[00:59] <phillw> hiyas, dead n00b question, whilst iptables -F will flush, how do stop top them being reloaded when there is an error?
[00:59] <sarnold> what kind of error? simple syntax error or you made a mistake in the rules and now ssh is dead? :)
[01:03] <phillw> sarnold: a simple error LoT made that took down the ZNC server...
[01:04] <phillw> as it reloads each time, where do I find and stop it from loading from that file?
[07:19] <prsoney> Hello im on a linux class and I need to get nmap. but when I do apt-get install nmap i get loads of extra tools that i dont need since im only running server. are there other similar tools i should use when running server only?
[07:20] <smw> prsoney, it doesn't install anything you don't need
[07:21] <prsoney> it asks me to install zenmap, and for zenmap i need x, which is like 80MB with all the dependencies?
[07:23] <smw> it should not require zenmap
[07:23] <smw> sudo apt-get install nmap adds zenmap to the list of deps?
[07:24] <simmel> Your apt uses suggested packages as deps?
[07:24] <smw> I don't even thing zenmap is suggested
[07:24] <prsoney> Need to get 47.0 MB of archives. After this operation, 130 MB of additional disk space will be used.
[07:24] <smw> sudo apt-cache show nmap      gives me no zenmap
[07:25] <smw> prsoney, what is it installing?
[07:25] <prsoney> loads and loads of stuff smw
[07:25] <smw> prsoney, copy and paste :-)
[07:25] <smw> or pastebin...
[07:27] <prsoney> http://pastebin.com/dqfzGYBa
[07:27] <smw> wtf? lol
[07:28] <smw> prsoney, pastebin the output of sudo apt-cache show nmap
[07:28] <smw> prsoney, and what command gave that output you just gave me?
[07:28] <prsoney> apt-get install nmap
[07:29] <smw> heh
[07:29] <smw> I think the issue is that liblua requires alot
[07:29] <smw> nope, nm
[07:29] <smw> liblua requires nothing
[07:30] <smw> so what is the output of sudo apt-cache show nmap ?
[07:38] <excalibr> smw:
[07:39] <excalibr> zenmap has its own package
[07:39] <smw> excalibr, right, but installing nmap should not drag zenmap in as a dep
[07:39] <smw> it would most likely be the other way around
[07:42] <excalibr> i dont seen zenmap in nmap dep
[07:43] <smw> excalibr, it isn't
[09:20] <doko> zul, Daviey: looking at component mismatches: ok to demote maas-dhcp and maas-dns? if not, please seed it somewhere
[09:20] <doko> ok to promote the other binaries?
[09:22] <doko> roaksoax, Daviey: any update on freeipmi?
[09:22] <bigjools> Daviey (or anyone), why would the BMC on my HP microserver not accept the IP address offered to it and keep doing more DHCPDISCOVER?  It used to work .... :/
[09:31] <linocisco> I followed that https://help.ubuntu.com/12.04/serverguide/samba-dc.html. but I dont know what to write inside logon.cmd
[09:33] <_ruben> linocisco: whatever you want to executed when a user logs in, it's optional
[09:46] <simmel> Unsure if this is the correct channel for this question, but: I'm trying to generate apparmor profiles using aa-complain and aa-logprof but my profiles are almost completely empty, no rules are generated at all. http://pastie.org/private/uoulvjifwy8dkvpgiwjwsw It probably has something
[09:47] <simmel> to do with that I've changed the rsyslog log format, but I can't find what it's supposed to look like. Ideas?
[09:51] <chambs> Can anyone tell me the name of the script which modifies /etc/motd on server 12.04?
[09:52] <frojnd> Hi there. I have ubuntu server 12.4 and I'm seeking any tutorial for setting up git server, I prefer any "official" tutorials. I've already checked help page of ubuntu but it's for later releases: https://help.ubuntu.com/community/Git
[09:54] <linocisco> _ruben, I want to know the format
[09:55] <_ruben> linocisco: it's just a batch file, which gets executed through cmd.exe
[09:55] <_ruben> unless you named it logon.ps1, then it'll get executed through powershell
[10:02] <Daviey> bigjools: hmm, are you sure it was offered ?
