| methods | is there any good resources on installing newer mysql server ? | 00:12 |
|---|---|---|
| EvilResistance | methods, define "installing newer mysql server" | 00:17 |
| methods | >=5.5.3 | 00:17 |
| EvilResistance | which version of mysqld (mysql daemon aka mysql server) do yo uwant to install? | 00:17 |
| EvilResistance | methods, unless its in sid or already packaged, you might have to compile/install from source | 00:20 |
| EvilResistance | lemme check debian sid | 00:20 |
| methods | sid ? | 00:20 |
| EvilResistance | Debian Sid | 00:20 |
| methods | well .. mysql download page offers debian downloads | 00:20 |
| methods | but the package names aren't the same and stuff so I know that might cause issues where apt-get'ing something may require mysql-server to be installed which clobbers my setup | 00:21 |
| methods | http://www.mysql.com/downloads/mysql/#downloads | 00:22 |
| EvilResistance | i know the page | 00:22 |
| EvilResistance | i'm reading atm | 00:22 |
| EvilResistance | methods: if you can wait to precise, there's a mysql-server-5.5 you can install | 00:24 |
| methods | precise ?? | 00:24 |
| EvilResistance | but it doesnt seem to be fully tested | 00:24 |
| EvilResistance | holy god, you arent aware of release names are you? | 00:24 |
| EvilResistance | !precise | 00:24 |
| ubottu | Ubuntu 12.04 LTS (Precise Pangolin) will be the sixteenth release of Ubuntu. Codename announcement here: http://www.markshuttleworth.com/archives/784 - Blueprints at https://blueprints.launchpad.net/ubuntu/precise - Discussion and support in #ubuntu+1 | 00:24 |
| EvilResistance | you should learn version numbers vs. codenames | 00:25 |
| methods | code names are kind of annoying rather just have versions | 00:25 |
| EvilResistance | Lucid is 10.04, Maverick = 10.10, Natty = 11.04, Oneiric = 12.04 | 00:25 |
| methods | but honestly i just upgrade my desktop whenever i don't care really to remember | 00:25 |
| EvilResistance | well in Debian, codenames need to be known anyways | 00:25 |
| EvilResistance | because Debian 6 is denoted as 'squeeze' in the sources.list entries :P | 00:25 |
| methods | well I'm on amazon anyway and they are behind | 00:26 |
| EvilResistance | sid = unstable | 00:26 |
| EvilResistance | etc. | 00:26 |
| EvilResistance | oh you didnt mention this was on an amazon "cloud vps" | 00:26 |
| methods | trying to find out what their latest is | 00:29 |
| methods | i wonder if i could get away with the precise package | 00:30 |
| methods | it might work fine on older ubuntu | 00:30 |
| hallyn | stgraber: well, i *think* that's pretty much what we want :) (wrt a typo) but, do you agree static-network-up is the way to go? | 00:31 |
| Lazerath | hay all | 00:35 |
| Lazerath | ok | 00:46 |
| Lazerath | my grub wont load | 00:46 |
| hallyn | i think i'm going to do the emit in pre-start after all | 00:46 |
| hallyn | should be ok if i do --no-wait | 00:47 |
| Lazerath | so it used to load but i moved the server and it loads to a grub prompt | 00:48 |
| Lazerath | I am pretty sure i am using a LVM | 00:49 |
| Lazerath | so when I am at the grub prompt and I do the ls command | 00:50 |
| Lazerath | I see the hd0 hd0,1 hd1 hd1,1 but none of them have my grub on them | 00:51 |
| Lazerath | it is on the drive "landl" | 00:52 |
| Lazerath | but I am not sure how to explain it | 00:52 |
| Lazerath | when i load from the live cd and recover broken system | 00:53 |
| Lazerath | I see the landl partition | 00:53 |
| Lazerath | and i shell to there but grub-install sg1 | 00:54 |
| Lazerath | does not work | 00:54 |
| Lazerath | the only drive is a scsi raid set up | 00:54 |
| twb | Lazerath: pastebin contents of /proc/partitions. | 00:55 |
| twb | Lazerath: run file -s across each of the partitions listed in /proc/partitions, it will report if a bootloader is installed on them. | 00:55 |
| Lazerath | do that from the recovery shell off the live cd | 00:56 |
| twb | Lazerath: another test is dd if=/dev/sda bs=512 count=1 | strings | 00:56 |
| Lazerath | it used to work | 00:56 |
| twb | That test will print a bunch of strings, including "GRUB" if grub is the bootloader on sda | 00:56 |
| Lazerath | i think it is sg1 not sda | 00:56 |
| twb | I have never seen disks called "landl" or "sg1" before; the latter is believably but the former sounds very unlikely. | 00:57 |
| twb | sg1 is more likely to be the CD or tape drive | 00:57 |
| Lazerath | ok landl is my own partion | 00:57 |
| twb | As I said, look at /proc/partitions. The number of partitions on a device and their sizes should make it very obvious which one(s) are you drive(s). | 00:58 |
| Lazerath | the only drives hooked up to the server are a cd rom and a SCSI External 7disk raid stack with 2 partitions | 00:58 |
| Lazerath | but i will check that out quick | 00:58 |
| Lazerath | should i check the /proc/partitions from the grub shell on my failed boot or from the recovery mode on the live cd | 00:59 |
| twb | Ah, OK, stupid hardware raid stuff often shows up in bizarre ways | 00:59 |
| twb | Lazerath: you need to check it from a live medium | 01:00 |
| Lazerath | I think the landl partition is the LVM | 01:00 |
| mtaylor | smoser: around? | 01:00 |
| twb | The grub shell is pretty useless | 01:00 |
| Lazerath | yeah it is | 01:00 |
| mtaylor | smoser: why does python2.7 conflict with python-profiler/ | 01:00 |
| mtaylor | ? | 01:00 |
| Lazerath | ok i will just boot ubuntu live cd and check it out | 01:00 |
| Lazerath | i was going to do that next anyway | 01:00 |
| Lazerath | i think i have grub and not grub2 not sure though but i will check this out... Thanks mate | 01:02 |
| twb | It doesn't matter much | 01:08 |
| Lazerath | WELL live cd is still loading | 01:43 |
| Lazerath | but i will check the /proc/partitions directory | 01:43 |
| Lazerath | what was the command with the -s variable you said | 01:43 |
| twb | 11:55 <twb> Lazerath: run file -s across each of the partitions listed in /proc/partitions, it will report if a bootloader is installed on them. | 01:45 |
| hex20dec | Can someone please help me pass this part??? | 01:46 |
| hex20dec | https://help.ubuntu.com/10.04/serverguide/C/dns-configuration.html#dns-caching-configuration | 01:46 |
| hex20dec | What IPs are they talking about?? | 01:46 |
| hex20dec | Anyone?? | 01:48 |
| kerframil | hex20dec: the DNS servers operated by your Internet Service Provider. that information is also false; you don't *have* to specify forwarders, as long as bind is aware of the 'root' name servers. | 01:49 |
| twb | hex20dec: "the IP Addresses of your ISP's DNS servers" | 01:49 |
| Vexiant | yo dawg | 01:49 |
| kerframil | hex20dec: I'll answer here. most folk want their DNS servers to answer queries for records outside of any 'local' zones that might be maintained. you may tell yours to forward such requests to a specific set of nameservers. rationale for using nameservers operated by your ISP is that they might be faster. | 01:54 |
| kerframil | hex20dec: if you don't specify forwarders, than the root hint servers will be used (see /etc/bind/db.root) | 01:55 |
| kerframil | hex20dec: if you do not perform recursion at all, then your own clients will not work very well. dns servers have to be able to ask other dns servers for information. | 01:56 |
| Vexiant | persay | 01:57 |
| Vexiant | forwarders { | 01:57 |
| Vexiant | 1.2.3.4; | 01:57 |
| Vexiant | 5.6.7.8; | 01:57 |
| Vexiant | }; | 01:57 |
| kerframil | well, they don't, but ... that's a common configuration | 01:57 |
| Vexiant | the IP's go in the two number sets | 01:57 |
| Vexiant | what IPs would be put there? | 01:57 |
| Vexiant | https://help.ubuntu.com/10.04/serverguide/C/dns-configuration.html#dns-caching-configuration | 01:57 |
| Vexiant | I'm talking to you, kerframil | 01:59 |
| Vexiant | hex and I are both trying to set the stuff up correctly | 01:59 |
| kerframil | Vexiant: if you don't know, then you don't have to specify any forwarders at all - as noted before. it's optional, even though the doc there suggests it isn't. | 02:00 |
| kerframil | Vexiant: a default bind setup will be aware of the 'root' nameservers, which can be used for upstream queries. | 02:00 |
| Vexiant | thanks | 02:03 |
| kerframil | Vexiant: if you do want to use your ISP nameserver, ask them. or look for the info in their knowledge base or such. | 02:03 |
| Vexiant | ok | 02:03 |
| kerframil | Vexiant: be careful though. some organisations operate really crap nameservers. you might actually be better off using the root servers, especially if you're caching anyway. | 02:04 |
| kerframil | Vexiant: Google's work alright too (8.8.8.8, 8.8.4.4) | 02:05 |
| Vexiant | thanks | 02:06 |
| stgraber | hallyn: static-network-up sounds good yes | 02:16 |
| twb | kerframil: IMO it is advisable to tell stupid noobs to use their ISP's DNS server, so as to reduce the load on the root servers | 02:31 |
| twb | If that *does* cause them problems, it can be dealt with then | 02:32 |
| kerframil | twb: a reasonable assertion | 02:32 |
| twb | My favourite annoying one is telstra seems to resolve all unqualified hosts to help.telstra.com or so, so when I'm at my dad's place, I can't just "getent hosts <an office host>" to test if I'm in the office | 02:33 |
| kerframil | twb: gah | 02:34 |
| hallyn | stgraber: i'm feeling burned out, think i'll wait on that. maybe next upload. Or maybe just on next merge. | 02:36 |
| stgraber | hallyn: would be nice to have for 12.04, but we can probably argue it's not technically a feature so can maybe wait post-Feature Freeze | 02:37 |
| hallyn | oh right. that. | 02:37 |
| hallyn | stgraber: ok, so if we switch to upstart, do you think 'stop lxc' should behave like 'stop libvirt' - and not shut down its network? | 02:38 |
| hallyn | (at all) | 02:38 |
| hallyn | right now, /etc/init.d/lxc stop will shut down the lxc nwetork if no devices are attached to the lxc bridge | 02:38 |
| hallyn | eh, ok, i'll see what i can do | 02:39 |
| hallyn | i'm trying to see if poor jodh has run into any other lxc-caused bugs | 02:39 |
| hallyn | and i wonder how your upstart lxc changes are faring | 02:40 |
| uvirtbot | New bug: #927519 in makedev (main) "makedev fails to upgrade in lxc container with error 'Device or resource busy'" [Medium,Confirmed] https://launchpad.net/bugs/927519 | 02:41 |
| stgraber | hallyn: I'll do some poking for the upstart changes, I know jodh has been pretty busy with upstart bugfixes and plymouth stuff ... | 02:41 |
| hallyn | and lxc-caused bugs | 02:42 |
| stgraber | hallyn: for the network, I'm not sure, same for the containers, I think killing the network if we no longer have anything in the bridge is fine, otherwise it's probably a bad idea | 02:42 |
| stgraber | I'm also not sure if "stop" should kill all the containers marked as "auto", I guess people would expect us to do so, but it won't react like libvirt then | 02:42 |
| stgraber | and we'll need to be careful not to "restart" lxc on package upgrade (or we'll suddently kill and start all the containers) | 02:43 |
| stgraber | hallyn: just saw your comment on the console/tty bug, I guess that'd work indeed, we could use a sub-directory too if that makes things easier | 02:55 |
| stgraber | hallyn: though I'm wondering how hardcoded /dev/console is, I wouldn't be surprised if software were using it directly (upstart being one of them) | 02:55 |
| hallyn | stgraber: maybe we can affect that with upstart arguments (akin to '--console /dev/ttyS0') | 02:58 |
| hallyn | still, i'm afraid you're probably right, it would end up missing something | 02:58 |
| stgraber | hallyn: indeed. Something else that we'd need to think about is how to deal with older version of the OS in the container | 02:58 |
| stgraber | though, I agree this would solve most of our problems (if not all) around upgrades/dist-upgrades in containers, so it's definitely worth investigating | 02:59 |
| hallyn | i was going to mention to you tomorrow i think we need a serious devns design discussion at uds | 02:59 |
| hallyn | anyway | 03:00 |
| stgraber | hallyn: yep, that seems to be our biggest pain nowadays (with these devices + udev storms), so indeed worth booking some time to sit and think about it | 03:00 |
| designbybeck | hello all, I am a server newbie i have a Mahara site setup, that someone walked me through, I am working on a Moodle site as well. My Mahara site points correctly to the subdomian, but the Moodle subdomain just pulls up the root domain | 03:01 |
| designbybeck | this works: http://mahara.triside.com ....but this doesn't http://moodle.triside.com | 03:02 |
| designbybeck | I made the VirtualHost file the same as each other with their directories pointing to their respective locations | 03:02 |
| designbybeck | I have restarted apache2 and it showed OK I think I forgot something, i just don't know what I missed | 03:03 |
| qman___ | servername, most likely | 03:03 |
| designbybeck | inside the vhost file qman___ ? | 03:04 |
| qman___ | to work properly, the virutalhost must be * (*:80 / *:443 ok too) | 03:04 |
| qman___ | and the servername directive defined for each host | 03:04 |
| qman___ | one bad one can mess up the whole config | 03:04 |
| Vexiant | Yo, how do I find out the IP of a server via terminal? | 03:05 |
| designbybeck | ok, I'll keep looking, thanks qman___ | 03:06 |
| Vexiant | anyone know? | 03:06 |
