the MAC address to determine uniqueness in the nic, right? [01:44] uh [01:44] I don't know what you mean by "determine the uniqueness of the nic" [01:44] nics are hardware devices, the computer doesn't need to use a MAC to know that there's two [01:44] well the actuall machine I am using only has one nic [01:44] okay [01:45] I've recommended that you talk to the virtualbox folks a number of times [01:45] both are emulated nics [01:45] why don't you do that? [01:45] k. [02:00] is anyone aware of a way to get my phpmyadmin/mysql back to stock? [02:00] ie, clear all db's, users, and start fresh? [02:01] i've tried apt-get purge mysql-server and phpmyadmin [02:01] then apt-get install mysql-server and phpmyadmin, but my db's are all still there with the same users [02:06] billybigrigger: there's a debconf setting for mysql-server-5.0 that controls whether databases are wiped on a purge [02:07] the easy way to change that may be to dpkg-reconfigure mysql-server-5.0, and see if it asks that question about it [02:09] just asked me for a password with dpkg-reconfigure [02:18] ok new question [02:18] all my mysql locale settings are ending with _ci [02:18] and its displaying some wierd stuff for characters [02:19] err for collation [02:19] what should i be using? [02:21] ci means 'case insensitive' [02:22] The weird stuff may be a an issue of some UTF-8 chars in a latin1 encoded field. [02:23] very common. [02:24] so _ci is fine to use [02:24] yes [02:24] roger, thanks [02:24] np [02:26] also... [02:27] im trying to install roundcube...its a webmail frontend...anywho.... [02:28] its root dir is /var/www/webmail, and it needs to be able to write to /var/www/webmail/temp and /logs [02:29] what would be the best way about giving the correct perms...chown'ing it to the webserver or chmod'ing it so that anyone can write to it? [02:29] chown or chgrp to the www-data user or group [02:30] thanks again :P [02:34] anyone here use roundcube? [02:35] !anyone [02:35] A large amount of the first questions asked in this channel start with "Does anyone/anybody..." Why not ask your next question (the real one) and find out? [02:35] Oh, sorry. [02:36] Hi there. [02:36] I'm not really sure how SELinux works, but isn't this SELinux related? [2603183.964203] audit(1242955897.392:147): type=1503 operation="inode_permission" requested_mask="rw::" denied_mask="rw::" name="/home/mysql/ibdata1" pid=20006 profile="/usr/sbin/mysqld" namespace="default" [02:38] Znuff: selinux is off by default on Ubuntu. [02:39] that's what I knew [02:39] and I'm pretty sure I haven't enabled it :-/ [02:40] twb, why would I be getting audit messages in my dmesg if I don't have selinux enabled? [02:45] Presumably, they are from some other subsystem. [02:45] For example, apparmour seems to be installed by default... [02:45] yup, so it was [02:45] ^^ [02:46] Znuff: Error: "^" is not a valid command. [02:46] ow [02:46] ^version [02:46] Znuff: The current (running) version of this Supybot is 0.83.3. The newest version available online is 0.83.4. [02:46] yup.. apparmor generates audit entries in your logs [02:46] holy s* [02:46] update to supybot o.O [02:49] thanks twb & cef [03:06] who's familiar with apache's virtual hosts? [03:06] im trying to setup a subdomain for /var/www/webmail [03:07] i want to set it up as webmail.mydomain.com [03:07] not www.mydomain.com/webmail [03:07] now the apache doc's aren't clear as where to put this virtual host file or what to name it [03:07] anyone? [03:30] billybigrigger: 'man a2ensite' [03:32] billybigrigger: gives you the help for the wrappers used by apache2ctl that help manage sites. very useful. [04:08] cef [04:08] a2ensite webmail.mydomain.com [04:08] ? [04:09] hehe not much of a man page [04:09] does that create the file in sites-available? [04:09] or do i have to create the file? [04:10] ie i have nothing in sites-available, a2ensite webmail.mydomain.com will create the file, and input the correct vhost data, and then sym link it to sites-enabled? [04:10] or is all a2ensite do is sym link an already created file? [04:20] no, you need to create the file in sites-available [04:20] using a2ensite will then enable that site [04:20] (sorry, stepped away from the keyboard.. work does that to you) [04:20] ahh [04:21] hehe [04:21] hmm [04:21] I tend to copy the default file and then edit that [04:21] i see lots of examples in apaache docs that explain multiple domains for 1 ip, or 1 domain for multiple IP's [04:21] i just want to add a bloody subdomain [04:21] as everything is pretty much there and you can tweak it however you like. esp if 90% of the time they're the