[10:03] <bigjools> Daviey: positive, I am looking at the syslog
[10:03] <Daviey> bigjools: Can't answer then.. i saw something similar, where it wasn't getting addresses.. turned out, i had exhausted my lease pool
[10:03] <bigjools> no other dhcp servers around either, so either there's a bug in quantal's isc-dhcpd or the bmc is doing something peculiar
[10:04] <Daviey> the bmc is peculiar
[10:04] <Daviey> :)
[10:04] <bigjools> definitely not exhuasted my lease pool :)
[10:04] <Daviey> I don't know, sorry :(
[10:28]  * koolhead17 pokes Daviey 
[10:31] <doko> zul, Daviey: looking at component mismatches: ok to demote maas-dhcp and maas-dns? if not, please seed it somewhere
[10:31] <doko> ok to promote the other binaries?
[10:34] <lepirat> Hello everyone
[10:34] <lepirat> I have a question if you don't mind, I'm new in Ubuntu world, and I'm trying really hard to get along with it, Ubuntu ROCKS
[10:34] <lepirat> :)
[10:35] <lepirat> my question is :
[10:36] <lepirat> How can I install xubuntu with LVMs on top of RAID 1
[10:36] <lepirat> ?
[10:36] <lepirat> best regards
[10:37] <xnox> lepirat: use alternate cd or mini.iso & follow server-guide advanced installation.
[10:40] <melmoth> lepirat, if the raid 1 is hardware raid, "as usual", it will be seen as a normal drive by the kernel.
[10:40] <linocisco> _ruben, can you show me sample?
[10:40] <melmoth> if it s software raid, i do not know
[10:42] <lepirat> thanks guys for replay, but here comes the output I have seen
[10:42] <lepirat> it shows me a three different HDDs
[10:43] <lepirat> ?
[10:43] <lepirat> it's /dev/sda
[10:43] <lepirat> 2nd /dev/sdb
[10:43] <lepirat> 3rd it's : /dev/mapper/isw_cffeiahdid_volume0
[10:43] <lepirat> sda I suppose is my HDD no.1
[10:44] <lepirat> sdb my HDD no.2
[10:44] <lepirat> what is  /dev/mapper/isw_cffeiahdid_volume0 ?
[10:44] <lepirat> is this the RAID 1 partition
[10:44] <lepirat> Lets say , YES it is !!
[10:44] <xnox> lepirat: stop.
[10:45] <lepirat> okay, something wrong !
[10:45] <xnox> lepirat: that one is Intel Rapid Storage Raid. Setup your RAID1 partitions in the BIOS and then only setup LVM on top of it. the /dev/sda & /dev/sdb will probably be masked by isw after you do that.
[10:46] <xnox> (or you can use the windows app)
[10:46] <xnox> lepirat: you will then need to activate "dmraid" during installation. This will be called "fakeraid" (although I don't like the term, there is nothing fake about it.... ;-) )
[10:47] <xnox> mdadm does support intel rapid storage, but it is not integrated in ubuntu yet.
[10:48] <lepirat> xnon : keep talking
[11:11] <lepirat> sorry I lost the connection befor
[11:11] <lepirat> e
[11:12] <frojnd> Hi there. I'm having a bit of a problem cloning the admin repo (for gitolite) when I do: $ git clone git@localhost:gitolite-admin it asks me for a git password (I type it in) and then it writes out: fatal: 'gitolite-admin' does not appear to be a git repository fatal: The remote end hung up unexpectedly
[11:12] <frojnd> I've successfully setup gitolite
[11:13] <frojnd> but now at the ste when trying to clone: http://sitaramc.github.com/gitolite/clone.html I get this error. Any help would be much appreciated
[11:13] <frojnd> Also I think it should not prompt for a password. Correct me if I'm wrong?
[11:23] <lepirat> Xnox : you were saying something about fakeRAID, how to apply it, or lets say how to make RAID 1 .. shall I go with Hardware RAID, which I stuck on that. Is there any alternative solution ?
[11:23] <lepirat> thanks in advance
[11:31] <soren> davi: I have a problem with "nova-manage db sync" in Folsom that I'm surprised you've not hit.
[11:31] <soren> davi: Whoops, not for you.
[11:31] <soren> Daviey: I have a problem with "nova-manage db sync" in Folsom that I'm surprised you've not hit.
[11:31] <soren> Daviey: https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/migrate_repo/versions/082_essex.py#L260
[11:31] <Daviey> soren: Oh
[11:31] <soren> Daviey: I get the error mentioned in that comment.
[11:32] <soren> Daviey: Why you don't is a mystery to me.
[11:32] <Daviey> soren: erm, i think we have hit that
[11:32] <Daviey> let me check the bug
[11:32] <soren> Daviey: You SRU'ed a fix for it, apparently, but I don't see the fix in Folsom?