| designbybeck | qman___, here is what both vhosts look like, but they aren't in the same file http://www.pasteall.org/28928 | 03:07 |
| kerframil | Vexiant: ip addr show | 03:07 |
| designbybeck | Vexiant, or you can try ifconfig i believe as well | 03:07 |
| Vexiant | ifconfig is for your own ip | 03:07 |
| Vexiant | thanks kerframil | 03:07 |
| qman___ | designbybeck, those are good, but make sure no other virtualhosts are defined that defy that convention | 03:08 |
| designbybeck | hmm none that i know of | 03:08 |
| qman___ | the way the apache config works is all those config files cross-reference each other into one big config | 03:08 |
| qman___ | so it could be anywhere | 03:08 |
| designbybeck | oh | 03:08 |
| qman___ | use grep -R to find any extraneous hosts | 03:08 |
| qman___ | you should also have "NameVirtualHost *" | 03:10 |
| designbybeck | http://www.pasteall.org/28929 | 03:10 |
| designbybeck | those are the vhosts files | 03:10 |
| designbybeck | hmmm... I haven't ran acros the NameVirtualHost | 03:11 |
| qman___ | do like so | 03:11 |
| qman___ | sudo grep -R NameVirtualHost /etc/apache2 | 03:11 |
| designbybeck | and here is the sample moodle file from their site: http://www.pasteall.org/28930 | 03:12 |
| qman___ | sudo grep -Ri ^\<virtualhost /etc/apache2 | 03:13 |
| qman___ | this particular issue doesn't have anything to do with the application being run, it's just a general apache virtualhost problem | 03:14 |
| designbybeck | qman___, here is that output | 03:14 |
| designbybeck | http://www.pasteall.org/28931 | 03:14 |
| qman___ | well that answers it | 03:15 |
| designbybeck | nto sure what the unable to resolve host triside is all about | 03:15 |
| qman___ | your hosts are not enabled, only the default is | 03:15 |
| designbybeck | but triside.com works and mahara.triside work? | 03:15 |
| qman___ | the default site is enabled | 03:15 |
| qman___ | the files are probably available through it | 03:15 |
| designbybeck | hmmm | 03:15 |
| qman___ | also, that message means 'triside' doesn't resolve in /etc/hosts | 03:16 |
| qman___ | you should add it such that it points to the LAN adapter's IP | 03:16 |
| designbybeck | i actually haven't edited the default file | 03:16 |
| qman___ | see a2ensite and a2dissite | 03:16 |
| qman___ | and your virtualhosts should be each in their own file in /etc/apache2/sites-available | 03:16 |
| designbybeck | i do have them in their own files | 03:17 |
| designbybeck | what do you mean by see a2ensite and a2disiste? what does that mean | 03:17 |
| qman___ | commands | 03:17 |
| designbybeck | i thought that mean /etc/apache2/sites-available | 03:17 |
| designbybeck | oohh | 03:17 |
| Lazerath_Phone | Ok | 03:18 |
| Lazerath_Phone | The boot drive is sda1 | 03:18 |
| Lazerath_Phone | But the files are on the lvm sda5 | 03:19 |
| qman___ | designbybeck, according to that output, your site configurations are either not formatted correctly or are not in the right place, because they'd show up in that list if they were | 03:20 |
| qman___ | like default-ssl does | 03:20 |
| designbybeck | ahhh qman___ i did see that on the mahara install instruction site it did have this command: sudo a2ensite mahara.conf | 03:20 |
| designbybeck | but i haven't done anything like that for the moodlevirtualhost file | 03:20 |
| Lazerath_Phone | the sda1 does have some files for booting | 03:21 |
| Lazerath_Phone | But the lvm is locked so how do I reinstall grub | 03:22 |
| qman___ | designbybeck, ah, they didn't show because of the whitespace | 03:22 |
| qman___ | not sure if that matters or not | 03:22 |
| qman___ | but anyway, you need to enable both of those sites, and should probably disable the default site if you're not using it | 03:22 |
| designbybeck | ok | 03:22 |
| designbybeck | let me try that command | 03:23 |
| designbybeck | LOL.... well....it doesn't point to the root anymore | 03:24 |
| designbybeck | qman___, i did the a2ensite and that seemed to work ok | 03:25 |
| designbybeck | then i restarted apace2 and that seemed to work ok | 03:25 |
| designbybeck | letme see how to disable default | 03:25 |
| qman___ | should be sudo a2dissite default | 03:25 |
| qman___ | and you can verify by doing ls /etc/apache2/sites-enabled | 03:26 |
| qman___ | those two commands are glorified symbolic linkers | 03:26 |
| qman___ | but it works pretty well | 03:26 |
| designbybeck | ok i dod see the moodlevirtualhost in sites-enable | 03:27 |
| designbybeck | and default too | 03:27 |
| qman___ | that should work as expected then | 03:28 |
| qman___ | provided the servername directive is set correctly for each site, when you visit by those URLs, you get the right site | 03:28 |
| qman___ | you can also use the serveralias directive to add more names | 03:28 |
| designbybeck | a2dissite is the other command? | 03:29 |
| qman___ | yes | 03:29 |
| designbybeck | so it would be sudo a2dissite default ? | 03:30 |
| qman___ | yes | 03:30 |
| qman___ | also, those servernames should resolve to your server, i.e. be in /etc/hosts | 03:31 |
| qman___ | but that's not strictly necessary, and in fact I just checked mine and it isn't | 03:32 |
| designbybeck | oh yeah i think i forgot to set those up | 03:34 |
| designbybeck | well....i think we're gettting smoewhere | 03:36 |
| designbybeck | ...but now triside.com is the same as mahara.triside.com | 03:36 |
| designbybeck | because i took off default? | 03:36 |
| designbybeck | let me check the /etc/hosts | 03:36 |
| qman___ | if that was your default site, yes | 03:36 |
| qman___ | each site must be defined, the default site takes over any that aren't | 03:37 |
| designbybeck | qman___, here is the /etc/hosts http://www.pasteall.org/28932 | 03:38 |
| qman___ | remove the second and third 127.0.0.1 lines | 03:38 |
| qman___ | actually, make it look like this: http://www.pasteall.org/28933 | 03:39 |
| designbybeck | do you have to restart apache afer that? | 03:39 |
| qman___ | yes | 03:39 |
| designbybeck | qman___, http://www.pasteall.org/28934 | 03:42 |
| designbybeck | i learned a few thigns this time | 03:42 |
| designbybeck | gave some errors | 03:42 |
| marrok5146 | anyone here familiar with openvas? | 03:42 |
| qman___ | ok, that means two things | 03:42 |
| qman___ | first, that that directory doesn't exist | 03:43 |
| qman___ | and second, that one of your servername directives doesn't resolve and needs to be added to /etc/hosts | 03:43 |
| designbybeck | ok...first thing, i thought i had the dir there | 03:43 |
| designbybeck | DOH! ...yeah i forgot i used the www for mahara, but for moodle (to keep them separate) i put it in its own dir | 03:44 |
| qman___ | that'll do it | 03:44 |
| qman___ | you may need to add your subdomains to /etc/hosts, if so, just tag them on the end | 03:45 |
| qman___ | like: 1.2.3.4 site site.com sub.site.com sub2.site.com | 03:45 |
| designbybeck | yeah i didn't get that error this time, but it still got the nameserver error | 03:47 |
| designbybeck | apache2: Could not reliably determine the server's fully qualified domain name, using 108.166.105.36 for ServerName | 03:47 |
| designbybeck | ok i'll try that in the /etc/hosts | 03:47 |
| designbybeck | Sweet!! moodle.triside.com resolved | 03:48 |
| qman___ | yeah, it's not super complicated, but unfortunately, unless you've done it before, it's not obvious where the problem is when it's not working as expected | 03:49 |
| designbybeck | well it was setting up apache and the file structure that obviously i'm still learning | 03:51 |
| designbybeck | i guess i thought things had to be in a very specific location | 03:52 |
| Lazerath | I the "cat partitions" command in the /proc folder | 03:52 |
| Lazerath | The boot drive is sda1 | 03:52 |
| Lazerath | But the files are on the lvm sda5 | 03:52 |
| Lazerath | the sda1 does have some files for booting | 03:52 |
| Lazerath | But the lvm is locked so how do I reinstall grub | 03:52 |
| designbybeck | but the important thing is that you just tell the file where to go | 03:52 |
| qman___ | yeah, it's a pretty flexible system, that's why it didn't throw any errors until you disabled the default site | 03:52 |
| designbybeck | ok so now i have to address the 2nd thing | 03:53 |
| designbybeck | i did put in the sub.site.com stuff in | 03:53 |
| designbybeck | is service apache2 reload the same as restart? | 03:53 |
| qman___ | they're not the same | 03:54 |
| qman___ | reload causes apache to reread all the configuration files live | 03:54 |
| qman___ | restart actually shuts down and restarts apache | 03:54 |
| qman___ | reload is faster and is in many cases good enough, but sometimes a full restart is needed | 03:54 |
| designbybeck | let me give it a good ol' restart then | 03:55 |
| designbybeck | * Restarting web server apache2 apache2: Could not reliably determine the server's fully qualified domain name, using 108.166.105.36 for ServerName | 03:56 |
| designbybeck | ... waiting apache2: Could not reliably determine the server's fully qualified domain name, using 108.166.105.36 for ServerName | 03:56 |
| Lazerath | yeahi found with apache i needed to restart sometimes when changing settings | 03:56 |
| designbybeck | ok that ServerName would be in the /site-available file correct? | 03:56 |
| qman___ | that servername could be anywhere | 03:56 |
| designbybeck | hmmm | 03:57 |
| qman___ | sudo grep -R ServerName /etc/apache2 to list them all | 03:57 |
| qman___ | and make sure all the ones listed (and which are not commented or disabled) resolve | 03:57 |
| qman___ | but bear in mind also that as long as your sites work the way you expect, that warning can be safely ignored | 03:58 |
| designbybeck | i'm not sure what this means qman___ , but is it missing smoething? http://www.pasteall.org/28936 | 03:58 |
| designbybeck | oh | 03:58 |
| designbybeck | yes my two resolve correctly | 03:59 |
| designbybeck | so for my testing purposes it is working thus far! | 03:59 |
| designbybeck | Thank you for your help qman___ !! I've learned a lot tonight ! | 04:00 |
| dforthman | Hi. I have Ubuntu 11.10 Server running, all updates applied. I'm trying to install Nagios NRPE plugin, but I'm getting "Unable to find SSL libraries" compilation error. I've installed libssl-dev, but I'm getting the same error on compilation. Is there another package I need for the ssl libraries? | 04:03 |
| qman___ | dforthman, suggest you install the nagios-nrpe-server package instead of rolling your own | 04:04 |
| qman___ | compiling your own software is not a best practice on production servers | 04:05 |
| dforthman | No, but when I installed the package from the repos, it was missing configuration files and wouldn't start. Uninstalled/re-installed several times and got the same result. | 04:06 |
| qman___ | I've installed said package without incident, did you modify them yourself? | 04:06 |
| qman___ | in any case, install then purge | 04:06 |
| qman___ | then remove the cached package file from /var/cache/apt/archives | 04:07 |
| qman___ | and download the fresh one | 04:07 |
| qman___ | purge gets rid of all configuration files so that the next install will recreate them | 04:07 |
| qman___ | remove does not | 04:07 |
| dforthman | Did all that. It was still missing the config files. | 04:07 |
| qman___ | what file(s) are you expecting which is missing? | 04:08 |
| qman___ | also, are you trying to set up your monitoring server, or the servers that are being monitored? | 04:10 |
| qman___ | because nagios-nrpe-server is what | 04:10 |
| qman___ | 's needed on the servers being monitored | 04:11 |
| dforthman | We're off-topic. Are there additional packages I need for the SSL libraries? Or where do I point --with-ssl-lib=/path/to/ssl/lib? | 04:11 |
| Vexiant | https://help.ubuntu.com/10.04/serverguide/C/dns-configuration.html | 04:11 |
| Vexiant | "Also, create an A record" | 04:11 |
| Vexiant | I keep seeing "A record" | 04:11 |
| Vexiant | What does that mean? | 04:11 |
| dforthman | "A record" is an address record. It points your domain name to an IP address. | 04:12 |
| Vexiant | @ IN A 127.0.0.1 | 04:12 |
| Vexiant | @ IN AAAA ::1 | 04:12 |
| Vexiant | ns IN A 192.168.1.10 | 04:12 |
| Vexiant | what would I put inplace of the A's? | 04:12 |
| qman___ | you don't | 04:12 |
| qman___ | the As belong there | 04:12 |
| Vexiant | /etc/bind/db.xxx.com:14: unknown RR type 'AAA' (Note, the X's are just there to hide the site) | 04:12 |
| Vexiant | welp, check BIND gave me that error | 04:13 |
| qman___ | a AAAA record is an ipv6 record | 04:13 |
| qman___ | your version of BIND is not interpreting ipv6 records | 04:13 |
| qman___ | remove the errant record or fix your BIND to accept ipv6 records | 04:13 |
| Vexiant | So what should I do? | 04:13 |
| Vexiant | errant? Sorry, I'm new to all of this | 04:13 |
| dforthman | Remove the line with AAAA | 04:14 |
| Vexiant | Ok | 04:14 |
| dforthman | then restart Bind | 04:14 |
| Vexiant | ok, thanks | 04:14 |
| qman___ | dforthman, you probably need libssl-dev, but again, there's nothing wrong with the packaged versions | 04:14 |
| qman___ | I use them, they work | 04:15 |
| Vexiant | Yo, thanks so much, guys. It just removed ALL my errors! | 04:15 |
| dforthman | I've installed libssl-dev but it's still giving the error. I'll try the packaged version again. | 04:15 |