same details [04:21] ok, ill edit the default then [04:21] no, copy the default [04:22] yeah, my bad [04:22] thats what i meant :P [04:22] then edit it with your settings, then run 'a2en filename' [04:22] ok, 1 more question that apache docs aren't too clear on...does the name of the file have to be subdomain.domain.com or can it be ilovecheese [04:22] ? === ssd7_ is now known as ssd7 [04:23] though making the filename something that'll identify what it is in it is useful. doesn't have to be the domain name [04:23] yes, it can be 'ilovecheese' or even 'ihavenopants' if you like [04:23] and yeah, 'a2ensite fillename' even. oops [04:24] ok [04:24] so using the domain or subdomain is usefull for identifcation only [04:24] thanks [04:24] same deal for modules too, but they're usually auto-created by packages, and you can a2enmod and a2dismod them by name. [04:24] no probs [04:25] ok [04:25] whatever is in the '-available' of each (sites and mods) are available to be used, but only what is symlinked in the '-enabled' are in use. [04:27] right [04:28] ok now this submdomain could be as simple as adding ServerName subdomain.domain.com and DocumentRoot /var/www/webmail inside of tags correct? [04:29] or it can be as elaborate as the default [04:29] ? [04:31] dammit, after a2ensite, it still doesn't work [04:34] or can i do this right in my httpd.conf? [04:37] did you reload apache? [04:38] yes [04:38] the a2ensite doesnt work [04:38] hrm weird. pastebin the contents of the file [04:38] and adding a vhost in apache2/httpd.conf after a restart doesn't work either [04:38] which file? [04:38] the site file [04:38] http://guide.opendns.com/?url=webmail.thefrozencanuck.ca [04:39] err [04:39] http://pastebin.com/f6846476a [04:40] should be [04:40] oh [04:41] can i remove directory / and directory /var/www and create one for /var/www/webmail since thats the docroot? [04:41] yup [04:42] will keeping them hurt anything? [04:42] or is this config only for webmail. [04:42] like the default config in sites-enabled still overrides this config correct as 000-default is loaded first? [04:43] and do i have to do anything with my domain provider? ie godaddy where i purchased the domain...do i have to link this subdomain somewhere or is it all done through apache? [04:45] here's the updated file [04:45] http://pastebin.com/f546dd67d [04:45] do i have to run a2ensite again? [04:45] it should already be enabled, just an apache2 force-reload eh? [04:47] just reload [04:47] or restart [04:48] are vhosts handled all by apache? theres nothing i have to do in godaddy's domain manager correct? [04:48] afaik correct [04:48] hmmm [04:48] well still not working [04:49] you should just need to point the subdomain at that machine though [04:49] eg: give it a CNAME or A record) [04:49] wtf [04:49] mail.tfc.ca [04:49] works [04:49] but webmail does [04:49] i have a CNAME for mail that points to @ [04:50] do you have a cname for webmail? [04:50] no [04:50] why do mail.tfc.ca take me to /var/www/webmail though? [04:50] well, how can the dns name resolve? you need a CNAME (or A) record [04:51] possibly a config issue [04:51] i changed mail to webmail [04:51] i had a cname for mail, not webmail [04:51] but mail pointed to /var/www/webmail, does this make sense to you? [04:52] http://mail.thefrozencanuck.ca/ [04:52] haha, wierd [04:52] it means you have a config that has changed the docroot to that... [04:52] ??? [04:52] where [04:52] all configs i have setup where for webmail to /var/www/webmail [04:53] s/where/were [04:53] yeah, your www. points to webmail now too [04:53] hmmm [04:53] put defaults and webmail sites files on pastebin [04:54] http://pastebin.com/f1c46a366 <---default [04:54] http://pastebin.com/f6672cc99 <---webmail [04:55] maybe i should have left the directory / and /var/www in my webmail config and just added a directory for /var/www/webmail [04:55] ??? [04:55] im lost haha [04:59] no it looks like it might be because you don't give the default a 'servername', or because webmail.tfc.ca isn't in the /etc/hosts file and it thinks it's the default. [04:59] I've seen someone have this issue before but I don't know what fixed it (work intervened) [05:00] ok [05:00] i will give default the servername of www.tfc.ca or just tfc.ca? [05:00] i don't think i have either webmail. or tfc.ca in /etc/hosts [05:01] http://pastebin.com/f2f8a4155 [05:04] ok [05:04] adding servername to default got my index.html page back [05:04] now webmail.thefrozencanuck.ca doesn't work [05:05] Is Jamie or Kees around? I'm wondering if there's any response yet to the SSH vulnerability discussed on