[11:33] <Daviey> soren: ah, that isn't bug 1062474
[11:33] <soren> It's also not bug 123456
[11:33] <soren> ..but why are we talking about bugs that it isn't?
[11:34] <Daviey> bug 993663
[11:34] <soren> Yes, that.
[11:34] <Daviey> soren: I /thought/ it might have been the same migration bug we saw
[11:35] <soren> Why the "¤% am I still affected by this bug, then?
[11:35]  * soren scratches head
[11:35] <Daviey> soren: Looking at the comments here, it seems zul asked appropriate questions https://review.openstack.org/#/c/8356/
[11:35] <Daviey> Dan & Mark seemed pretty sure it was kosher
[11:37] <soren> Since then, fix_dns_domains.py seems to have disappeared.
[11:37] <soren> Gah.
[11:37]  * soren goes digging
[11:39] <soren> Oh, that fix only went to stable/essex?
[11:40] <soren> Ok, found the folsom equivalent.
[11:41] <soren> ..but I don't understand why it works. :(
[11:44]  * soren glances at uvirtbot 
[11:46] <codescience> is there really much difference between ubuntu servers of old compared to newer updated versions?
[11:51] <glance> codescience: yes.
[11:51] <soren> Daviey: Nope, still puzzled why it works for anyone else. The first migration attempts to create dns_domains with a domain field as varchar(512) and then defines it as the primary key. 512 characters in utf-8 encoding is much more thn the max of 767 bytes.
[11:52] <soren> Daviey: I don't see anything attempting to prevent this from being created. I see a later migration try to shrink the domain field to make it work, but how it manages to apply the first migration is mysterious.
[11:52] <Daviey> soren: did you see why the gerrit review from above was seemingly reverted for Folsom?
[11:53] <Daviey> https://review.openstack.org/#/c/7139/
[11:53] <soren> Daviey: No, didn't see that.
[11:53] <soren> Daviey: But I do see a comment in migrate_repo/versions/096_recreate_dns_domains.py claiming that "the old dns_domains table is in the 'latin1' charset".
[11:53] <Daviey> soren: do you have a minimal testcase that would be suitable for a git bisect?
[11:54] <soren> Daviey: ..but I don't see anything attempting to make sure that dns_domains is indeed latin1 in the 082_essex migiration.
[11:54] <codescience> glance, for internal server though, old versions are fine when they're internal only servers, right?
[11:55] <Daviey> soren: Generally, my confidence in encoding for Folsom isn't as strong as essex.
[11:56] <soren> If "CREATE TABLE" doesn't specify a charset, where does the default come from?
[11:56] <soren> brb
[11:59] <Daviey> soren: the distro
[12:01] <Daviey> doko: sorry, i missed your pings
[12:01] <Daviey> doko: I am dealing with that today.. leave it with me. Thanks
[12:01] <doko> ok, thanks
[12:03] <soren> Daviey: Well, sure, but how? Where? How can I determine what it is and how can I change it?
[12:04] <Daviey> soren: grep default-character-set /etc/mysql/
[12:05] <Daviey> soren: client & server option
[12:05] <soren> Daviey: No hits.
[12:06] <soren> Daviey: But I can see that if I "create database foo" and mysqldump it, it says its default charset is latin1. My nova db has default charset utf8.
[12:06] <Daviey> I thought we switched from latin1 -> utf8 about 3 cycles ago
[12:06] <soren> I know where the utf8 comes from.
[12:07] <Daviey> Really a question for SpamapS.
[12:07] <soren> My Puppet magic does that explicitly.
[12:07] <soren> Hmm..
[12:10] <Daviey> I really thought we set default-character-set now :/
[12:10] <soren> I think that particular setting got deprecated?
[12:11] <soren> "The syntax '--default-character-set' is deprecated and will be removed in a future release. Please use --character-set-server instead."
[12:18] <Daviey> soren: bah
[12:19] <SpamapS> soren: re where default charset comes from.. there are no less than *4* variables that can influence that
[12:19] <Daviey> SpamapS: Hmm, can we distro patch some more in?
[12:20] <SpamapS> soren: you have the server charset, the db charset, the connection charset, and the "coin flip" charset
[12:21] <SpamapS> and by coin flip I mean "table and/or column"
[12:22] <SpamapS> soren: so, re your question.. create table w/o charset should come from the db charset.
[12:23] <SpamapS> soren: but if that is left unset.. the server charset
[12:23] <SpamapS> I've been through so many nightmares with mysql charsets... was so happy to see drizzle just said "F it, we have utf-8. Thats it."