| dforthman | And I'm trying to install the NRPE addon to the Nagios monitoring server. | 04:16 |
| qman___ | that one needs the -plugin package | 04:16 |
| qman___ | nagios-nrpe-plugin | 04:16 |
| twb | qman___: or he just typed AAA instead of AAAA | 04:17 |
| qman___ | true, I assumed his paste was what was in his records but it could not have been | 04:18 |
| qman___ | could have not been* | 04:18 |
| Vexiant | it was AAA | 04:19 |
| Vexiant | the problem is solved, as I stated. Thanks to qman___ and dforthman | 04:19 |
| twb | Vexiant: there are 4 A's in an AAAA | 04:19 |
| qman___ | but odds are you don't need any AAAA records, especially if you don't already know what they are | 04:20 |
| twb | Vexiant: if you get that wrong dumb things happen | 04:20 |
| dforthman | thanks to qman___, I just translated. | 04:20 |
| Vexiant | twb, I got nothing wrong. It was already set like that | 04:20 |
| twb | I have lost interest in this discussion | 04:21 |
| Vexiant | yes, the problem has been colved | 04:21 |
| Vexiant | solved* | 04:21 |
| dforthman | alright, let's see how this nagios on the repos goes this time | 04:23 |
| dforthman | "Not replacing deleted config file" | 04:25 |
| qman___ | purge and try again, that normally fixes it | 04:25 |
| qman___ | it might be a dep too | 04:26 |
| Lazerath | ok all | 04:26 |
| Lazerath | I ran the "cat partitions" command in the /proc folder...The boot drive is sda1...the sda1 does have some files for booting..like: initrd.img-2.6.32-32-generic-pae..But the files are on the lvm sda5..like: initrd.img..But the lvm is locked so how do I reinstall grub.... I can access the lvm when booting recovery from the cd.... i got some type of error when i ran "grub-install hd0" which i cannot remember right now. | 04:26 |
| dforthman | Total Warnings: 0 Total Errors: 7 | 04:28 |
| dforthman | it won't let me purge, either. i'm getting error messages | 04:29 |
| qman___ | pastebin the whole log on the purge | 04:29 |
| qman___ | it usually amounts to touching the missing files to satisfy apt | 04:30 |
| qman___ | so it can remove them | 04:30 |
| Lazerath | bah | 04:31 |
| Lazerath | +-++++++ | 04:31 |
| dforthman | http://pastebin.com/xw7mAj0e | 04:31 |
| qman___ | ah, looks like an in-nagios checking issue | 04:32 |
| qman___ | try commending out / removing all the checks on localhost's config | 04:33 |
| dforthman | ugh. no good | 04:37 |
| dforthman | now i'm getting "No checks defined" | 04:38 |
| dforthman | and it errors out | 04:38 |
| dforthman | "No services defined" rather | 04:39 |
| qman___ | ok, try purging nagios3-common nagios3-core nagios3 nagios3-cgi all at once | 04:39 |
| qman___ | nagios3 looks like a metapackage and that may be why it's wanting things to work before you can remove it | 04:40 |
| dforthman | seems to be working | 04:40 |
| dforthman | sh: getcwd() failed: No such file or directory sh: getcwd() failed: No such file or directory sh: getcwd() failed: No such file or directory | 04:40 |
| dforthman | should that worry me? | 04:40 |
| qman___ | not necessarily | 04:40 |
| qman___ | what's your current working directory | 04:41 |
| dforthman | var/cache/apt/archives | 04:41 |
| qman___ | a little odd then, but it could just be missing files/directories | 04:42 |
| qman___ | in any case, a successful purge marks it as totally gone | 04:42 |
| qman___ | and then you can manually delete any files left over | 04:42 |
| qman___ | and install fresh | 04:42 |
| qman___ | I've had bad downloads make for very bad days before | 04:43 |
| dforthman | ok, so which nagios package has nrpe? | 04:44 |
| qman___ | nagios-nrpe-plugin installs the monitoring server bits for nrpe | 04:45 |
| qman___ | you also need a working nagios (or icinga) for it to be useful, obviously | 04:45 |
| dforthman | yeah. i'm gonna set that up now | 04:46 |
| qman___ | nagios-nrpe-server are the bits for the machines being monitored | 04:46 |
| qman___ | that naming threw me for a loop for quite a bit | 04:46 |
| dforthman | eh, that's what the NSClient++ is for haha | 04:46 |
| dforthman | we don't run anything linux except for the nagios server | 04:46 |
| qman___ | I use nsclient++ too, works pretty well | 04:47 |
| qman___ | that is, when the untangle firewall isn't corrupting the download, which it was with 0.38 for some reason | 04:49 |
| dforthman | haha | 04:49 |
| twb | 10.04 ubuntu fortune file has a link to https://help.ubuntu.com/9.10/serverguide/C/networking.html | 04:53 |
| twb | That should be checked and fixed for precise | 04:53 |
| twb | fortunes-ubuntu-server 0.3, I mean | 04:54 |
| dforthman | ok, so where are the actual check command file(s) located on the repository nagios? | 04:56 |
| dforthman | nevermind, found 'em | 04:57 |
| twb | dforthman: /etc/nagios3/*.conf is sourced by the master file therein IIRC | 04:57 |
| dforthman | i was looking fore /etc/nagios-plugins/http.cfg | 04:59 |
| dforthman | had to add a command to check for non-80 http servers | 04:59 |
| clone_ | hi all, know somthing about closed connection with Ubuntu Server 11.10 | 05:24 |
| clone_ | I got some servers with that problem | 05:25 |
| clone_ | after few bytes of transfer the connection wiht the services are close | 05:25 |
| clone_ | like ssh or apache web server | 05:25 |
| clone_ | but only with some Ip | 05:26 |
| clone_ | not all of it | 05:26 |
| === KeyGruin is now known as KeyGruin-zzz | ||
| dforthman | so, now i'm getting "Could not parse" errors on my check_nt commands. Is the formatting different? or should check_nt!SERVICESTATE!-d SHOWALL -l "SQL Server (BKUPEXEC)" still work? | 05:37 |
| dforthman | needed to add -p in the nt.cfg file | 05:40 |
| clone_ | any Idea why apache2 - Connect to Apache times out randomly in ubuntu 11.10 | 06:02 |
| clone_ | the same thing with ssh | 06:02 |
| clone_ | just after call a command the ssh frozed | 06:02 |
| stiv2k | hi | 06:22 |
| stiv2k | anyone here alive | 06:22 |
| === clone_ is now known as clonemx | ||
| === clonemx is now known as cloneMX | ||
| === cloneMX is now known as clonemtz | ||
| clonemtz | hey | 06:31 |
| chelz | clonemtz: look for errors in the logs | 06:37 |
| chelz | and google those | 06:37 |
| stiv2k | i have a couple commands in my /etc/rc.local file that runs at startup | 06:39 |
| stiv2k | just two, like this | 06:39 |
| stiv2k | # Run IRCD | 06:39 |
| stiv2k | /home/ircd/Unreal3.2.7/ircdcron/ircdchk >/dev/null 2>&1 | 06:39 |
| stiv2k | # Run BOPM | 06:39 |
| stiv2k | /home/ircd/bopm/bin/bopm | 06:39 |
| stiv2k | but it runs them as root | 06:39 |
| stiv2k | what can i do to not run my ircd as root? | 06:39 |
| chelz | stiv2k: su <user> /home/ircd/Unreal3.2.7/ircdcron/ircdchk | 06:43 |
| chelz | su stiv2k /home/ircd/Unreal3.2.7/ircdcron/ircdchk >/dev/null 2>&1 | 06:44 |
| chelz | i dunno if rc.local has issues with arguments, i know cron stuff does. might have to put that line in a two line shell script | 06:44 |
| twb | chelz: apt-get install ircd-irc2; it runs as irc by default, not root | 06:45 |
| twb | Sorry, @stiv2k | 06:46 |
| twb | You probably should use start-stop-daemon(8) or upstart(8), though. | 06:46 |
| twb | For the latter ref. init(5) also. | 06:47 |
| === ts2 is now known as tsimpson | ||
| twb | kirkland: hey, did you leave Canonical? I notice your bzr commit emails changed :-) | 06:52 |
| stiv2k | twb: irc2? | 06:53 |
| stiv2k | whats that | 06:53 |
| stiv2k | chelz: i thought that was su -c ? | 06:53 |
| twb | stiv2k: people who need *an* IRC and don't care about the details | 06:55 |
| chelz | stiv2k: right, might be. i take no responsibility for my syntax :P | 06:55 |
| stiv2k | twb: i already have an irc server setup | 06:55 |
| twb | stiv2k: running out of /home? Real professional. | 06:56 |
| stiv2k | twb: it runs as its own user | 06:56 |
| chelz | yeah but to do stuff 'properly' using upstart and all that is the 'proper' way. if you start running a few custom services then you should look into it, but for the first few i figure rc.local is fine | 06:56 |
| stiv2k | chelz: what is upstart | 06:56 |
| twb | chelz: depends on whether you want to be lazy now or lazy in twelve months | 06:56 |
| twb | stiv2k: ubuntu's clusterfuck of a init replacement | 06:57 |
| stiv2k | twb: clusterfuck? so you dont like it? | 06:57 |
| twb | I do not like it. | 06:57 |
| chelz | twb, stiv2k it's what we must use. as it has been decided. so say we all | 06:58 |
| stiv2k | ok | 06:58 |
| chelz | stiv2k: yeah but it's basically init++ | 06:58 |
| chelz | gl with that :P | 06:58 |
| stiv2k | but then i have to write my own stop and start script for it right | 06:58 |
| twb | chelz: no, that's insserv. | 06:58 |
| chelz | eh | 06:58 |
| chelz | ok nvm then | 06:58 |
| twb | chelz: upstart is not an incremental improvement, but a complete paradigm shift | 06:58 |
| twb | stiv2k: yes, but simple upstart jobs are straightforward | 06:59 |
| chelz | stiv2k: right yeah it's kind of an ordeal. for a production env you should use it, for personal stuff i don't think it really matters. if you want to learn it go for it | 06:59 |
| twb | http://paste.debian.net/155377/ example | 07:00 |
| stiv2k | i will look into it | 07:00 |
| stiv2k | twb: whats wrong with running my ircd out of /home | 07:00 |
| twb | stiv2k: it's not integrated properly with the rest of the system | 07:00 |
| stiv2k | how so | 07:00 |
| twb | stiv2k: so, for example, apt will happily remove the C libraries it needs to run when nothing else keeps them installed | 07:01 |
| chelz | stiv2k: wait, for one thing, why aren't you using an ircd from the repos? | 07:01 |
| twb | chelz: EXACTLY | 07:01 |
| * chelz hits stiv2k with a stick | 07:01 | |
| chelz | stop doing wrong things | 07:01 |
| stiv2k | because man | 07:01 |
| twb | chelz: but probably because irc servers upstreams have a hard-on for running stuff out of CVS HEAD | 07:01 |
| stiv2k | i compiled mine myself | 07:01 |
| stiv2k | with specific options | 07:01 |
| stiv2k | that i want | 07:02 |
| twb | Gentoo is <over there> | 07:02 |
| stiv2k | i dont want to compile the whole OS | 07:02 |
| twb | But we discourage that kind of thing | 07:02 |
| stiv2k | just my ircd | 07:02 |
| chelz | eh | 07:02 |
| chelz | stiv2k: look into deb-src | 07:02 |
| chelz | i think that's the name | 07:03 |
| chelz | stiv2k: you grab the src for a package in the repos, then you can compile your own tweaked version of it, but it's still mostly within the package management system | 07:03 |
| dforthman | where'd qman___ go? | 07:03 |
| dforthman | to bed? | 07:03 |
| twb | chelz: add deb-src entries to sources.list; apt-get source foo; cd foo-1.0; sensible-editor debian/rules; dch -i "custom build"; debuild; dpkg -i ../*deb | 07:04 |
| chelz | twb: yeah that's looks good | 07:04 |
| stiv2k | what O_o | 07:05 |
| twb | Example: http://paste.debian.net/155378/ and http://paste.debian.net/155379/ | 07:05 |
| chelz | just have to do that again when it updates but that's the price of a custom package, but that'll be hardly ever | 07:05 |
| twb | Although the latter can be substituted by a simple dpkg -i, or an elaborate reprorepo or so | 07:05 |
| chelz | twb: there a good guide anywhere for ubuntu/debian deb-src stuff? | 07:05 |
| chelz | i think i saw on one reddit some time ago. didn't end up reading it but the idea of a guide like that is a good one | 07:06 |
| twb | Those pastes run out of cron and are responsible for recompiling openldap against openssl whenever ubuntu issues a new security update, you see. It lands in the in-house apt repo where it is signed, and unattended-upgrades on the various hosts are configured to upgrade to the new in-house build automatically. | 07:06 |
| chelz | that's pretty neat | 07:07 |
| twb | chelz: there isn't really a one-stop document, but the #debian-mentors group on OFTC can direct you to LOTS of documentation that helps, even if a lot of it is various levels of stale | 07:07 |
| twb | e.g. dh_make tends to encourage newbies to use pre-dh7 style rules files | 07:07 |
| chelz | oh hm, so this gets into some packaging stuff. i've yet to master those dark arts. | 07:08 |
| chelz | on my todo list though since quite a few projects are in need of some packagers | 07:08 |
| twb | A lot of it is easy, you just need to say "hey I want to do <blah>" and have someone point you to the appropriate bit of helper code | 07:08 |
| stiv2k | hmm | 07:08 |
| stiv2k | i wonder if i should keep the ircd files on the SSD or the HDD | 07:09 |
| twb | stiv2k: both | 07:09 |
| stiv2k | the / is on the SSD | 07:09 |
| twb | RAID1 them with HDD configured as write-mostly write-behind | 07:09 |
| stiv2k | and /home is the HDD | 07:09 |
| chelz | oh hmm | 07:09 |
| chelz | https://help.ubuntu.com/community/UpdatingADeb | 07:09 |
| chelz | found that | 07:09 |
| stiv2k | twb: i dont think the machine supports raid | 07:09 |
| stiv2k | twb: http://stats.stiv2k.info | 07:09 |
| twb | stiv2k: md raid. Hardware raid is completely useless unless you are paying $200 at least, and it has a BBU or equivalent. | 07:10 |
| chelz | also the obligatory cyberciti.biz guide, dunno if what they say is good or not on this one: http://www.cyberciti.biz/faq/rebuilding-ubuntu-debian-linux-binary-package/ | 07:10 |
| stiv2k | i have good uptime thought :) | 07:10 |
| stiv2k | though* | 07:10 |
| chelz | more guides: https://www.google.com/search?q=rebuild+ubuntu+package+deb-src | 07:10 |