http://news.zdnet.co.uk/security/0,1000000189,39653852,00.htm ? [05:05] tonyyarusso: the SHA-1 vulnerability? [05:06] tonyyarusso: try #ubuntu-hardened? [05:06] twb: I think this is a separate thing actually. [05:06] That seems to be the security team's channel. [05:06] fair enough [05:13] cef::: added webmail.thefrozencanuck.ca /var/www/webmail to /etc/hosts [05:13] correct? [05:14] bah [05:14] stupid ass subdomains...who thought this would be so bloody hard [05:15] billybigrigger: um, no... [05:15] billybigrigger: subdomains go in your apache config, not /etc/hosts. /etc/hosts does name resolution for machines. [05:16] no it looks like it might be because you don't give the default a 'servername', or because webmail.tfc.ca isn't in the /etc/hosts file and it thinks it's the default. [05:16] i don't know, im reaching right now...so im trying any and all suggestions [05:16] i can't get this to work for the life of me [05:17] is there anything im missing besides my .../sites-available/webmail.tfc.ca [05:17] ? [05:17] do i need to call that config in httpd.conf somewhere? [05:17] like a subdomain should be simple, i want webmail.tfc.ca to point to /var/www/webmail [05:18] create the config in .../sites-available then a2ensite webmail.tfc.ca and then reload apache...correct? [05:20] billybigrigger: okay, sites-available should have stuff about the subdomain, and /etc/hosts should have IP addresses - could you pastebin your configs? [05:22] http://pastebin.com/f1c46a366 <---default [05:22] http://pastebin.com/f6672cc99 <---webmail [05:22] thats my /sites-available dir [05:25] aparrently it doesn't resolve...and the httpd guys are telling me to make it resolve hehe [05:26] What were the boot scripts afected by the change from bash to dash that happened in Ubuntu some time ago? Only those located in /etc/init.d ? [05:26] KurtKraut: why do you ask? [05:28] billybigrigger: Is 68.146.139.247 the correct IP address for webmail.thefrozencanuck.ca ? [05:28] yes [05:28] do i need a CNAME or an A record for webmail? maybe thats my problem [05:29] twb: I'm willing to request the same migration from dash to bash to other distros and I've realised that most scripts placed on /etc/init.d call /bin/bash directly instead of /bin/sh. [05:29] ohhh [05:29] billybigrigger: I'm reasonably certain you want an A record for this. [05:29] im about to get a boot from #httpd...hehe chalk that up as #2 for the day :P [05:29] ok [05:29] i think thats my problem... [05:29] i have a CNAME for webmail to my ip [05:30] ie....webmail @ [05:30] twb: In other to understand better the impact on making such change, I'd like to know if there other scripts related to boot that are not located in /etc/init.d that would impact on boot performance by this sort of change [05:30] billybigrigger: The usual wisdom is "If you don't know why you need a CNAME, you don't want a CNAME." [05:30] KurtKraut: well, I believe Debian recently (as in, the last twelve months) fixed that issue. [05:30] tonyyarusso::: cause i have cnames that where set when i got the domain for www and ftp....so i assumed a cname for webmail would work for webmail.mydomain.com [05:30] :P [05:30] KurtKraut: stuff like RHEL is simply fucked; they have a LOT more bashisms in their code. [05:31] KurtKraut: and they have no convention of actually, you know, making packages that adhere to a comprehensive integration policy. [05:31] twb: I see. Bu do you know what scripts are run at boot besides those located in /etc/init.d? [05:32] KurtKraut: that depends on many things. But I suspect you want the system to do more than just boot. [05:32] tonyyarusso::: why did you ask about my ip address? [05:32] tonyyarusso::: and also should the host be webmail.tfc.ca or just webmail? [05:32] billybigrigger: Because you mentioned that it wasn't resolving. [05:32] tonyyarusso::: ahh [05:33] billybigrigger: the full thing. [05:33] tonyyarusso::: k, well i added it [05:33] * billybigrigger waits [05:33] soab [05:33] webmail.thefrozencanuck.ca now points to thefrozencanuck.ca [05:33] it points to /var/www [05:33] twb: what I'm trying to test/measure is the impact of the migration from bash to dash on boot. And what I want to know at the moment is how many *boot* scripts declare they need precisely bash instead of /bin/sh [05:33] not /var/www/webmail [05:34] KurtKraut: OK, what you want to do is get the checkbashisms(1) script and run it over, basically, all of /etc/. [05:35] twb: that sounds great. Thanks [05:35] KurtKraut: from the list it emits, it should be obvious which files AREN'T used during boot. [05:35] KurtKraut: note that checkbashisms will not report on e.g. /etc/sysconfig/networking, because those files don't have shebangs -- but I bet they are sourced by a sh script. [05:36] The other obvious thing to do, is just make sh -> dash, and see what breaks when you reboot. [05:36] twb: this would be the step 2 of my experiment, seing what breaks [05:37] twb: but the first data I want to collect is how many scripts right at the begining say they need bash to run instead of /bin/sh, that would in theory and in a ideal world, mean that they don't have bashisms [05:38] i want to strangle all the smartasses in #httpd [05:38] KurtKraut: oh, then try something like grep -rn /etc '^#!.