[12:24] <Daviey> ++
[12:24] <SpamapS> unfortunately, they (the devs) also said "f it, we want a paycheck" and moved on to other projects
[12:25] <SpamapS> now Drizzle is basically funded by GSoC
[12:35] <Kingsy> anyone in here use rsync? I am looking to get some support for it
[12:38] <SpamapS> Kingsy: half the world uses rsync
[12:38] <SpamapS> Kingsy: whats the question?
[12:39] <Kingsy> SpamapS: a command like this, rsync -av --dry-run --exclude-from '.gitignore' -e ssh chris@site.co.uk:/srv/site.co.uk/ .    the output is not just the files tha thave changed.. it is EVERYTHING.. why?
[12:40] <Kingsy> shouldnt that command only sync through what has changed? in regards to its timestamp and size etc?
[12:40] <Kingsy> or at least thats what I assumed
[12:41] <SpamapS> Kingsy: its telling you all things considered
[12:41] <SpamapS> Kingsy: you'll see the "speedup" number which will tell you all the things skipped
[12:41] <SpamapS> Kingsy: btw, your -e is not necessary
[12:41] <SpamapS> Kingsy: '-e ssh' has been assumed for about 10 years now
[12:42] <Kingsy> ah ok
[12:42] <Kingsy> so, is it possible to only echo the files that it is going to sync on the dry run?
[12:42] <SpamapS> unless you're still on RHEL 4 or Debian 3.0 .. nobody expects rsh to work. :)
[12:42] <Kingsy> rather than what it is considering ?
[12:43] <SpamapS> Kingsy: dunno.. I'm sure thats possible
[12:44] <SpamapS> Kingsy: 'man rsync' says that --dry-run + --itemize-changes does what you want
[12:45] <soren> SpamapS: Cool, thanks for the clarification.
[12:47] <Kingsy> hmm let me see
[12:47] <Kingsy> SpamapS: nope thats even worse.. even more verbose
[12:48] <Kingsy> each file/directory seems to be a given a .d..tp..... key along with it, which I suppose means something.. but if you have 100,000 of those, kinda impossible to go through them all
[12:50] <SpamapS> Kingsy: those things are all the mods that will be made
[12:50] <SpamapS> Kingsy: so you can filter out the ones that won't change anything
[12:51] <Kingsy> how? all i want is a dry-run list of the files that are going to be synced
[12:51] <Kingsy> it seems impossible.. I just get a MASSIVE soup of information.. heheh
[12:52] <Kingsy> I have tried about 20 variations on the command
[12:55] <SpamapS> Kingsy: I've used rsync for such things before, but always programmaticaly
[12:57] <patdk-wk> is there anyway to make a raid0 with like loop? without mdadm?
[12:58] <SpamapS> patdk-wk: lvm can do stripes
[12:58] <patdk-wk> I am attempting this from readonly media
[12:58] <xnox> patdk-wk: and raid0 is more or less just plain lvm.
[12:58] <tasaras> hi, I have Ubuntu Server 10.04.4, the system tomcat6 is 6.0.24. I wanna run the latest version, how can I do that?
[12:58] <patdk-wk> lvm is not plain raid0, it's totally different
[12:58] <SpamapS> tasaras: Ubuntu 12.04 is out.. :)
[12:59] <patdk-wk> attempting to recover a raid10 array
[12:59] <patdk-wk> that wasn't made with mdadm
[12:59] <tasaras> SpamapS, i know but it is not within my power to upgrade to the new LTS
[12:59] <xnox> patdk-wk: what was it made with?
[12:59] <tasaras> so is there a "work-around" ?
[12:59] <patdk-wk> hw raid card
[12:59] <patdk-wk> so mainly just need to unstripe the data
[12:59] <SpamapS> tasaras: if you can't upgrade the OS, you probably can't upgrade the system tomcat either
[13:00] <SpamapS> tasaras: you can always just install a newer one in your home dir
[13:00] <xnox> patdk-wk: so... activate it with dmraid. If you can't there is not much help as the hw controller can put data however it likes. I daubt it will be straight forward to "mount" directly.
[13:00] <xnox> but good luck figuring out offsets and see if they work....
[13:00] <SpamapS> me
[13:00] <patdk-wk> xnox, dmraid?
[13:00] <SpamapS> doh
[13:01] <patdk-wk> dmraid doesn't add any headers?