| chelz | stiv2k: rsnapshot to do backups of important stuff from one to the other | 07:11 |
| twb | I'm the rsnapshot upstream :-) | 07:11 |
| stiv2k | chelz: what i wanted was to keep OS files on the SSD, things that mostly read from but not written to | 07:11 |
| stiv2k | chelz: and user data, torrents, websites, etc, on the HDD | 07:11 |
| chelz | twb: whaa | 07:12 |
| twb | chelz: this is a reasonable example of how to package something: http://darcs.debian.org/collab-maint/mg | 07:12 |
| twb | Except ignore the get-orig-source and the override rules to deal with a non-autoconf ./configure script. | 07:12 |
| twb | But the best way to learn packaging is to lurk on #debian-mentors (on OFTC) | 07:13 |
| chelz | otfc | 07:14 |
| * chelz shakes fist | 07:14 | |
| chelz | also oftc | 07:14 |
| twb | HTFU you big baby | 07:15 |
| chelz | yeah yeah | 07:15 |
| chelz | well i dnno how good they are but googling for packaging docs sure brings up a lot | 07:15 |
| twb | Of uncertain vintage and reliability, sure | 07:16 |
| twb | It staggers me how many jackasses come in here "following a blog they found" and they're trying to apply 8.04 or 6.06 docs to a 12.04 install | 07:16 |
| chelz | haha | 07:17 |
| chelz | yeah, that doesn't work out | 07:17 |
| chelz | can't say i know the right or wrong ways to package things but that mg looks pretty good | 07:18 |
| chelz | also looks like a neat project | 07:18 |
| === smb` is now known as smb | ||
| === dforthman_ is now known as dforthman | ||
| Vivek | kirkland: Are you around ? | 08:37 |
| Vivek | roaksoax: Are you around ? | 08:41 |
| Vivek | Daviey: Are you there ? | 08:41 |
| Vivek | kirkland suggested that I have a word with both of you regarding some queries I had with dnsmasq on Orchestra. | 08:42 |
| Vivek | Since I am in the IST (Indian Standard Time). I am not able to get to talk to the developers while they are available. | 08:42 |
| === jodh is now known as jhunt` | ||
| twb | Try asking your real question first | 08:47 |
| erichammond | Vivek: You could try the ubuntu-server mailing list: https://lists.ubuntu.com/mailman/listinfo/ubuntu-server | 08:47 |
| Vivek | erichammond: Thanks. | 08:47 |
| Vivek | twb: The question was asked on Friday itself :) | 08:52 |
| Vivek | twb: Let me repaste the chat log with kirkland for you. | 08:53 |
| Vivek | twb: Pastebin it rather. | 08:54 |
| Vivek | twb: http://paste.debian.net/155392/ | 08:54 |
| * twb reads | 08:57 | |
| Vivek | jono: Hi | 08:57 |
| jono | hey Vivek | 08:58 |
| Vivek | jono: Vivek Cherian here. | 08:58 |
| twb | Vivek: please also pastebin dnsmasq.conf | 08:58 |
| Vivek | It was nice meeting you here mate. | 08:58 |
| Vivek | twb: sure give me couple of mins. | 08:58 |
| twb | No worries | 08:59 |
| twb | Vivek: so basically the problem is dnsmasq appears to be serving requests to both your ifaces? | 09:01 |
| jono | hey Vivek :-) | 09:01 |
| Vivek | twb: The problem is that inspite of having configured inteface=eth1, dnsmaq seems to be (I figured out that it might not be) serving I.P Addresses via eth0 also. | 09:04 |
| Vivek | Now the reason why we suspect that dhcp requests are being made via eth0 is because of the presence of directories named after I.P Addresses in /var/log/orchestra/rsyslog/2012/01/25 | 09:05 |
| twb | I need the pastebin | 09:06 |
| Vivek | Sure | 09:06 |
| jamespage | morning all | 09:07 |
| Daviey | Vivek: hello | 09:08 |
| Vivek | Daviey: Hi | 09:09 |
| Daviey | Vivek: Hi! | 09:13 |
| SpamapS | crap | 09:13 |
| SpamapS | Daviey and jamespage are up | 09:13 |
| SpamapS | I should probably have been in bed an hour ago then | 09:13 |
| jamespage | SpamapS, almost certainly | 09:13 |
| jamespage | SpamapS, good morning! | 09:14 |
| SpamapS | jamespage: indeed, good morning. :) | 09:14 |
| Vivek | Daviey: I have some queries which I am working out with twb. | 09:14 |
| Vivek | SpamapS: Hi again and prolly bye :) | 09:15 |
| Daviey | SpamapS: sleep well :) | 09:15 |
| Daviey | SpamapS: if it helps, i've been up 2 hours :) | 09:15 |
| Daviey | Vivek: sounds good, shoot away! | 09:15 |
| Vivek | Daviey: For some background http://paste.debian.net/155392/ | 09:16 |
| Vivek | twb: http://paste.debian.net/155397/ | 09:17 |
| === jhunt` is now known as jhunt_ | ||
| Vivek | twb: http://paste.debian.net/155398/ | 09:19 |
| * SpamapS must finish this epic email about mysql :-P | 09:19 | |
| twb | Vivek: ok, check /proc/<pid of dnsmasq>/cmdline to see if it's getting any overriding args | 09:19 |
| twb | tr '\0' '\n' </proc/N/cmdline if you have trouble with NULLs | 09:20 |
| Daviey | Vivek: firstly, can i check - you are not doing this (or targetting Precise?) | 09:20 |
| twb | Also I would try using exclude-interface=eth0 instead of interface=eth1 (check syntax in dnsmasq manpage) | 09:20 |
| uvirtbot | New bug: #928147 in mysql-dfsg-5.1 (main) "package mysql-server-5.1 5.1.41-3ubuntu12.10 failed to install/upgrade: EOF on stdin at conffile prompt" [Undecided,New] https://launchpad.net/bugs/928147 | 09:20 |
| twb | Also pastebin output of "ip a" and "ip r" please | 09:21 |
| Vivek | Daviey: It is Oneiric 32 bit i386 | 09:21 |
| Vivek | twb: ok | 09:21 |
| twb | btw here is a working dnsmasq (non orchestra) PXE server: http://paste.debian.net/155401/ | 09:22 |
| twb | I notice you're not doing TFTP inside dnsmasq | 09:23 |
| Vivek | twb: paste.debian.net/155402/ | 09:23 |
| Vivek | twb: http://paste.debian.net/155403/ | 09:24 |
| twb | Yeah OK so you also need to check /etc/dnsmasq.d/* if there is anything in there | 09:25 |
| Vivek | ok | 09:25 |
| Vivek | It only has a README :) | 09:25 |
| twb | How confident are you that it's dnsmasq that's responding on eth0, and not a rogue DHCP server? | 09:26 |
| twb | Because I can't see how it's happening | 09:26 |
| Vivek | Me too and I need to prove to the management here that it is not happening :) | 09:27 |
| twb | btw I would never use 192.168.[01].* as a range, due to conflicts with everyone else | 09:27 |
| Vivek | The Corporate range is the 10.x.x.x | 09:27 |
| twb | Vivek: on eth0, do a dhclient -n -v or so, see who is responding by MAC | 09:27 |
| dforthman | or 192.168.100.0 | 09:27 |
| twb | Using 192.168.<not 0 1 or 100>/24 or something in 172.16/12 | 09:27 |
| Vivek | Why are those directories getting created in /var/log/rsyslog/orchestra ? | 09:27 |
| twb | Vivek: that would be up to /etc/rsyslog.conf.d/*.conf | 09:28 |
| Vivek | The directories in the 10.x.x.x range ? | 09:28 |
| dforthman | i see 10.0.0.0 and 172.20.0.0 everywhere i go | 09:28 |
| twb | Vivek: but probably because someone in 10/8 is talking to your rsyslog. | 09:28 |
| Vivek | twb: exactly | 09:28 |
| twb | dforthman: I haven't ever seen anything in 172.16/12 AFAIK | 09:28 |
| Vivek | I have a theory | 09:28 |
| Vivek | I need you to tell me if that will sell :) | 09:29 |
| twb | cosmic rays? | 09:29 |
| Vivek | dnsmaq is broadcasting on eth0 saying hey I am a DHCP server out here... | 09:29 |
| Daviey | Vivek: the log is pushed to rsyslog, thanks to the kernel command line options sent to the installer. | 09:29 |
| Vivek | 10.x.x.x machines are asking for a I.P Address. | 09:29 |
| Vivek | dnsmaq is refusing it saying I can serve only on the 192.168.1.x range. | 09:29 |
| dforthman | electronic healthcare resellers love to use the 172.20.0.0 range dunno why | 09:30 |
| Vivek | The 10.x.x.x are created in rsyslog simply by the virtue of the 10.x.x.x contacting dnsmaq and on the virtue of leasing and I.P Address in the 10.x.x.x range. | 09:31 |
| * Vivek stops and waits for brickbats or an applause. | 09:31 | |
| twb | Vivek: that's not what happens, no | 09:32 |
| Vivek | twb: ok. | 09:32 |
| Vivek | twb: What is happening ? | 09:33 |
| twb | Vivek: dnsmasq says "I was told to serve 192.168.1.5 through .200 or whatever you said. That is routed via eth1, so I will respond on eth1" | 09:33 |
| Vivek | ok | 09:33 |
| twb | Under no circumstances should dnsmasq respond AT ALL (for DHCP, anyway) on eth0 | 09:33 |
| twb | Not with that config | 09:33 |
| Vivek | Then why is it responding, which means creating those 10.x.x.x directories in the rsyslog ? | 09:34 |
| twb | If your eth1 link is guaranteed to be up before dnsmasq (unlikely if you use upstart), then you could use --bind-interfaces instead of --interfaces. | 09:34 |
| twb | Vivek: I am not convinced it *is* responding | 09:34 |
| twb | Vivek: so far your only evidence is rsyslog hits, which is why you should be testing with dhclient -n -v eth0 or so | 09:35 |
| Vivek | Me too but we need to prove it, currently the presence of the 10.x.x.x cannot make us say for certain that no dchp activity is happening via eth0. | 09:35 |
| Vivek | ok | 09:35 |
| Vivek | Going to do that now. | 09:35 |
| twb | Another thing you could do, is simply have dnsmasq serve a bullshit next server programmed to do something silly, like eject the CD drive. | 09:37 |
| twb | Then wander around the office looking for ejected CDs | 09:37 |
| twb | But that's probably more elaborate than you can be bothered doing | 09:37 |
| twb | Also you may want --no-dhcp-interface instead of --except-interface | 09:38 |
| Vivek | ok | 09:38 |
| twb | Telling it to keep a leasefile and then keeping an eye on that, is also a handy check for stupidity | 09:39 |
| Vivek | ok. | 09:39 |
| twb | And of course --log-dhcp | 09:39 |
| twb | At this point I'm just looking at the dnsmasq manpage, which you should be doing | 09:39 |
| Vivek | I have already done that couple of times before and edited that dnsmasq.conf file which I gave you. | 09:40 |
| Vivek | I'll try the other options too. | 09:40 |
| twb | Good luck | 09:41 |
| xokvictor | hi all! who can resolv my problem? when my ubuntu server (10.04.2) try booting system is stoped on "run-init: /sbin/init: No such file| Kernel panic" | 09:45 |
| Vivek | twb: I even tried taking a tcpdump on eth0 | 09:48 |
| dforthman | xokvictor did you update your kernel recently? | 09:49 |
| xokvictor | dforthman, yes | 09:49 |
| Vivek | Nothing conclusive to day that a dhcp request was made via eth0, those directories in the rsyslog are what is creating the confusion. | 09:50 |
| dforthman | in grub, there should be an option to boot to the old kernel - if that works you can just keep running that one and try to update again. otherwise, it'll be time to reinstall | 09:50 |
| Vivek | Also it would be kind if some one guides me to the part of the code that creates those directories. | 09:50 |
| Vivek | I have the orchestra, dnsmasq and ryslog source code downloaded via apt-get source. | 09:51 |
| Vivek | They source code is the best documentation so a pointer to where I can find the directory creation part in /var/log/rsyslog/ would be nice. | 09:51 |
| dforthman | Vivek - Can you grep out the directory it creates from the code? That'd be my first stop if I was looking for it | 09:52 |
| xokvictor | dforthman, i press shift button before system boot and select boot olders kernels, but problem persists :( | 09:52 |
| SpamapS | jamespage: I'd love to discuss automated reboot testing when I wake up .. perhaps after the meeting? | 09:53 |
| jamespage | SpamapS, sounds OK to me | 09:53 |
| SpamapS | ok, time to go pass out | 09:54 |
| jamespage | SpamapS, night night! | 09:54 |
| * jamespage thinks it unlikely that SpamapS will chair todays IRC meeting... :-) | 09:54 | |
| xokvictor | dforthman, my way - only reinstall? | 09:55 |
| dforthman | If you can't boot to any kernel, that's the only thing I know to do. Someone else might have a better option for you, though. | 09:56 |
| xokvictor | dforthman, thx | 09:56 |
| Vivek | dforthman: Tried grepping once, could not find it. | 09:57 |
| dforthman | I wish I slept. Today's gonna suck haha | 09:57 |
| === diplo- is now known as diplo | ||
| dforthman | Vivek did you grep through all the subdirectories also? I have no idea what you're trying to do, though. | 09:59 |
| dforthman | alright, time to start getting ready for work. thanks for all your help with the issues i had earlier, guys and gals | 10:01 |
| xranby | jamespage: hi, are the apache cassandra database in ubuntu? | 10:01 |
| jamespage | xranby, no - although the cassandra project does produce some .deb's of their own | 10:02 |
| xranby | OK, i will try those | 10:02 |
| jamespage | Daviey: do meeting minutes go to -server and -devel? | 10:23 |
| jamespage | (well when they get sent at least :-)) | 10:23 |
| Vivek | jamespage: In which channel is the meeting held ? | 11:45 |
| jamespage | Vivek, #ubuntu-meeting | 11:46 |
| Vivek | jamespage: Am I allowed in ther ? | 11:46 |
| jamespage | Vivek, yep! | 11:46 |
| jamespage | its where most ubuntu related IRC meetings are held | 11:47 |
| Vivek | jamespage: Thanks. | 11:47 |
| jamespage | Vivek: np - its normally 1600UTC on a Tuesday | 11:48 |
| Daviey | jamespage: just sevrer is suitable IMO | 11:49 |
| koolhead17 | Daviey: is meeting over? | 11:49 |
| jamespage | Daviey, great - thats what I did | 11:50 |
| jamespage | koolhead17, 1600UTC today | 11:50 |
| * koolhead17 is allready in the channel!! :) | 11:50 | |
| irvie | so i have a fresh ubuntu server install on a linode and i'm running drupal (which uses php's mail settings to send mail by default). what's the easiest way to set up a mail server to send mail from this machine without opening myself up as a spam relay. the drupal install and the mail server would be on the same host | 12:45 |