*bash' | grep :1: [05:39] Oops, swap the pattern and /etc around, etc. [05:41] twb: what a curious result. In Ubuntu, only 9 files found. In Mandriva, 290. [05:42] KurtKraut: culturally, Debian packages are encouraged to avoid bash and bashisms. [05:42] I don't think policy actually REQUIRES them to avoid bash shebangs... [05:46] twb: those data were really helpful. Thanks a lot for your help and commands. [05:47] KurtKraut: no worries. [05:48] In Ubuntu the default /bin/sh is dash, so we've been trying to get rid of bashisms for some time. [05:51] what script or process is supposed to make /var/run/apache2? [05:56] a_ok: Generally the init script should handle that. [05:56] I'm not sure specifically about apache2 === ssd7_ is now known as ssd7 [06:47] i installed squid and i want to authenticate with ncsa, my users still can connect to internet and never authenticate with ncsa. what could be wrong? i use 8.04.2 [06:48] i followed some tutorials to edit my squid.conf and it's still no result [06:51] !squid [06:51] squid is a caching proxy for the Web. See: https://help.ubuntu.com/community/SquidGuard See: http://www.squid-cache.org [06:51] !ncsa [06:51] Sorry, I don't know anything about ncsa [07:02] anyone? [07:05] NCSA? The people that brought us the Mosaic browser in 1993? http://www.ncsa.uiuc.edu/Projects/mosaic.html ? === jmarsden_ is now known as jmarsden [07:07] * jmarsden would suggest not using a discontinued and very very old browser...? Or is the NCSA you refer to something else? [07:07] I think he means an auth mechanism [07:08] what's the MS one? [07:08] ntlm? [07:09] cef: Yes, but that means he got only one out of 4 letters correct in the acronym, which is pretty poor spelling? [07:09] jmarsden: I've seen worse.. ;) [07:11] Maybe something like http://www.opensourcehowto.org/how-to/squid/squid-with-ntlm-authentication.html would help, then? [07:12] wow, NSCA also made a httpd server. never knew that [07:13] Yes, wasn't that the one the pile of patches turned into "a patchy" server, which became Apache? [07:14] probably [07:15] I ran that, back in about 1994 or so, on NetBSD :) [07:15] ahh back then, I was still running a BBS.. those were the days [07:20] it's solved [07:21] unewbie: Are you really using NCSA ? The web server or the Mosaic browser? [07:22] i put the http_access allow ncsa_users exactly below the acl ncsa_users proxy_auth REQUIRED [07:22] i used ncsa module at /usr/lib/squid/ncsa_auth [07:22] hrm! there ya go (??) [07:23] thanks guys [07:23] :) [07:27] oh bah..it's just a module that reads any ncsa compliant passwd file (eg: htpasswd) [07:28] Yes... squid "NCSA" auth is what most of us would call "Basic Auth" [07:28] yup [08:09] i'm thinking using ldap authentication [08:09] any advice? [08:29] unewbie: If you have to authenticate using LDAP before you can think, you are in deep trouble :) [08:31] I need to go to sleep... but if you already have an LDAP server set up and working, it probably makes sense to use it for Squid auth too. If not... it's a fair amount of work to set one up. [08:32] New bug: #379320 in mysql-dfsg-5.1 (universe) "mysql-server-5.1 not installed" [Undecided,New] https://launchpad.net/bugs/379320 [09:15] jmarsden: thanks for the advice [09:16] looks like i need someone to teach me to install and configure openldap :D [09:55] New bug: #379329 in openssh (main) "Security flaw in openSSH prior to 5.2" [Undecided,New] https://launchpad.net/bugs/379329 === yann2_ is now known as Yann2 === Thirsteh is now known as thirsteh [13:13] Hello. I keep reading articles at HowToForge.com and most of them say to disabled app armor to run thing in a chroot jail, but is this really necessary? [13:41] anyone familiar with ispconfig [13:41] 3 [13:43] i need help with setting up ispconfig3 and godaddy [13:43] am i talking to ghosts [13:43] hell? [13:44] hello? [13:45] infekteddeath: Ubuntu doesn't (last i looked) ship an ispconfig package, so this may not be the best place to ask. [13:50] does