[13:01] <xnox> patdk-wk: yes, dmraid a software package used to mount / talk to some hardware(ish) raid controllers.
[13:01] <patdk-wk> no, no hardware raid here
[13:01] <patdk-wk> the array was built with a raid card in another system, that system is toast
[13:01] <patdk-wk> I just need to unstripe the data
[13:02] <xnox> patdk-wk: if it was not build using linux software raid mdadm... then you really need to get a raid card.... unless you know the ondisk layout that toasted raid card produced.
[13:02] <patdk-wk> xnox, and I do know it :)
[13:02] <patdk-wk> like I said, I only need to unstripe the data
[13:03] <tasaras> SpamapS, so is there not a way to get a new .deb and install over it?
[13:03] <tasaras> or it is a recipe for disaster?
[13:03] <SpamapS> tasaras: you can probably rebuild the precise tomcat6 packages..
[13:04] <SpamapS> tasaras: there's a script called 'backportpackage' that can be helpful there
[13:04] <SpamapS> tasaras: I'm just not sure thats worth the trouble
[13:04] <tasaras> its for supporting a client, not personal use
[13:06] <doko> roaksoax, Daviey: freeipmi promoted. again, please look at component mismatches once this gets updated
[13:07] <Daviey> doko: funny, i did it aswell
[13:07] <Daviey> :)
[13:08] <Daviey> doko: I wasn't lieing when i said i was "on it" :)
[13:09] <doko> Daviey, that was for maas ...
[13:10] <Daviey> doko: Okay.. well we both seem to have promoted freeipmi aswell.
[13:10] <lepirat> Hello there
[13:10] <lepirat> guys I have one question ?
[13:10] <lepirat> How can I install LVM on RAIDed Machine ?
[13:11] <tasaras> where do i get more info about what version of tomcat is contained in Ubuntu 12.04?
[13:13] <melmoth> tasaras, apt-cache search tomcat ? (mention 6 and 7)
[13:14] <tasaras> melmoth, i dont have access to ubuntu 12.04
[13:14] <melmoth> install one, it s free :-)
[13:15] <Pici> tasaras: http://packages.ubuntu.com
[13:16] <SpamapS> tasaras: https://launchpad.net/ubuntu/+source/tomcat6
[13:16] <SpamapS> tasaras: or https://launchpad.net/ubuntu/+source/tomcat7
[13:20] <zul> Daviey: ping can you have a quick peek http://paste.ubuntu.com/1269240/
[13:23] <Daviey> zul: lgtm.. I think that is the approach you discussed with adam_g ?
[13:23] <zul> Daviey:  yeah
[13:24] <Kingsy> SpamapS: so there is no way you know of to just list the files that are going to be synced?
[13:27] <npws_marc> so last night my hard drive had 14GB available.. now it has 0... is there some way i can find out what happened over night?
[13:27] <Kingsy> wow..
[13:28] <Kingsy> df to find out what partition, du to find out what files specifically
[13:28] <Kingsy> I guess
[13:29] <npws_marc> how so?
[13:30] <npws_marc> i know the partition
[13:30] <npws_marc> it's the partition for my /var mount point
[13:30] <npws_marc> how can i use du to find out what happened in the past 12 hours
[13:30] <Kingsy> oh no, I meant use du to find what files have been created that are taking the space
[13:31] <Kingsy> or ls
[13:31] <npws_marc> also: du -hc * on /var claims 66GB, but the partition size is 41GB
[13:31] <Kingsy> then you will be able to tell which process is responsible
[13:32] <npws_marc> oh ithink i know what happened
[13:33] <npws_marc> so i have a hard drive mounted on /var/backup, but i forgot to remount it last night.. so my daily backup script went ahead and backed up to /var/backup anyway
[13:34] <Kingsy> oh.. hehe
[13:48] <SpamapS> Kingsy: yes there's a way. No I don't know it.
[13:48] <zul> Daviey:  can you review cinder in the queue please?
[13:48] <Daviey> ok
[14:38] <jamespage> Daviey, I just uploaded  a fresh merge of bacula from debian including a security fix
[14:38] <jamespage> bug 1064435
[14:39] <Daviey> jamespage: All i read is, "< jamespage> Daviey, i've created some work for you" :)
[14:39] <TheLordOfTime> lol
[14:39] <jamespage> Daviey, absolutely!
[14:39] <jamespage> probably less work that it took to test it tho...
[14:40]  * jamespage must write an automated test suite for bacula
[14:40] <Daviey> jamespage: My review is bassed on, I'm flipping a coin on weather to accept.