| irvie | 10.04 | 12:48 |
| acidflash | hi all, in ubuntu 10.04, if i want to add this line -> options ip_conntrack hashsize=32768 ; where do I put it (its supposed to go int /etc/modprobe.conf but i cant find it | 12:55 |
| acidflash | there is only modprobe.d, what should i put inside modprobe.d instead? | 12:55 |
| Vivek | jamespage: Any other ubuntu specific channels that you would recommend ? | 12:59 |
| mdeslaur | jamespage: thanks for testing tomcat6 and for the call for testing! | 13:10 |
| mdeslaur | jamespage: is that another package that showed up on your SRU report? | 13:11 |
| jamespage | mdeslaur, no problem | 13:28 |
| jamespage | it did show up on the SRU report; Daviey poked me about it this morning.... | 13:28 |
| mdeslaur | jamespage: hrm, I wonder how we can make sure it's clear those are security updates...just to make sure the security team publishes the proper USN, and it gets put into -security when it's ready to be published | 13:29 |
| mdeslaur | jamespage: should I always write so in the bugs? I was assuming the SRU report was checking the SRU team membership on the bug, but that's probably not the case | 13:30 |
| smb | zul, bug 928182, not a too big deal and for a change I even tried to be helpful by providing a debdiff. :) | 13:30 |
| uvirtbot | Launchpad bug 928182 in xen-common "xendomains produces ugly output on shutdown" [Low,Triaged] https://launchpad.net/bugs/928182 | 13:30 |
| irvie | so i have a fresh ubuntu server install on a linode and i'm running drupal (which uses php's mail settings to send mail by default). what's the easiest way to set up a mail server to send mail from this machine without opening myself up as a spam relay. the drupal install and the mail server would be on the same host. ubuntu server 10.04 | 13:31 |
| jamespage | mdeslaur, not sure exactly how the report is generated but I suspect its based off the server team packageset + magic! | 13:31 |
| mdeslaur | jamespage: hehe, ok, I'll just write a note in the bug | 13:32 |
| mdeslaur | jamespage: thanks for the testing! | 13:32 |
| irvie | what do i need to install on ubuntu server 10.04 in order to have php send mail? | 13:39 |
| ikonia | irvie: an MTA and php | 13:44 |
| zul | smb: yay! | 13:44 |
| irvie | so if i just install sendmail i'll be fine? is there a way to restrict it to only listen to the local machine? | 13:46 |
| irvie | ie only send mail coming from 127.0.0.1? | 13:46 |
| irvie | i just don't want spammers hijacking the sendmail server | 13:47 |
| ikonia | yeah, thats standard lock down | 13:49 |
| ikonia | ubuntu comes with postfix I think, so you could use that | 13:49 |
| ikonia | unless you know/like sendmail better | 13:49 |
| irvie | i just installed sendmail, so it should already be restricted to 127.0.0.1? | 13:50 |
| smb | zul, And just one observation: seems that the updated xen-common has the effect that you cannot use xm and xl at the same time anymore. Which is probably a good thing. It defaults to xm. I think that is right now a better default because libvirt rather interacts with how xm does things | 13:50 |
| smoser | mtaylor, sorry, just saw your question. i'm not sure why, i'd just have to dig around. maybe ask in #ubuntu-devel | 13:50 |
| irvie | or do i have to actually tell it that | 13:50 |
| zul | smb: right | 13:50 |
| memoryleak | how can I see response body using Curl command line ? | 14:09 |
| andol | memoryleak: response body, as in the http headers? | 14:10 |
| memoryleak | andol: the HTTP headers I see, but no body (like html content) | 14:11 |
| andol | memoryleak: odd, because the default curl behavior is to send the actual content to stdout | 14:12 |
| andol | memoryleak: On the other hand, it doesn't show http headers by default, so perhaps you have a .curlrc tripping you up? | 14:13 |
| memoryleak | what would the option to show the http headers also? | 14:14 |
| memoryleak | andol: ok got it, mixed up uppercase and lowercase -i option | 14:15 |
| stgraber | hallyn: hey, just had an idea, not sure if it'd work though :) Can't we have LXC create /dev/lxc/, put console and all the ttys in there, then have /dev/console and the other be symlinks? | 14:22 |
| stgraber | hallyn: AFAICS this would work whatever the version of Ubuntu in the container and should make upgrading udev/makedev/... work | 14:22 |
| === dduffey_afk is now known as dduffey | ||
| stgraber | hallyn: it won't be perfect because /dev/console and /dev/tty* won't be usable post-upgrade until you reboot, but at this stage you should already have the gettys started anyway and if you upgrade udev/makedev you should reboot post-upgrade regardless | 14:23 |
| stgraber | hallyn: can you remove/mark as merged/abandoned https://code.launchpad.net/~serge-hallyn/ubuntu/precise/upstart/upstart-containers now that all these changes are in my branch? | 14:55 |
| hallyn | ok | 14:57 |
| hallyn | so would lxc while starting up the container do 'rm -f /dev/console; ln -s /dev/tty/console' ? | 14:57 |
| hallyn | i guess so | 14:57 |
| stgraber | hallyn: yes | 15:03 |
| stgraber | hallyn: so on upgrade, the symlink will be replaced by an actual node but at least won't fail and will be replaced next time the container reboots | 15:04 |
| stgraber | hallyn: as we'd put that code in lxc itself, it should just work, whatever the distro in the container | 15:04 |
| hallyn | hopefully no gettys are out there refusing to work over symlinks | 15:10 |
| hallyn | also i could see selinux policy getting in the way | 15:10 |
| hallyn | stgraber: so beside that, getting upstart changes in, and the apparmor mount enforcements, how much more've we got to add to lxc this cycle? | 15:11 |
| hallyn | (perusing bugs) | 15:12 |
| hallyn | all of the upstart-contingent ones of course | 15:12 |
| stgraber | I think that should be pretty much it for 12.04 | 15:13 |
| stgraber | getting a new upstream would be nice though | 15:13 |
| stgraber | *upstream release | 15:13 |
| hallyn | daniel said he'd cut one "in the next few days" (last weekend) | 15:13 |
| stgraber | good, hopefully merging your branch before that :) | 15:14 |
| hallyn | well i did say that a new release is more important, so if merging my branch first is a blocker, then just do that | 15:14 |
| hallyn | oh, right | 15:15 |
| === KeyGruin-zzz is now known as KeyGruin | ||
| hallyn | gary_poster: do you care about lxc-start-ephemeral working with lvm-backed contaienrs? | 15:15 |
| stgraber | hallyn: btw, jodh told me he'd have a look at my upstart branch today/tomorrow | 15:15 |
| stgraber | hallyn: so hopefully by the end of the week we can get rid of lxcguest and announce that you can run a clean Ubuntu system in a container ;) | 15:16 |
| hallyn | unless my luck of this cycle continues, | 15:16 |
| hallyn | where each closed bug opens 2 new ones | 15:16 |
| stgraber | hehe, well, that's what post-FeatureFreeze is for no? :) | 15:17 |
| hallyn | heh | 15:17 |
| hallyn | i guess so. i've got the wrong point of view! | 15:17 |
| hallyn | now, i if i don't test and clean up lxc-clone upstream first, the next merge could be a bear... | 15:19 |
| hallyn | oh! the lxc-ubucloud template | 15:19 |
| roaksoax | Vivek: i'm here now | 15:22 |
| gary_poster | hallyn, I'm pretty sure the answer is "no" because LVM is not part of our current story. | 15:23 |
| Vivek | roaksoax: Hi | 15:23 |
| hallyn | gary_poster: ok, thanks - then i'll leave that to deal with later | 15:23 |
| Vivek | I am still testing some issues dnsmasq was throwing up. | 15:23 |
| hallyn | stgraber: i gather you're not doing the lxc device move right now? | 15:24 |
| Vivek | roaksoax: http://paste.debian.net/155436/ | 15:25 |
| Vivek | That's my dnsmasq.conf | 15:25 |
| gary_poster | ack thanks hallyn | 15:25 |
| Vivek | I've configured it to lease I.P Addresses only though eth1 | 15:25 |
| stgraber | hallyn: correct, though I ran a quick test running getty on a symlink and it seems to work | 15:27 |
| roaksoax | Vivek: ok... | 15:27 |
| stgraber | hallyn: I can do the change if you're busy with other things though | 15:27 |
| Vivek | I find that directories with the 10.x.x.x is created in /var/log/rsyslog even when it is not supposed to do it. | 15:28 |
| Vivek | eth0 serves the 10.x.x.x in my case | 15:28 |
| hallyn | stgraber: i'm happy to do it, but wouldn't get to it until this afternoon (if i try to do it during meetings, experience says it will go badly) | 15:28 |
| Vivek | and eth1 the 192.x.x.x | 15:28 |
| hallyn | stgraber: if you do, just take the bug? i'll check before i start | 15:28 |
| hallyn | thx | 15:28 |
| Vivek | The dnsmasq.conf is as I have posted above. | 15:28 |
| Vivek | My questions is why are those directories appearing ? | 15:29 |
| roaksoax | Vivek: ah yes, i've heard that before | 15:29 |
| roaksoax | though, rsyslog sends the ip address of the orchestra server to the client | 15:29 |
| roaksoax | Vivek: I don't really think dnsmasq has anything to do with that | 15:29 |
| roaksoax | Vivek: let me check the config for a sec | 15:29 |
| Vivek | ok. | 15:29 |
| Vivek | ok. | 15:29 |
| smoser | rbasak, ping | 15:32 |
| smoser | utlemming, maybe you or even Daviey also answer. | 15:33 |
| rbasak | smoser: pong | 15:34 |
| smoser | is there a qemu-system-arm that supports hard float ? | 15:34 |
| rbasak | The beagle one should, I think | 15:35 |
| smoser | ok. thank you. | 15:35 |
| rbasak | Most of what we call hard float is really a userspace ABI thing. All (almost all?) armv7 hardware has actual hard float support | 15:35 |
| roaksoax | Vivek: So yeah, as I ws thinking, the orchestra server preseeds the IP address of itself to the clients. That server address is then used for the clients to fconnect to rsyslog | 15:37 |
| hallyn | stgraber: on second thought i'd prefer to concentrate on the lxc-ubucloud template this afternoon if nothing else comes up | 15:37 |
| roaksoax | Vivek: so I'm guessing 1 of two things. Either rsyslog uses eth0 by defaulkt, or the address of the orchedstra server is in the same 10.0.0.X range | 15:37 |
| hallyn | don't want smoser giving me a noogie at uds | 15:37 |
| roaksoax | Vivek: which causes the clients to use that subnet | 15:37 |
| smoser | rbasak, good enough for me. | 15:38 |
| smoser | rbasak, see email. | 15:38 |
| drPoO | hi all. I am trying to free up some space on my ubuntu server 10.04 /boot directory. I want to remove old kernels to fix a problem I have. I am running 'sudo apt-get remove linux-image-2.6.32-21-server' and get an unmet dependencies error: http://pastebin.com/cP61hzXG | 15:41 |
| drPoO | I dont know how to override the unmet dependencies error so that I can remove the kernel | 15:41 |
| Vivek | roaksoax: I have 2 ethernet interfaces eth0 and eth1, the clients are configured to only use the range 192.168.1.x supplied by eth1 | 15:41 |
| Vivek | roaksoax: They are supposed to ignore the 10.x.x.x provided by the eth0 interface. | 15:41 |
| roaksoax | Vivek: that is in terms of dhcp, correct? | 15:41 |
| Vivek | roaksoax: Yes | 15:42 |
| roaksoax | Vivek: ok so rsyslog and dhcp are not tied up together | 15:42 |
| roaksoax | Vivek: what you would have to do is to change rsyslog config and provide the IP address of the server which is from the range of 192.x.x.x | 15:42 |
| Vivek | Why am I getting the directories with I.P Addresses in the 10.x.x.x range in /var/log/orchestra/rsyslog ? | 15:42 |
| Vivek | roaksoax: What parameter of rsyslog.conf should I change for this to happen ? | 15:43 |
| Vivek | I don't find any range provided in rsyslog.conf now. | 15:44 |
| roaksoax | Vivek: could you pastebin your /etc/rsyslog.d/99-orchestar.conf (or orchestra itself) | 15:44 |
| Vivek | pl | 15:44 |
| Vivek | ok | 15:44 |
| roaksoax | Vivek: you are getting those because the orchestra clients connect to the rsyslog server using the IP address of the 10.x.x.x range (that's what I'm assuming | 15:44 |
| drPoO | is it possible for me to just 'rm' old kernels from the /boot partition without using apt??? | 15:45 |
| Vivek | roaksoax: http://paste.debian.net/155440/ | 15:46 |
| roaksoax | Vivek: that's the server, what about any of the clients? | 15:49 |
| drPoO | is there a way of emptying /boot and reinstalling kernels?? | 15:54 |
| pmatulis | drPoO: treat the kernels like any other package. remove the oldest ones | 16:02 |
| Daviey | Ursinha: What meeting clashes? | 16:02 |
| Ursinha | desktop meeting | 16:03 |
| Ursinha | starting a while after server one | 16:03 |
| Ursinha | Daviey, ^ | 16:04 |
| === Skaag_ is now known as Skaag | ||
| smoser | bug 893926 | 16:14 |
| uvirtbot | Launchpad bug 893926 in eucalyptus "Contains traces of UEC" [High,In progress] https://launchpad.net/bugs/893926 | 16:14 |
| smoser | Daviey, that is you. | 16:16 |
| smoser | dagummit | 16:17 |
| jjohansen | Daviey, hallyn: so mount rules, should be up in a ppa by eod for me | 16:21 |
| hallyn | jjohansen: \o/ thanks! | 16:22 |
| jjohansen | hallyn: the labeling patches, won't hit this week, and if we do them will need a FFE | 16:22 |
| hallyn | jjohansen: is that what's in the test kernel attached to that bug? | 16:23 |
| hallyn | no, that was a bug fix. this is enhanced feature? | 16:23 |
| jjohansen | hallyn: no, that is what I would call the fallback patch. That will be available in the ppa, and part of the pull request | 16:23 |
| hallyn | ok | 16:23 |