this channel deal with just the software side of servers? [13:50] it deals with the content of the inhabitants brains more or less [13:51] and I doubt that's limited to software only. [13:51] :) [13:53] alright [13:54] well i'm looking at picking up a multiprocessor server, supposedly rated at 550mhz [13:55] would that be the same effectively as having a single 2.2ghz processor? [13:55] and would ubuntu run on such a server? [13:56] How much RAM? [13:56] 3.5GB [13:57] It's not the same as a single 2.2ghz processor. In some ways it will be better, in some ways worse. [13:58] Do you know if the BIOS is from before 2000 or after? [14:00] timedout_: If it's before 2000 (and thus won't use acpi), Hardy is the last release it will run on. If it's new enough to use ACPI, then it should run on any Ubuntu release. [14:01] timedout_: i'd say that there isn't that much value in a old PIII with that low frequency. [14:01] and ScottK is right. [14:01] thanks [14:01] ewook: it's bascially just for learning purposes [14:01] It's fine for that. [14:01] timedout_: for learning it's neat :) [14:02] I have an old dual PIII 450 machine that I use for backups (new hard drives). [14:03] New bug: #379412 in openvpn (universe) "openvpn server startup script broken" [Undecided,New] https://launchpad.net/bugs/379412 [14:04] anyone know of some resources for finding used network components? servers, routers, switches, etc [14:04] other than craigslist [14:04] Ebay? === asac_ is now known as asac [14:39] has anyone install cacti on 9.04 [14:39] ? [14:39] I seem to have an issue where it is not showing graphs on the main page but if I click on the missing graph like object it shows graphs [14:45] ah screw it. I am re installing from source === Nicke_ is now known as Nicke [15:06] ivoks, heya master... have some good new for you :) [15:07] RoAkSoAx: do you? [15:07] RoAkSoAx: i can't merge anything now... i'm at the airport [15:09] ivoks, not merge related... last night I've talked with horms, the heartbeat Debian mantainer (i think hi's also upstream??)... and well he said something that will make our lives easier [15:10] what? [15:10] horms was at last uds, iirc [15:10] ivoks, they are waiting for final test on heartbeat / pacemaker / openais to get them in Debian archives. [15:10] oh, nice... [15:11] ivoks, and they are going to create a debian-ha group to maintain heartbeat / pacemaker / openais ... and I also said we had the same project, and that now we can work close together [15:12] great [15:12] i was expecting something like that [15:12] ivoks, yep, I'll email you my conversation [15:13] great [15:13] <\sh> ivoks: on your way to pretty barcelona? :) [15:13] Today? I hope not. [15:14] \sh: yep [15:14] ivoks, done :) [15:14] \sh: i'm in bonn [15:14] <\sh> ivoks: WHAT? just drive 300km down to KA and visit me ;) [15:14] :D [15:14] \sh: if only you said that 3 hours ago :) [15:15] <\sh> ivoks: and it's cologne/bonn airport ;) it's more cologne then bonn ;) [15:15] \sh: i doubt i'll make it to you and back in 60 minutes, without a car [15:15] \sh: i know it as Koeln :) [15:15] <\sh> ivoks: yeah...well hopefully next year I'm able to attend another UDS myself.. [15:16] ivoks, and here are the packages that might go to Debian archives: http://packages.vergenet.net/experimental/ [15:16] <\sh> ivoks: drink some "kölsch" beer ;) (1. reissdorf 2. Früh 3. Sion) ;) [15:16] those are apt-gettable so we can test them [15:17] RoAkSoAx: great job! [15:17] \sh: on my way back, maybe :) [15:17] <\sh> .oO(dojotoolkit package looks like it's ready for upload...just a few lintian messages to fix) [15:19] ivoks, so anyways... as you said, our cluster stack should be around as heartbeat / pacemaker /openais as long as Debian supports it [15:20] <\sh> guys...any status on mysql-cluster for karmic? [15:20] what's wrong with it? [15:21] time to go... [15:21] take care guys [15:22] ivoks, have a good flight [15:22] thanks [15:22] <\sh> ivoks: give a hug to ogra from me, pls :) [15:23] hehe [15:26] anybody using rabbitmq on ubuntu intrepid? is there an apt repository? [15:26] I need to do RAID 1 with 9.04..Do I need the Ubuntu-server standard or the alternate? [15:41] To answer my own question, either Server Edition or alternate will do RAID 1. User Server Edition... [15:48] Hello. I keep reading articles at HowToForge.com and most of them say to disabled app armor to run thing in a chroot jail, but is this really necessary? [15:59] hello [15:59] is there a hook, there I can hook in on every user logoff (forced by timeout, or manually by userrequest)? [16:05] hkais: http://thedaneshproject.com/posts/how-to-set-default-session-timeout-in-linux/ [16:05] hkais: worth trying [16:08] heath|work: Pardon? [16:55] I have mysql 5.0.51a installed on ubuntu server 8.04. when I attempt to stop mysql using /etc/init.d/mysql stop I get the message * Stopping MySQL database server mysqld [fail]. I'm trying to track down why it's failing and how to fix it [16:55] note: I have asked at #mysql and they recommend that I ask here. [16:55] also note: /var/log/mysql/mysql.err and mysql.log are both 0 byte files [16:55] also note: syslog makes no mention of mysql [16:56] correction: /var/log/mysql.err and /var/log/mysql.log are 0 bytes. there are no files in /var/log/mysql/ [17:01] zanberdo: on ubuntu, mysql logs to /var/log/syslog by default [17:02] right, so I've read. sadly, I'm getting no data in syslog [17:02] relating to mysql [17:02] it's clearly running [17:10] zanberdo: This sometimes happens if for instance the pid file is bypassed. If you manually ran the mysqld directly for instance instead of using the init.d method. In this type of case you have to manually kill the process and start it properly from the init script. [17:11] genii, thanks for the suggestion. No, I did not start it manually, though I suppose I will kill the pid's and see if I can start it again. [17:19] hi all, having a little problem with cups on my server, i activated remote administration throuh cupsctl --remote-admin but when i try to login, all i get is a measly 403 forbidden :( any ideas where i can fix it, or what i have to put into the cupsd.conf file to make it work? this is before even the login page is shown [17:19] maybe it has to do with the installed apache server? [17:21] or that i am in a different subnet....hmmm [17:23] jeiworth: have you tried to restart the service? [17:23] cups have apache style configuration file, so maybe you should check there === yml_ is now known as yml [17:27] it is documented in ubuntu server guide [17:27] radovan: hmm coming to think of it, i didnt restart it after the cupsctl command, i thought cupsctl would take care of that for me, let me check [17:27] the thing about cups [17:32] well, appears to be the subnet, if i access from a virtual machine in the same subnet i can configure it [17:40] i'm looking for a tool that let's me send a popup msg to all systems on my lan, e.g. "Rebooting proxy in 5 minutes. Please stand by." any ideas? [17:41] dayo: what kind of systems? [17:42] jmedina: ubuntu desktops and laptops. 8.04 [17:48] dayo: so you want to send this messages when you reboot your proxy (squid)? [17:48] if so, you can redirect your users to a mainteinance web page [17:49] * jmedina loves virtualization (cero downtime) [17:49] jmedina: that was just an example. sometimes i need to let all my users know things like "The ISP is here. Expect about 20mins downtime, while we install the new router" stuff like that [17:50] mmm then I dont know [17:51] we use internal IM with jabber, I can send broadcast messages [17:51] or you can try something like this: [17:52] autologin with ssh and something like: "logger -t adminmessage -p local7.emerg "The ISP is here....!!!:" [17:53] by default most syslog daemons logs emerg messages to everything [17:53] for example, KDE launches a little window with the message, not sure about gnome [17:54] then you can use something like parallel ssh connections and automatic login using rsa keys [17:57] dayo: http://www.manpagez.com/man/1/xmessage/ [17:57] when I install postgress with apt, there is not /etc/postgres/* directory, and nothing works...I am stuck here, asking for direction, why would this happen, how to fix it? [18:02] could postgres have used another directory for the config files for some reason? [18:16] 'lo all. if you could bear with me...I haven't used IRC for about a half decade [18:17] I was sent here by #ubuntu [18:17] I'm the tech guy at a school. Really new to linux. I purchased 4 Asus eee PC 4G. Hated the native OS so I installed Ubuntu and followed the instructions to get the Kernel that works best with the eee pc's. I'd love to be able to use our server (2003) to allow the students to login with their windows username. I discovered the activedirectoryhowto on the wiki. Unfortunately it states... [18:17] ...that I need ubuntu server to connect to AD. Is there a way that I can connect to AD with Ubuntu Desktop, or is there an easy way to configure the server to have a x windows environment? [18:17] I've currently got server installed on my test eee box, but before I blow away another, I'd like to ask those that know more than I. [18:17] I was told that desktop and server are