[15:11] <zul> adam_g:  lemme know when you are around
[15:15] <Kingsy> SpamapS: you still there?
[15:35] <ahs3> hallyn: sorry 'bout the delay -- uploaded netcf to experimental last night.  wasn't sure if you wanted to leave that bug open or not, tho
[15:39] <SpamapS> Kingsy: stepped out for a bit... back now
[15:40] <hallyn> ahs3: i saw the email late last night - thanks for pushing.  you mean wait to close the bug until it hits unstable?
[15:42] <ahs3> hallyn: i wasn't sure how you wanted to handle it.  it could have been closed with the push to experimental, or we can wait until it's time to push the same bits to sid.  it was kinda hard to tell from the bug.
[15:42] <ahs3> (either way WFM...)
[15:44] <zul> soren: ping the ubuntu-openstack-testing has a patch that solves the nfs testsuite bug fyi
[15:48] <jamespage> zul, whats the story with nova-vncproxy  in folsom?
[15:48] <hallyn> ahs3: i'm not sure either.  how does promotino to unstable happen?
[15:48] <ahs3> hallyn: we'd have to decide to do it, iirc.  then it's just another upload
[15:48] <zul> jamespage: bad documentation in the docs
[15:49] <frojnd> Hm.. I've added a line AllowUsers got myusername, but then when I try to ssh to server I get: ssh: connect to host localhost port 10001: Connection refused. If I remove this line in sshd_config and try to login with the same user I'm allowed. What I be missing here?
[15:49] <jamespage> zul, so which vncproxy package should be used?
[15:49] <zul> jamespage: lemme get back to you on that
[15:49] <jamespage> zul, I'd guess nova-novncproxy
[15:49] <zul> jamespage: yeah
[15:50] <Kingsy> SpamapS: I don't think it matters now thanks
[15:51] <hallyn> ahs3: let's leave the bug open so we remember :)
[15:52] <ahs3> hallyn: okey dokey :)
[15:53] <hallyn> ahs3: i guess i'll put a note down for a week from today to look into promoting
[15:54] <frojnd> Anyone here using gitolite?
[16:03] <frojnd> Anyone willing to help me out with tihs problem? I googled tons of stuff and I still don't know what I'm missing for not successfully cloning at third step: http://sitaramc.github.com/gitolite/clone.html
[16:04] <SpamapS> frojnd: I know marcoceppi in #juju has at least played w/ gitolite
[16:05] <frojnd> SpamapS: thank you I'll try and msg him/her
[16:05] <SpamapS> frojnd: there's a juju charm that he wrote which automates the deployment. Might be worth looking at
[16:06] <frojnd> SpamapS: thank you
[16:07] <TheLordOfTime> hallyn, ping?
[16:08] <hallyn> TheLordOfTime: ubuntu-meeting time
[16:08] <TheLordOfTime> i'll ping you later then :P
[16:08]  * TheLordOfTime returns to beating php5 with a stick
[16:13] <frojnd> One general question, why if I add AllowUsers git defaultuser into sshd_conf I can't login with defaultuser afterward?
[16:14] <sarnold> frojnd: did you put an allowed command on the key that forces starting a git server / process ?
[16:16] <frojnd> sarnold: when I was mading keys: ssh-keygen -t rsa I just pressed enter a couple of times
[16:16] <frojnd> sarnold: but no I don't knwo how to put allowed command on the key o_o
[16:16] <frojnd> sarnold: when I was making keys*
[16:19] <frojnd> sarnold: Was that the question you were asking me?
[16:19] <sarnold> frojnd: ah, my mistake; ForceCommand is used with Match blocks in the sshd_config to force commands...
[16:21] <frojnd> sarnold: I do have ForceCommand line, that I needed to setup chrooted sftp environment
[16:22] <frojnd> I don't understand why I'm not allowed to login after I add user to AllowUsers
[16:22] <sarnold> frojnd: is it properly specific-enough to also still allow your git user to work?
[16:22] <frojnd> sarnold: yes, I setit up for a group sftponly
[16:22] <frojnd> let me paste code to make this clear
[16:23] <Kentos> yo everyone
[16:23] <frojnd> sarnold: http://pastebin.com/j0vmmXEw
[16:25] <Kentos> got a dell 2800, ordered 2 146GB 10k hdd's and the system won't boot, stalls at spinning the drives. I tried finding a way to et the spin speed in the bios but Im having troubles finding anything about it. Also, could potentially be bad drives
[16:25] <Kentos> possible reasons for thi?