| jjohansen | it is independent of labeling | 16:23 |
| hallyn | excellent. thanks. | 16:24 |
| jdstrand | zul: re bug #881464> you gave my the url to the keystonelight branch in the keystone MIR report | 16:24 |
| uvirtbot | Launchpad bug 881464 in keystone "[MIR] keystone" [High,Incomplete] https://launchpad.net/bugs/881464 | 16:24 |
| jdstrand | s/my/me/ | 16:24 |
| hallyn | i skimmed the aa doc (despite it being libreoffice :) So do i gather i'd be able to say "label /some/path XYZ" and later refer to it as XYZ in permissions? | 16:24 |
| zul | jdstrand: i did | 16:24 |
| jjohansen | hallyn: yep, labeling is start are of being able to access the label on files. | 16:25 |
| jjohansen | eg. label=foo rw, | 16:25 |
| jjohansen | much better for files passed into a namespace than connect_disconnected | 16:25 |
| jdstrand | zul: I'm confused. I thought keystone and keystonelight were different (with keystonelight being changed to quantum) | 16:25 |
| zul | jdstrand: no they are different basically keystone->keystonelight | 16:26 |
| hallyn | jjohansen: and so we can assign a label recursively? "everything under /var/lib//lxc/p1 is label=xyz"? | 16:26 |
| hallyn | do that in parent profile, use it in child profile? | 16:26 |
| jdstrand | zul: so is the code I reviewed for this mir not relevant any more? | 16:26 |
| hallyn | if so, nifty. if not, i'm sure it's still nifty :) | 16:26 |
| jjohansen | hallyn: err not exactly, but yes you will be able to use it in a child. | 16:26 |
| zul | jdstrand: right | 16:26 |
| === Skaag_ is now known as Skaag | ||
| hallyn | cool. i'll stop taking your time. thanks :) | 16:27 |
| jdstrand | zul: so I reviewed keystone for nothing?!? | 16:27 |
| zul | jdstrand: no at the time it was every intention of getting it into main, but upstream decided to switch mid cycle | 16:27 |
| zul | jdstrand: im not impressed either | 16:28 |
| jdstrand | impressed. I'm ticked. I don't have a lot of spare cycles for reviews that are obsoleted a month later | 16:28 |
| zul | jdstrand: i totally understand | 16:29 |
| jdstrand | zul: would you mind updating the bug to change the package accordingly and clearly explain the upstream transition so that everyone is clear on what needs to be done? | 16:29 |
| zul | jdstrand: i will | 16:29 |
| jdstrand | fyi, I'm usually spending a full day or more on requested security reviews | 16:29 |
| jdstrand | zul: thanks | 16:29 |
| === shadeslayer is now known as shadeslayer_ | ||
| === shadeslayer_ is now known as shadeslayer | ||
| FunnyLookinHat | Anyone here know how I can utilize SOCKS proxy with SSH to tunnel my SSH connections? I'm doing ssh -D 9999 user@remotehost and I want to go through remotehost to remotehost2 | 16:35 |
| kraut | did you used socks4 in ff f.e.? | 16:38 |
| kraut | it's a socks proxy, not a http proxy | 16:38 |
| rbasak | bug 926160, smoser and utlemming - interesting! Perhaps it's an appropriate time to normalise by passing it through a tarball stage, then mkfs directly to destination size and unpack? The images would probably be more compressible then as well. I have no idea how much work this would be though. | 16:39 |
| uvirtbot | Launchpad bug 926160 in ubuntu "precise cloud-images significantly larger than oneiric" [Medium,Confirmed] https://launchpad.net/bugs/926160 | 16:39 |
| utlemming | rbasak: the initial analysis is that the root cause _is_ the mkfs difference. Between the instance-store and EBS, the journal alone is 96MB bigger for EBS | 16:41 |
| rbasak | yeah, that would make sense. mkfs will make different decisions for performance reasons depending on the requested size | 16:41 |
| utlemming | rbasak: for the instance-store the rootfs is created as a 1.4GB and then grown to 8GB, while the EBS has the rootfs created as a 10GB volume. | 16:42 |
| utlemming | but for EBS we do exactly what you suggest -- grab the tarball, mkfs and then lay down the tarball. | 16:42 |
| rbasak | Then perhaps the size difference is acceptable. If this means what I think it does, then performance will be reduced for the smaller images otherwise. | 16:46 |
| === kentb_ is now known as kentb_dell | ||
| roaksoax | jamespage: | 16:49 |
| roaksoax | jamespage: fence-agents and cobbler now ship the fence_cud and template respectively, if you can it it a try would be very much appreciated | 16:49 |
| jamespage | roaksoax, great - will do - is that in precise only? | 16:50 |
| roaksoax | jamespage: yes | 16:50 |
| Daviey | smb: you are welcome to join the chair roster if you want :) | 17:02 |
| smb | Daviey, Thanks a lot but .... no | 17:02 |
| Daviey | lol | 17:03 |
| zul | we should get community people to do it as well like ttx :) | 17:03 |
| ttx | zul: you still had community people, by then :P | 17:05 |
| zul | ttx: hehe | 17:05 |
| smoser | i've recently seen ttx commenting on how he is ubuntu core dev | 17:05 |
| smoser | i think part of that requirement shoudl be running meetings | 17:05 |
| koolhead17 | ttx: i liked your forkhub related blog though!! :P | 17:06 |
| jamespage | utlemming: issues? https://jenkins.qa.ubuntu.com/view/Precise%20ISO%20Testing%20Dashboard/view/Daily/job/precise-server-ec2-daily/ | 17:09 |
| jamespage | utlemming, no EBS anymore? | 17:10 |
| Vivek | roaksoax: I have to call it a day now. | 17:10 |
| utlemming | jamespage: smoser fixed that.... | 17:10 |
| Vivek | roaksoax: Catch you tommorow. | 17:10 |
| === sixstringsg|away is now known as sixstringsg | ||
| smoser | rbasak, the difference that is not understood in that is that apparent size (per du) is identical, but we still have used filesystem growth. | 17:15 |
| smoser | and significant | 17:15 |
| rbasak | smoser: that's expected, surely? | 17:16 |
| smoser | why? | 17:16 |
| rbasak | filesystem overhead | 17:17 |
| smoser | no | 17:17 |
| smoser | rbasak, see 'file tot' lines | 17:18 |
| smoser | filesystem total went up by 2.9% | 17:18 |
| smoser | but 'df' grew by 15% | 17:18 |
| smoser | utlemming, suggested that maybe it is symlinks causing the additional fs usage | 17:19 |
| smoser | due to python3 | 17:19 |
| CantWinn | Hello peeps | 17:20 |
| CantWinn | Does anyone know of a great "exchange" style open source server? | 17:21 |
| hallyn | smoser: there is no way to have ubuntu-cloudimg-query directly spit out the '*-root.tar.gz' url right? | 17:31 |
| Mez | ok, I have a new dedicated server - which is provided with a custom kernel - trying to boot up off of the normal kernel causes it to not come up (I believe this is cause it's not starting the network card) | 17:34 |
| Mez | can anyone suggest a way of getting it up? I can provide custom boot dmesg vs ubuntu kernel dmesg | 17:34 |
| === zz_FridgeBoX is now known as FridgeBoX | ||
| === sixstringsg is now known as sixstringsg|away | ||
| roaksoax | Daviey: you want me to ship a preseed just for the enlist stuff or would it be better to just add a snippet into the regular preseed | 18:29 |
| Daviey | roaksoax: Requirements are: it gets pushed to a system that cobbler hasn't previously seen. The second requirement is, it should be possible to overide variables via the API | 18:31 |
| Daviey | outside that, whatever you think best :) | 18:31 |
| roaksoax | Daviey: hehe right, but "system that cobbler hasn't previously seen" that's handled by the enlistment process, isn't it? | 18:32 |
| Daviey | roaksoax: erm, right - you are doing the enlistment stuff, right? | 18:33 |
| Daviey | roaksoax: The default preseed is a way of passing values where enlist should push back it's details.. | 18:33 |
| roaksoax | Daviey: i'm doing the cobbler side | 18:33 |
| Daviey | such as hostname | 18:33 |
| Daviey | right | 18:33 |
| roaksoax | Daviey: excalty, so my point being is should I install something like: enlistment.preseed and work from there, or use orchestra-client.preseed with a snippet that passes the enlistment stuff | 18:34 |
| Daviey | d-i boots dumb, the preseed provides values to the enlist udeb on how to operare. Right? | 18:34 |
| Daviey | Oh, and tftp default flavour, right? | 18:34 |
| Daviey | roaksoax: i think a seperate snippet sounds tastu | 18:35 |
| Daviey | tasty | 18:35 |
| roaksoax | Daviey: yeah I'm doing this: 1. create a cobbler-enlist profile with a special preseed. 2. create a 'default' system 3. cobbler sync --> next client that tries to boot and has no system, defaults to the cobbler-enlist profile | 18:36 |
| Daviey | roaksoax: that sounds beautiful | 18:36 |
| roaksoax | Daviey: and I'm basing this on a i386 image | 18:36 |
| Daviey | roaksoax: I think for now that works just fine, rbasak and Justin are currently investigating ways to differentiate arm.. You'll need to be prepaired to take their changes | 18:38 |
| roaksoax | Daviey: cool | 18:38 |
| Daviey | (probably something like a system default-armhf or something) | 18:38 |
| adam_g | smoser: if this looks okay with you, i'll merge into my branch. https://code.launchpad.net/~gandelman-a/+junk/smoser-deployer | 18:39 |
| smoser | looking | 18:40 |
| adam_g | smoser: nothing major, renamed the argument and avoided an error when running it against an environment that already has all services deployed, which is useful when a relation has failed, you've resolved and want to continue on with relations | 18:41 |
| * Daviey goes afk | 18:41 | |
| smoser | adam_g, seems good enough to me. | 18:43 |
| adam_g | cool | 18:44 |
| roaksoax | adam_g: by any chance do you have the latest enlistment preseed available? | 18:47 |
| adam_g | roaksoax: i never had one, i was only working on the cobbler-enlist utility. i think there is something on the CI cobbler server, let me see | 18:49 |
| adam_g | roaksoax: http://paste.ubuntu.com/832978 | 18:50 |
| roaksoax | adam_g: cool thanks | 18:51 |
| hallyn | (going afk for awhile) | 19:19 |
| args[0] | i wanna add a sub-domain to my server, how can i do that using CLI? | 19:24 |
| soren | zul, Daviey: Do you guys not you dch when updating changelogs? | 19:38 |
| soren | Err... | 19:38 |
| soren | zul, Daviey: Do you guys not use dch when updating changelogs? | 19:38 |
| zul | soren: yes | 19:38 |
| soren | zul: You do? | 19:38 |
| zul | soren: yes | 19:38 |
| soren | zul: So you manually go in afterwards and turn the multi-maintainer changelog things into a non-standard format? | 19:39 |
| zul | soren: how do you do it? | 19:39 |
| soren | dch "Did some stuff" | 19:40 |
| soren | done | 19:40 |
| zul | soren: ok | 19:40 |
| soren | How do you do it? | 19:40 |
| zul | vi | 19:40 |
| soren | zul: Do you use vi or do you use dch? | 19:41 |
| zul | i use vi | 19:41 |
| soren | Ok. Can you please invoke vi by calling dch from now on? | 19:41 |
| zul | yep | 19:41 |
| soren | With no arguments, it runs $EDITOR | 19:41 |
| soren | Only difference is that dch ensures standard format is used for changelog entries where multiple people participate. | 19:42 |
| soren | Also, can you please not leave those UNPUBLISHED changelog entries in there? | 19:42 |
| zul | in which branch? | 19:43 |
| soren | All of 'em? | 19:43 |
| zul | k | 19:43 |
| bencer | is anybody willing to sponsor new zentyal (old ebox) packages? we mailed -server ml but nobody stepped, only robbie suggesting to mail -sponsors. soren maybe you as our first sponsor? :) | 19:43 |
| soren | bencer: Where are the packages? | 19:44 |
| bencer | soren: https://launchpad.net/~jacalvo/+archive/zentyal-precise/ | 19:45 |
| soren | bencer: How are upgrades handled? | 19:46 |
| bencer | no upgrade path, we have a migration tool from 2.0 to 2.3, but is beta and not packaged | 19:47 |
| bencer | actually, packages on lucid and later releases are quite broken | 19:47 |
| bencer | nothing that depends on ldap works | 19:48 |
| bencer | so i think doesn't worth to provide a migration path | 19:48 |
| bencer | anyway, we can improve that, but we are in kind of a hurry now | 19:48 |
| bencer | because feature freeze is in a week, and after that no new packages will be accepted without an exception | 19:49 |
| kagz | hi!! | 20:08 |
| robo | is apt-get clean safe to run? It says that it "clears out the local repository of retrieved package files." and that sounds important | 20:09 |
| Daviey | soren: wait, what have i done? | 20:09 |
| Daviey | soren: UNPUBLISHED or UNRELEASED ? | 20:09 |
| ChmEarl | robo, /var/cache/apt/archives gets cleared. Back it up first if needed | 20:10 |
| Daviey | bencer: Are these NEW packages, or new versions? | 20:10 |
| robo | ah, ty che3ver | 20:11 |
| robo | ChmEarl, | 20:11 |
| Driip | Hello Ubuntu Community! I was wondering if any of you know any Web interface Firewall for ubuntu. Cant seem to find any decent once which is "easy" to port forward etc | 20:11 |
| Daviey | as in, all of these packages have a lower version already in the archive? | 20:11 |
| ChmEarl | robo, locate "*.deb" -> results should be from that dir | 20:11 |
| kagz | robo iits always good to first backup as chmearl says | 20:11 |
| robo | i'm in a bit of a spot here. I built a package for node.js called custom-nodejs.deb which I told apt that it requires libssl-dev. So I went to install libssl-dev and I got this error: "libssl-dev: Depends: zlib1g-dev but it is not going to be installed". Then I try to install zlib1g-dev and I get: custom-nodejs: Depends: libssl-dev but it is not going to be installed | 20:15 |
| _ruben | robo: try sudo apt-get install -f | 20:16 |