the same, but server has some extra packages installed [18:18] AD is ldap [18:18] Could anyone point me towards a wiki that would explain what to install to get a GUI environment to talk to my server2003 box? [18:18] login go throug PAM, so you need to update your pam config [18:19] or....is it possible to have an eee pc talk with the server via wifi at boot so that they can login with their AD username and pwd? [18:19] http://ldots.org/ldap/ [18:19] section 2 is interesting one [18:20] looking at it [18:21] afaik there is no exact howto do it. maybe google will help you, but AD is ldap. [18:21] without creating users on each eeepc will they be able to log in? [18:21] or would it be easier to create users on a box and image the other comptuers? [18:22] yes, they will be able to log in [18:22] sweet [18:22] second question, if you've got time [18:23] yes? [18:23] once logged in, would I be able to connect them to their windows home folder on the server, or would I need a separate NFS share that would be publicly accessed? Did that make sense? [18:24] you san use existing samba share, but i'm not sure single sign on will work, so they may need to authenticate to samba share [18:24] or you can use nfs on windows, but generally that's not a good idea [18:25] implementation really sux :) [18:25] why? [18:25] :) [18:25] windows implementation of course [18:25] right [18:25] i've tryed it and had performance issues [18:26] Maybe I'm working at it backwards...I've already created a share for them on the windows box [18:26] I don't know a lot about samba [18:26] or linux for that matter [18:26] We've got a linux firewall with dansguardian, but I only know enough to be dangerous [18:27] I'll keep digging. Thanks for the pointers [18:27] no prob :) [18:32] does anybody know if there is a good irc-services package for ngircd? I'm not having luck finding one [18:36] New bug: #379488 in mysql-dfsg-5.0 (main) "package mysql-server-5.0 None [modified: /var/lib/dpkg/info/mysql-server-5.0.list] failed to install/upgrade: el subproceso pre-installation script devolvi? el c?digo de salida de error 1" [Undecided,New] https://launchpad.net/bugs/379488 [18:42] Hi all, question: I have just set up apache/mysql/php to enable a simple webserver. /var/www/ is the web root. So if I put index.html in there it gets displayed to the web. But sometimes you want things higher up than that, i.e. on other hosted sites, the web root is not accessible by the public and youu have a public_html folder where the public acessible info goes...how do I make a folder within /var/www/ be the default 'public_html'? [18:43] is it something like editing /etc/apache/sites-available/default and adjusting the root directory? [18:45] Erm [18:45] That's a slightly weird thing to do [18:46] Put the things which are not part of the website somewhere other than /var/www [18:47] j0nr: If your looking for per-user public_html folders (which is the only place I know of that dir name being used) you need to look into mod_userdir for apache [18:48] ok, but somethings are part of the website... maybe I am just confused (n00b)... I am trying to install gallery software... the frontend of it is in a sub-directory of the webroot as it is viewable by the public, but the data for it, i.e. the raw images, are stored outside the document root...think I am answering my own question a bit... [18:48] i can just tell it where they are in relation to '/' [18:48] this is just because its my own server rather than a hosted one [18:48] ok no worries... :) [18:50] for gallery, the standard place to store the data (i.e. photos) files is /var/lib/gallery2/g2data/ and that is something that you set in the config.php === stickystyle1 is now known as stickystyle [18:51] FWIW, gallery is already in the repos so you don't have to manage all this yourself. [18:51] under 'gallery2' [18:51] howdy all [18:52] I got the two nic thing to work [18:52] thanks for the help [18:53] how do I tell ubuntu server to do certain requests on one nic and all the others on another? [18:53] can someone just point me in right direction? Not sure what to search for as far as effective keywords...keep getting #$@#Q!!