[16:25] <sarnold> Kentos: will it boot to USB stick?
[16:26] <Kentos> if i remove the drives it will boot just fine
[16:26] <sarnold> frojnd: okay, that looks fine. hrm. Are there any errors in /var/log/ that look related?
[16:26] <Kentos> but on the boot up, it stalls as it scans and spins the drives
[16:26] <sarnold> Kentos: have you been passed to the OS yet? Or does it stall in the BIOS?
[16:27] <Kentos> its a bios bootup for the server, spins and chacks for the drives and configuration, tries to spinup the channels for the newer drives and never finishes
[16:28] <sarnold> Kentos: can you spin up the drives using USB<->SATA adapters?
[16:28] <Kentos> cant go to OS with these drives, since i cant boot with them in the machine
[16:28] <Kentos> they are scsi drives
[16:29] <Kentos> and i dont have any of those in any case
[16:29] <frojnd> sarnold: no no errors
[16:29] <sarnold> oof, that does complicate things :/
[16:29] <sarnold> frojnd: darn. :/
[16:30] <frojnd> yeah
[16:30] <Kentos> yeah im just wondering what the problem could be, i know that the other drives are 15k speed but the channels for the new drive are not configured
[16:30] <Kentos> like they arent supposed to be other drives for a raid, two new drives for two unused drive spaces
[16:31] <sarnold> Kentos: have you pushed beyond the available power of the powersupply?
[16:31] <Kentos> that is possible, but im not sure what would be doing that
[16:32] <Kentos> not sure if hooking up the redundant would help though, i dont think it would amnyway
[16:41] <TJ-> Kentos: Will it fire up with only one of the 10k drives connected? Are the SCSI terminations correct?
[16:41] <frojnd> One newbie q. How do I check if $HOME/bint exists and is in my $PATH? just with $PATH ?
[16:47] <sarnold> frojnd: can you re-phrase your question?
[16:48] <frojnd> How do I chek on the server that $HOME/bin exists and is in your $PATH
[16:48] <frojnd> check*
[16:48] <TJ-> frojnd: "echo $PATH | grep $HOME/bin"
[16:49] <frojnd> Hm. If it's not how do I add it
[16:51] <sarnold> TJ-: oops, that's vulnerable to shell injections :) here's hoping that is only ever used by trusted programs...
[16:51] <sarnold> frojnd: usual is to just stick a bit like this into your ~/.profile:
[16:51] <sarnold> if [ -d "$HOME/bin" ] ; then
[16:51] <sarnold>     PATH="$HOME/bin:$PATH"
[16:51] <sarnold> fi
[16:52] <frojnd> so everytime user logins it sets bin directory to path
[16:52] <sarnold> frojnd: this makes the assumption that (a) you'll run it early enough so that it will be added before every interactive shell starts (b) it doesn' actually matter if it gets added twice, because it doesn't actually matter. Heh. :)
[16:53] <TJ-> frojnd: Look at the last few lines of ~/.profile that'll show you where it is added and why
[16:53] <frojnd> TJ-: sarnold ok tanx
[17:00] <TheLordOfTime> hallyn, are you still busy?
[17:01] <TheLordOfTime> more breakage!
[17:01]  * TheLordOfTime looks after finding power
[17:15] <TheLordOfTime> !irclogs
[17:15] <TheLordOfTime> oops
[17:15] <TheLordOfTime> that was destined for a query window
[17:15] <TheLordOfTime> working from a laptop sucks :/
[17:17] <TheLordOfTime> uvirtbot is reporting any bugs for server team packages, including private ones, or no?
[17:17] <TheLordOfTime> silence, bot
[17:22] <TheLordOfTime> also, hallyn, ping.
[17:28] <TheLordOfTime> https://bugs.launchpad.net/ubuntu/+source/php5/+bug/191251  <-- Per http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=380731, this was updated in 5.4.1-1 of php5, which is now in quantal as part of 5.4.6-1, probably part of libphp5-embed.  Shouldn't that needs packaging bug be changed to say this was alread implemented?
[17:29] <TheLordOfTime> (old bug, yes, but old ancient bugs that arent closed annoy me :P)
[17:30] <TheLordOfTime> SpamapS, you'd probably know how to handle that one.
[17:31] <SpamapS> TheLordOfTime: yes!
[17:32] <SpamapS> TheLordOfTime: Fix Released
[17:32] <TheLordOfTime> SpamapS, you want to take it, or should I?