| robo | _ruben, will that cause unexpected behavior? | 20:16 |
| bencer | Daviey: are new packages that replace the old ones, now the project is called zentyal, was called ebox before | 20:16 |
| _ruben | robo: it'll try to fix any inconsitencies by installing packages that are missing by dependencies | 20:16 |
| bencer | Daviey: they have the replace headers in control and so on | 20:17 |
| Driip | bencer: what is Zentyal exactly? is it some kinda wrapper for ubuntu or does like change the whole operating system | 20:17 |
| Driip | like would i be able to still use the ssh to do normal commands in ubuntu. | 20:17 |
| robo | _ruben, I think i'm going to take a snapshot just in case this breaks something :-) | 20:17 |
| bencer | Driip: it manages some services in a "best as we can" way | 20:17 |
| bencer | we you enable a module, admin gets a notification of files going to be managed | 20:18 |
| Driip | bencer: so its basicly just like webmin? | 20:18 |
| bencer | you can merge you changes with zentyal configuration editing these templates | 20:18 |
| bencer | Driip: actually not, we try to be distribution friendly | 20:18 |
| bencer | and not to smash the system | 20:18 |
| _ruben | robo: it shouldn't .. and it'll ask for confirmation after telling you what it'll do .. iirc :) | 20:18 |
| robo | ah | 20:19 |
| Driip | bencer: but apart from that it would be a yes? :P | 20:19 |
| robo | okay | 20:19 |
| bencer | Driip: is not only a web interface, we do integration between the modules and services | 20:19 |
| === sixstringsg|away is now known as sixstringsg | ||
| bencer | all modules talk to ldap, firewall, etc modules | 20:19 |
| bencer | and we use .d folders where possible | 20:19 |
| Driip | bencer: But lets say, i have other services like Game servers etc. would this be a problem to run on ubuntu after a Zentyal installation. or is the base system still accessible? What im afraid of is that my system goes all clearOS on me | 20:20 |
| bencer | Driip: of course not, we are a set of packages that we touch only the modules we manage | 20:21 |
| bencer | we dont destroy the system, you can still work with apt like you are used to | 20:21 |
| bencer | and if you disable a module, you can keep on modifing the conf files as you would expect on plain ubuntu | 20:21 |
| Driip | that sounds sweet | 20:22 |
| Driip | does this work with debian aswell or is it only ubuntu support ATm? | 20:22 |
| bencer | this is one of the things more people like of zentyal, they install the packages to manage the server, and the things they found missing, they configure them by hand | 20:22 |
| bencer | should be pretty easy to use it on debian | 20:22 |
| bencer | but i didnt have the time to test it | 20:23 |
| bencer | ubuntu server is our priority | 20:23 |
| Driip | mhm i see | 20:23 |
| bencer | we use upstart for many things, but that's present on debian too | 20:23 |
| Driip | i use ubuntu on most of my servers. but debian tends to be more stable | 20:23 |
| Driip | ill actually try it on debian first and if it all works smoothly ill go for it on rest of em :) | 20:24 |
| Driip | looks awsome | 20:24 |
| bencer | well i don't think there is a huge different, both are quite similar and i prefer ubuntu server for some things | 20:24 |
| bencer | like the guarantee support lifecycle | 20:24 |
| bencer | or having upstart to respawn servicies out of the box, and not having to install runit or things like that | 20:25 |
| Driip | thats true :). I had a problem with my server a while ago where my NIC or whatever went to sleep without any reason. And i randomly appeared on the network again after x amount of time (normally hours) | 20:25 |
| Driip | after that i had to port some of my servers to debian | 20:25 |
| Driip | never figured it out | 20:25 |
| bencer | soren, Daviey is you find the time to sponsor the packages, would be great, let me know, please | 20:26 |
| Driip | Anyways. thats for the help/explanation. ill get back to work now! | 20:27 |
| Driip | thanks* | 20:27 |
| soren | Daviey: Looking more closely, it seems it may have been bzr rebase that screwed with me. I'm not entirely sure, though. More proof the rebasing is evil. | 20:39 |
| === sixstringsg is now known as sixstringsg|away | ||
| roaksoax | Daviey: still around? | 20:54 |
| Eruadan | hi | 20:58 |
| RoyK | ho | 21:03 |
| Eruadan | sup RoyK | 21:06 |
| roaksoax | smoser: around? | 21:08 |
| smoser | roaksoax, here | 21:08 |
| jamespage | SpamapS, ping | 21:09 |
| hallyn | smoser: so the cloud image has lxcguest built in, right? template shouldn't have to do anything after downloading it? | 21:11 |
| smoser | hallyn, correct | 21:12 |
| smoser | it should boot in lxc. | 21:12 |
| smoser | if not, you have something to fix | 21:12 |
| smoser | :) | 21:12 |
| hallyn | and what about username/pwd? | 21:13 |
| SpamapS | jamespage: oh my! you're still here. :) | 21:16 |
| jamespage | SpamapS, took a break and eat pizza | 21:16 |
| jamespage | ate that is | 21:17 |
| SpamapS | MMMMMmmm pizza | 21:19 |
| SpamapS | jamespage: I'm actually modifying the jenkins charm to allow plugins. :) | 21:19 |
| jamespage | SpamapS, sweet! How are you enabling them? | 21:20 |
| SpamapS | jamespage: dropping them in /var/lib/jenkins/plugins :) | 21:20 |
| SpamapS | 2012-02-07 13:20:18,416 unit:jenkins/0: hook.output ERROR: Issued certificate has expired. | 21:20 |
| SpamapS | Unfortunately, updates.jenkins-ci.org's cert expired yesterday | 21:20 |
| jamespage | bah! | 21:20 |
| jamespage | are you using the distro packaging in oneiric? | 21:21 |
| SpamapS | jamespage: yes still oneiric | 21:22 |
| jamespage | SpamapS, hmm - thats tickling something at the back of my mind... | 21:22 |
| SpamapS | jamespage: more plugins in precise now? | 21:22 |
| * jamespage goes to take a look | 21:22 | |
| jamespage | SpamapS, I remembered something but if was for plugin signing not the cert on the website. | 21:24 |
| jamespage | hmm - I know that the core jenkins folk have been travelling | 21:24 |
| SpamapS | jamespage: its ok, this just needs --no-check-certificate to wget until they fix it | 21:24 |
| jamespage | SpamapS, you should take a look at jenkins-cli as well | 21:25 |
| jamespage | SpamapS, I see discussion about this issue on #jenkins | 21:25 |
| jamespage | SpamapS, jenkins-cli -s http://localhost:8080 install-plugin gives more help | 21:26 |
| jamespage | it talks to jenkins over its RESTful API and tells it to download X plugin from the update-centre | 21:27 |
| jamespage | you can get it to restart once they have been installed as well | 21:27 |
| === Jasonn is now known as TEH-JASON | ||
| jamespage | SpamapS, so you wanted to discuss reboot testing? | 21:30 |
| Daviey | roaksoax: long enough to write this :) | 21:30 |
| SpamapS | jamespage: yes, give me a couple minutes... and thanks for the tip on jenkins-cli | 21:30 |
| roaksoax | Daviey: so i'll give you something to think | 21:30 |
| roaksoax | Daviey: the enlister has d-i cobbler-enlist/host-profile string precise-x86_64 | 21:30 |
| roaksoax | enlister preseed* | 21:31 |
| roaksoax | so that's the profile for the system. How do we select on automatically? | 21:31 |
| roaksoax | Daviey: the same applies for user/pass | 21:31 |
| roaksoax | Daviey: there's no way to determine it automatically | 21:31 |
| Daviey | roaksoax: make it a templatetag that can be updated/overidden via the api? | 21:33 |
| jamespage | SpamapS, if I manage to pull my finger out in the next few days I'll see if I can enable the SSH module in the precise version | 21:34 |
| jamespage | means you don't have to have java installed | 21:34 |
| jamespage | todo remote admin | 21:35 |
| roaksoax | Daviey: what do you mean? | 21:35 |
| roaksoax | Daviey: http://paste.ubuntu.com/833205/ | 21:38 |
| Daviey | roaksoax: I'm sorry, but i can't talk right now.. give me 30mins | 21:38 |
| roaksoax | Daviey: no worries ;) | 21:39 |
| adac | I'm trying to install etherpad (via etherpads deb repo) and it tries then to install all this packages: http://pastebin.com/hJ80QBkK I was wondering if thes are grafical packages? | 21:42 |
| adac | mean if it does intalla grafical UI | 21:42 |
| jamespage | adac, etherpad uses java itself and also uses openoffice for document generation | 21:44 |
| rallias | is there software that I can employ the use of that allows me to encrypt the contents of my server's files and transparently decrypt them upon use? | 21:45 |
| adac | jamespage, hmm so it isntalls a graifacl desktop? | 21:45 |
| adac | grafical | 21:45 |
| rallias | preferably that I can install after the server is up and running. | 21:45 |
| jamespage | I would suggest looking at etherpad-lite - its entirely node.js and uses abiword instead for doc generation which is a much smaller footprint | 21:45 |
| jamespage | adac, unfortunately I think that a full Java JDK is required which does pull in some X related libraries; but not a full desktop | 21:47 |
| adac | jamespage, i see yes | 21:47 |
| adac | jamespage, maybe light would be enough though...but i guess both need these grafical packages | 21:47 |
| jamespage | adac, no - etherpad-lite does not use Java at all | 21:48 |
| jamespage | (not that I have anything against Java :-)) | 21:48 |
| adac | :) | 21:48 |
| adac | jamespage, yes I understand...but abiword is grafical, right? | 21:48 |
| jamespage | adac: well it runs headless under etherpad-lite - and its dependency footprint is much smaller | 21:50 |
| SpamapS | jamespage: ok, so.. did you get my email about the test stuff? | 21:50 |
| jamespage | SpamapS, I did - just checked it out | 21:50 |
| jamespage | want to walk me through it? | 21:50 |
| adac | jamespage, i guess etherpad-light just won :) | 21:50 |
| jamespage | adac: its much more active in terms of development community | 21:51 |
| jamespage | adac: https://github.com/Pita/etherpad-lite | 21:52 |
| jamespage | 1.0 works pretty well | 21:52 |
| adac | cool thanks! | 21:52 |
| SpamapS | jamespage: Its been updated a lot since I sent that.. but it seems to work for me | 21:52 |
| SpamapS | jamespage: should get the result of both the post-reboot test and the regular test.. | 21:52 |
| rallias | is it possible to change the home directory of a service user such as www-data and enable encryptfs with it in the new home directory? | 21:52 |
| Eruadan | i'm learning to program...what should i do firt, ror or node.js? | 21:53 |
| jamespage | SpamapS, so I remember that run_test gets injected into the initrd that boots and installs the system; | 21:53 |
| Eruadan | first*# | 21:53 |
| jamespage | how are you adding boottests.py and the reboot_test_cases? | 21:53 |
| SpamapS | jamespage: same way | 21:53 |
| rallias | Eruadan, Whichever you see as more useful in the field you wish to program in. | 21:54 |
| SpamapS | jamespage: boottests.py goes in the initrd too.. reboot_test_cases goes in couchdb | 21:54 |
| jamespage | SpamapS, so did you make changes to ubuntu-server-iso-testing as well? | 21:54 |
| Eruadan | i see a lot of rails apps ready, i feel more secure with ruby. But i see javascript and node beautiful, just imature | 21:54 |
| SpamapS | jamespage: yes | 21:55 |
| SpamapS | jamespage: oh I thought I included that.. hang on | 21:55 |
| Eruadan | and i'm into web programming | 21:56 |
| Eruadan | it seems ^^ | 21:56 |
| SpamapS | jamespage: lp:~clint-fewbar/ubuntu-server-iso-testing/add-post-reboot-tests | 21:56 |
| SpamapS | jamespage: this is a bit of a mess really.. hard to follow and understand without days of covering yourself in the mud | 21:56 |
| Eruadan | rallias, do you know ruby? | 21:57 |
| jamespage | SpamapS: agreed - I've hardly touched run_test since it was first written many moons ago | 21:57 |
| jamespage | hence why qa want to stand back from this and re-develop | 21:57 |
| SpamapS | jamespage: I thin juju would actually be an easy choice for this | 21:59 |
| SpamapS | jamespage: it just needs a libvirt provider. :) | 21:59 |
| jamespage | SpamapS, lol | 21:59 |
| Eruadan | i might leave node.js for a second step, ror seems a nice way to get out of drupal chain, and learn something useull | 21:59 |
| Eruadan | i do site building for 2 years already | 22:00 |
| SpamapS | Eruadan: trading drupal for ror is like trading a sperm whale for a grey whale... they're both mostly blubber | 22:01 |
| rallias | Eruadan, No. The only languages I know are perl and php. | 22:02 |
| jamespage | hm - I seem to have a magical new keyboard shortcut that minimizes all my windows! | 22:02 |
| jamespage | SpamapS, well bearing in mind the context of the code changes it looks OK to me | 22:06 |
| SpamapS | jamespage: the end result is an extra results file.. | 22:07 |
| jamespage | yeah | 22:07 |
| SpamapS | -rw-rw-r-- 1 clint clint 847 Feb 4 12:06 POST-REBOOT-TEST-lamp-reboot.xml | 22:07 |
| SpamapS | jamespage: actually I think there's a bug as that is duplicated from the regular result | 22:08 |
| SpamapS | jamespage: but, once that is fixed.. whats the next step to get those feeding jenkins? | 22:08 |
| jamespage | you might want to rename that to TEST- POST-REBOOT-lamp-reboot.xml | 22:08 |
| === TEH-JASON is now known as Jasonn | ||
| jamespage | Jenkins looks for **/TEST-*.xml | 22:08 |
| jamespage | SO; Step 1) would be to propose the changes to ubuntu-server-iso-testing | 22:09 |
| Eruadan | SpamapS, i just cannot learn drupal module development, it's a mess | 22:09 |