#. Thanks [18:55] bin1010: what do you exactly want to do? [18:57] I am using one nic mainly for incoming requests on local network and the other nic for internet. Sometimes when the server makes a request like for aptitude update, it has to wait for some timeout before it gets the information it seeks. I want to go ahead and tell server for most requests uses nic 2 and only use nic 1 for these specific things. [18:57] did that make sense? [18:59] bin1010: which nic is your default gateway? [18:59] good question...hold on [18:59] and it sounds more like a DNS resolution issue. [19:00] well one nic can see a local network just fine....but it can't get to the internet at all. The other is hooked directly to our internet infrastructure [19:00] or probably both interfaces connected to same switch and same broadcast domain [19:03] bin1010: does this computer serve as a gateway/firewall for computers on the the LAN side? [19:04] ip addresses are 192.168.56.101...thats the local network guy....and the other is 10.0.2.15 which has the ability to get to internet. [19:04] no...no gateway/firewall [19:05] i am using it as local web development environment...so I only allow connections to it via the local network NIC [19:05] connections to the machine work greate [19:06] but apt-get updates take a while to get going? [19:08] yes. it hangs for a time, then goes...kindof like a timeout [19:09] remember when firefox had a similar problem...it would hang on a socket drop [19:09] The DNS server(s) set in /etc/resolv.conf, are they on the internet or local? [19:10] the one it is pointed to is the one that can get to the internet....I just use the boxes ip address to check the webpage...sorry [19:10] BTW, there is no 'timeout' type thing with having more than one NIC. You have a set of routes that the computer tries, and then the default gateway. [19:10] okay....cool That's much better ;) [19:11] To expand upon that, if your DNS server is not on those predefined set of routes (which it sounds like it is not) then the computer will always use the default gateway interface. [19:12] that makes sense [19:12] I'm leaning towards a DNS issue if you haven't noticed ;) [19:12] is see that [19:12] LOL [19:13] yes...the local network NIC will not be able to get to the DNS server in resolv.conf [19:16] so that is the problem...but there is no way for me to get that NIC to that DNS server or even the gateway...they are really setup for the internet nic [19:16] sorry the DNS and Gateway are setup for the "internet"nic... [19:17] that sounds like a valid setup from what you have described so far. [19:18] but wont both nics try to use those settings? [19:20] sorry for my ignorance...I am more of a developer than a linux admin. I can get around most things, but I still get lost/confused on occasion. :) [19:20] Yes, the OS will handle which NIC to use for a particular request. [19:20] No worries. [19:20] sweet...thanks [19:22] how many DNS servers do you have in resov.conf ? [19:22] just one [19:24] Just for fun... change it to '4.4.4.2' (that is verizion's DNS server), we won't leave it on that but just for testing lets see what happens. [19:25] Whoops. I meant to type 4.2.2.2 [19:26] He'll be back :) [19:41] guys, i just create a .iso with remastersys, when i boot it the boot process stop in "configuring network interfaces", any idea ? [19:42] "ctrl+c" does not work.. the boot just keeps stuck [20:28] i have just picked up likewise open for my workstation. I have a primo setup I dont want to lose. I figured our winbind usermap, so when I log in as AD username, it appropriately comes in as the local user, but I think that is a bandaid. [20:29] sommer: here? :) [20:29] is it safe to cp -R ~ to the ad users homedir, chown it all, and expect gnome settings and seuch to be safe? Think gconf will have a conniption? [20:32] ivoks, heya master... any news on the HA track for the UDS? [20:33] RoAkSoAx: i haven't checked [20:33] you already in barcelona? [20:33] yes [20:33] awesome [20:33] I wish I can attend to an UDS someday :( [20:33] hahaha [20:34] work hard and you might :D [20:35] indeed [20:35] I think they have not scheduled the HA talk just yet [21:03] radovan: thx! I will give it a try [21:05] Question: I tried shutdown -F, touch /forcefsck on separate occasions. Neither triggered a fsck on the reboot, any ideas how to do so? [21:17] hello [21:18] hello [21:18] exit [21:19] hi, I have installed through aptitude mysql-server-5.1.13 [21:20] but I cannot find php5-mysql for the version of 5.1.13 [21:20] as a result [21:20] to have a php mysql client 5.0.x and not in the same version of the db server [21:21] does anyone know how to upgrade php mysql client to 5.1.13? [23:21] Hi. I have my own private repository and I'm trying to permanently remove a package. I'm running "reprepro -b . remove jaunty packageName" and it looks like it succeeds. But when I do an apt-get update and an apt-cache search on one of my servers I can still see the package. Is there something I'm missing? [23:23] Anyone? === yml_ is now known as yml