| jamespage | jibel and hggdh will want to review as well | 22:09 |
| Eruadan | rails is much more clean, in my view | 22:09 |
| jamespage | Minor glitch maybe that we agreed not todo any more feature development in this project | 22:09 |
| jamespage | but its prob OK. | 22:10 |
| jamespage | Once its merged we can then get it deployed in the lab with some new jenkins config; that will really need QA team involvement | 22:10 |
| jamespage | I can do it but they are taking more of a lead these days.... | 22:11 |
| SpamapS | jamespage: heh, yeah, nobody told me that we can't do features. :) | 22:11 |
| hallyn | stgraber: would it be deemed an unholy hack to have dnsmasq by default expliclty say --except-interface=virbr0,lxcbr0? | 22:11 |
| jamespage | SpamapS, well there ain't nowhere else to deliver it from yet! | 22:12 |
| hallyn | because ppl get bitten by it more often than i'd like | 22:12 |
| SpamapS | jamespage: ok I'll get it a little cleaned up. | 22:12 |
| stgraber | hallyn: what kind of issue is it causing by not having --except-interface? (yes it'd be a pretty big hack causing even weirder bug reports) | 22:13 |
| hallyn | there's bug 925511 | 22:15 |
| uvirtbot | Launchpad bug 925511 in lxc "lxc init script should fail when it ... failed" [High,Fix released] https://launchpad.net/bugs/925511 | 22:15 |
| hallyn | (looking ofr the most recent libvirt one) | 22:15 |
| SpamapS | jamespage: hrm, jenkins-cli would be helpful if I could coax it to ignore bad certs. ;) | 22:16 |
| jamespage | SpamapS, lol | 22:17 |
| hallyn | stgraber: bug 918807 and bug 925511 i guess | 22:17 |
| SpamapS | jamespage: thanks btw for the review. I release you to bedtime. :) | 22:17 |
| uvirtbot | Launchpad bug 918807 in libvirt "[precise] virbr0 not created during fresh install when dnsmasq is running with --except-interface=virbr0" [Low,Confirmed] https://launchpad.net/bugs/918807 | 22:17 |
| uvirtbot | Launchpad bug 925511 in lxc "lxc init script should fail when it ... failed" [High,Fix released] https://launchpad.net/bugs/925511 | 22:17 |
| jamespage | well once that is fixed .... | 22:17 |
| SpamapS | jamespage: actually I think it can just take a bare http:// url for the .hpi so it should work fine | 22:18 |
| robbiew | zul: ping | 22:19 |
| SpamapS | jamespage: https://code.launchpad.net/~clint-fewbar/charms/oneiric/jenkins/add-plugins/+merge/91930 .. if you get bored. :) | 22:19 |
| stgraber | hallyn: well, the problem is going to be when someone creates virbr1 | 22:24 |
| robbiew | zul: nevermind | 22:24 |
| kantlivelong | anyone here use NUT for UPS monitoring? | 22:25 |
| hallyn | stgraber: at that point they can surely take care of themselves | 22:25 |
| hallyn | i don't suppose --except-interface can handle wildcards :) | 22:26 |
| robbiew | kantlivelong: roaksoax/RoAkSoAx *might* be able to help | 22:26 |
| robbiew | kantlivelong: Uzuul is the real expert -> https://launchpad.net/~aquette | 22:27 |
| robbiew | I don't think he's around, but you can certainly email him...he's driving this spec for 12.04 https://blueprints.launchpad.net/ubuntu/+spec/servercloud-p-cloud-power-management | 22:27 |
| kantlivelong | robbiew: its just kinda weird.. | 22:28 |
| kantlivelong | instead of waiting for a low battery warning it just shuts down after X seconds of being on batt.. kinda sillly.... | 22:29 |
| MTecknology | kirkland: You happen to have a PPA for oVirt? :D | 22:30 |
| kirkland | MTecknology: no, sorry | 22:30 |
| kirkland | robbiew: is anyone working on oVirt at this point at Canonical? | 22:31 |
| MTecknology | kirkland: :'{ You should. I'm having a bitch of a time trying to build it | 22:31 |
| kirkland | MTecknology: heh | 22:31 |
| robbiew | in terms of building it...nope | 22:31 |
| kirkland | MTecknology: ^ | 22:31 |
| MTecknology | so.. so work being done to make it a nice sexy package? | 22:32 |
| hallyn | smoser: if i set root:root password in lxc-ubucloud, are you going to cry? | 22:32 |
| MTecknology | Any chance you can help me with this?.. http://dpaste.com/699214/ | 22:32 |
| hallyn | smoser: because otherwise, since agetty clears the screen on startup, i miss the supposed printing out of ubuntu user password | 22:32 |
| hallyn | maybe i'll do ubuntu:ubuntu | 22:33 |
| robbiew | MTecknology: yuck :/ | 22:33 |
| hallyn | oh no then your cloud still will overwrite it | 22:33 |
| MTecknology | robbiew: I was trying to follow with http://www.ovirt.org/wiki/Building_Ovirt_Engine using the repos for everything except jbossas | 22:34 |
| utlemming | hallyn: couldn't you just use "--noclear" for agetty? | 22:34 |
| hallyn | utlemming: it's your f'ing image | 22:34 |
| hallyn | you tell me | 22:34 |
| robbiew | MTecknology: yeah...so this is why we haven't tackled ovirt...JBoss | 22:35 |
| hallyn | i'll just do root:root until someone has a better idea | 22:35 |
| hallyn | or did i misread? is passw0rd always the ubuntu pwd? | 22:36 |
| robbiew | MTecknology: jamespage is our resident java man...but he's probably asleep..or should be (in the UK) | 22:36 |
| robbiew | you could try posting to ubuntu-server...loads of folks on there | 22:36 |
| === sixstringsg|away is now known as sixstringsg | ||
| MTecknology | robbiew: perhaps I'll bug him when he's awake. I'm about to go home for the day, if I don't catch him before the morning, I'll try the ML. :) | 22:37 |
| MTecknology | robbiew: I _should_ be able to do this not horribly painfully, right? | 22:37 |
| robbiew | heh | 22:38 |
| utlemming | hallyn: the default password for the ubuntu images is ubuntu:ubuntu with the user password disabled | 22:38 |
| robbiew | MTecknology: sure..on Redhat :/ | 22:38 |
| MTecknology | :( | 22:38 |
| hallyn | utlemming user password disabled? so how do i log in? | 22:38 |
| robbiew | unfortunately other distros (even SuSE) have additional hurdles | 22:38 |
| jamespage | MTecknology, try running maven with a -X flag - it should tell you a bit more | 22:39 |
| MTecknology | robbiew: I don't wanna; are there any other things I could use that give non-linux guys a pretty way to manage a few hundred hosts? | 22:39 |
| utlemming | hallyn: cloud-init needs to enable it...or we need to rethink the tooling of the images in that context | 22:39 |
| MTecknology | jamespage: I restarted it with -e, I'll do with -X | 22:39 |
| robbiew | MTecknology: landscape.canonical.com..but of course, I assume you mean for free ;) | 22:40 |
| MTecknology | robbiew: landscape will manage deploying VM's on a KVM host? | 22:41 |
| robbiew | MTecknology: no idea, tbh...fwiw, we are involved with ovirt...just not actively working on enabling it this cycle...plate is more than full, if you know what I mean | 22:41 |
| robbiew | oh..no | 22:41 |
| robbiew | lol | 22:41 |
| robbiew | sorry | 22:41 |
| robbiew | misunderstood | 22:41 |
| MTecknology | I do think landscape is nice and pretty, but higher up guys wouldn't go for it | 22:42 |
| robbiew | understand | 22:42 |
| MTecknology | k- it's go home time. | 22:42 |
| hallyn | utlemming: would it be fair for me to push the lxc-ubucloud template as I ahve it, and let you open a bug to explain how it should be done better? (cause i don't understand what 'cloud-init needs to enable it' means) | 22:42 |
| robbiew | MTecknology: have a good one | 22:42 |
| utlemming | hallyn: do you have a link to your template? | 22:43 |
| hallyn | smoser: utlemming: stgraber: and, what woudl be a better name than ubucloud | 22:43 |
| jamespage | MTecknology, this might be a stupid question but how much free RAM do you have? | 22:43 |
| hallyn | utlemming: one sec, i'll push it | 22:43 |
| utlemming | but yeah, that is probably the way to do it | 22:43 |
| MTecknology | -/+ buffers/cache: 408 88 | 22:43 |
| robbiew | MTecknology: http://openetherpad.org/ovirt-on-debianubuntu ...fyi | 22:43 |
| jamespage | When java forks it can be very memory intensive at the point it forks and it won't push into swap | 22:43 |
| hallyn | utlemming: http://people.canonical.com/~serge/lxc-ubucloud | 22:44 |
| MTecknology | jamespage: you think that might be killing it? | 22:44 |
| jamespage | MTecknology, probably | 22:44 |
| hallyn | utlemming: 'lxc-create -t ubucloud -n cloud1' works fine for me with that template (and i can start it just fine too) | 22:45 |
| jamespage | if it hit 650M usage and tried to fork it would need another 650M of non-swap memory to fork successfully | 22:45 |
| jamespage | (or so I remember - got caught out with forking solr cache warmups in an application server) | 22:45 |
| jamespage | it might not need that afterwards.... | 22:46 |
| jamespage | MTecknology, leave me a message if that does not sort you out - as robbiew says I should be in bed... | 22:47 |
| smoser | hallyn, ubuntu-cloud-images | 22:47 |
| MTecknology | jamespage: will do- thanks :D | 22:48 |
| Eruadan | just have this node.js running my head....I might learn javascript, it seems more purposes than ruby.Besides, you need to know ajax anyway... | 22:51 |
| Eruadan | yes, i will go for javascript, i decided now | 22:52 |
| utlemming | hallyn: how about http://paste.ubuntu.com/833283/ | 22:52 |
| * robbiew wonders if one could install the dependencies and then use alien on the ovirt packages | 22:54 | |
| hallyn | utlemming: so cloud-init itself won't mess with it? | 22:56 |
| === fysics__ is now known as fysics | ||
| utlemming | hallyn: yes, cloud-init will leave the Ubuntu user alone unless the user sends a cloud-config to mess with it | 22:56 |
| hallyn | utlemming: testing, but it's fine with me long's it works | 22:57 |
| hallyn | utlemming: hm, usermod -U ubuntu: No such file or directory | 22:58 |
| * utlemming looks | 22:58 | |
| hallyn | (trying with full path, just in case) | 22:59 |
| utlemming | and it should be just "ubuntu" not "ubuntu:" | 22:59 |
| hallyn | utlemming: that's just part of the error msg | 23:00 |
| utlemming | ah | 23:00 |
| hallyn | oh, i don't think you can do the quotes :) | 23:00 |
| utlemming | :(, really? | 23:00 |
| utlemming | surprising | 23:01 |
| hallyn | there is command called "usermod -U ubuntu" :) | 23:01 |
| hallyn | nah, chroot isn't that smart | 23:01 |
| hallyn | or, it doesn't want to maek assumptions about yoru pathanmes | 23:01 |
| utlemming | for image building we do chroot "meh" all the time | 23:01 |
| hallyn | (after all, space is ok in a pathname) | 23:01 |
| hallyn | yes | 23:01 |
| hallyn | but not "meh hello workd" | 23:01 |
| hallyn | yeah that worked | 23:01 |
| hallyn | well, haven't fired it up yet | 23:01 |
| hallyn | stgraber: feh! my reboot patch for lxc is messed up still. my check for reboot support is bogus. | 23:03 |
| hallyn | utlemming: ok, all good. I'll do it that way, thanks. | 23:04 |
| hallyn | utlemming: question still stands, do you want a different name, or is ubucloud ok? | 23:06 |
| utlemming | hallyn: ubucloud is kind of catchy, but probably ubuntu-cloud | 23:06 |
| utlemming | for branding consistency | 23:06 |
| hallyn | bleh. that's a lot of typing | 23:07 |
| hallyn | but ok | 23:07 |
| hallyn | maybe we need bash completion for lxc-create | 23:07 |
| === dduffey is now known as dduffey_afk | ||
| mgw | has anybody experienced an issue where the partition manager (in the installer) keeps returning to the main screen if you select anything other than "Guided - Entire Disk"? | 23:14 |
| mgw | (11.10) | 23:15 |
| JanC | mgw: I used it fairly recently and didn't see that | 23:19 |
| mgw | JanC: thanks, i wonder what's going on; do you know if the server images are updated with bug fixes? | 23:20 |
| mgw | within a specific release (e.g., 11.10) | 23:21 |
| JanC | serious bugfixes yes (security bugfixes certainly, serious bugfixes if somebody takes the time?) | 23:23 |
| mgw | JanC: thanks | 23:25 |
| === Lcawte is now known as Lcawte|Away | ||
| JanC | mgw: just thinking about it, only LTS server images get updated ;) | 23:26 |
| mgw | Yeah, i just noticed everything on 11.10 still has Oct 13 datastamp | 23:27 |
| === sixstringsg is now known as sixstringsg|away | ||
| uvirtbot | New bug: #918807 in libvirt (main) "[precise] virbr0 not created during fresh install when dnsmasq is running with --except-interface=virbr0" [Low,Confirmed] https://launchpad.net/bugs/918807 | 23:34 |
| uvirtbot | New bug: #928524 in lxc (main) "lxcbr0 fails to come up when dnsmasq is installed" [Low,Confirmed] https://launchpad.net/bugs/928524 | 23:35 |
| uvirtbot | New bug: #928550 in php5 (main) "PHP Comparison Issues ... 0 equates to 'D'" [Undecided,New] https://launchpad.net/bugs/928550 | 23:35 |
| abrotman | I'm trying to dist-upgrade from 8.04 to 10.04, and for whatever reason, the 8.04 is using EVMS. When the system boots, it stops, and says it cannot find the device for /, and asks me to wait, hit S or M for Skip or Manual. If I hit Skip, i can then mount everything as normal (though I have to remount / as rw). | 23:44 |
| MTecknology | three away nick changes in one screen of text... | 23:44 |
| abrotman | It seems to be a bug in mountall and/or udev, but I'm looking for a way to maybe have udev rescan the devices or something along those lines .. I'm really not sure | 23:44 |
| abrotman | i found a few bugs in launchpad related to this, but they reference an issue with LVM, not EVMS .. and the fixed packages seem to already be in place. | 23:45 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!