[17:32] <TheLordOfTime> (you've got more "say" than I on those matters, you actually work with the repos on occasion)
[17:33] <SpamapS> TheLordOfTime: make sure to copy/paste the changelog
[17:33] <SpamapS> TheLordOfTime: all thats needed is evidence of the release that contains php5-embed
[17:34] <TheLordOfTime> SpamapS, i assume i have to pull from the debian/changelog from the package in Ubuntu :p
[17:34] <TheLordOfTime> ... right...
[17:34] <TheLordOfTime> i'll get on that once they allow Linux on this netowrk
[17:34] <TheLordOfTime> :/
[17:34] <SpamapS> doh
[17:34]  * TheLordOfTime is stuck on Windows atm
[17:34] <SpamapS> TheLordOfTime: the changelog is on launchpad too
[17:34] <TheLordOfTime> SpamapS, the *complete* changelog>?
[17:34] <TheLordOfTime> or the diff'd changelog
[17:35] <TheLordOfTime> (5.4.1 was never in Ubuntu, i think it got synced as 5.4.somethingAfter2
[17:35] <TheLordOfTime> )
[17:36] <TheLordOfTime> ... wait... waiit...
[17:36]  * TheLordOfTime facepalms
[17:38]  * TheLordOfTime elevates that to facedesking
[17:39] <frojnd> I'm still having a difficult with ssh keys. When trying to clone git: git clone git@localhost:gitolite-admin.git it asks me for a password. Even though I've configured authorized_keys in git's .ssh dir. This is the /var/log/auth.log http://paste2.org/p/2316558 I'm starting to belive my sshd is corrupt: http://paste2.org/p/2316568 Any ideas what might be wrong?
[17:40] <TheLordOfTime> note to self: all packages in Ubuntu have bzr branches containing code and files.  use that if you don't have access to php5 source packages.  *facedesk*
[17:44] <TheLordOfTime> SpamapS, do i have  to pull the entire changelog (5.4.1-1 -> 5.4.6-1ubuntu1), or can i just pull the excerpt for the 5.4.1-1 changelog entry?
[17:44] <TheLordOfTime> considering i pulled that entry from the debian/changelog in the ubuntu source stuffs
[17:45] <SpamapS> TheLordOfTime: just the excerpt for 5.4.1-1
[17:45] <hallyn> SpamapS: say, i fired up a ceph instance with juju, but install doesn't complete bc package radosgw isn't found <shrug>
[17:46] <SpamapS> hallyn: new ceph charms incoming :)
[17:46] <hallyn> ah
[17:46] <SpamapS> hallyn: its been quite neglected
[17:46] <hallyn> SpamapS: i also don't see any support under ceph/hooks for the fs mounts...  is that addressed in the new charms, or am i mis-reading and ti's actually done somewhere?
[17:46] <hallyn> i.e. metadata.yaml says there is a fs mount relation, but there is no fs-* under hooks/
[17:47] <TheLordOfTime> SpamapS, http://pastebin.com/Qnndm3EY  <-- looks good as a description?
[17:47] <SpamapS> hallyn: the new charm focuses on the mon service
[17:47] <TheLordOfTime> hallyn, got a spare instance of php5-fpm lying around on a Precise VM?
[17:47] <SpamapS> hallyn: the FS itself is quite a bit more tricky
[17:47] <TheLordOfTime> s/VM/install/
[17:48] <hallyn> TheLordOfTime: no, but i can fire one up.  note that php is REALLY not my bag :)
[17:48] <hallyn> SpamapS: sigh.
[17:48] <TheLordOfTime> hallyn, just need something confirmed/refuted is all, and its just testing the init.d script ;)
[17:48] <hallyn> TheLordOfTime: ok, one min
[17:48] <TheLordOfTime> lemme get the bug number, i forgot what it is offhand
[17:49] <TheLordOfTime> hallyn, https://bugs.launchpad.net/ubuntu/+source/php5/+bug/1059272  <-- that's what i'd like confirmed/refuted
[17:49] <TheLordOfTime> if i remember, back in Natty, it used to say "Starting PHP5-FPM...   [OK]" or "Stopping PHP5-FPM....  [OK]" or similar
[17:50] <TheLordOfTime> doesn't seem to do that in Precise...
[17:50] <TheLordOfTime> that's probably a wishlist bug, asking for a minor change in the init.d script
[17:50] <TheLordOfTime> it just is weird as a server admin when you issue "stop' and you've got no statement that its been stopped
[17:51] <hallyn> TheLordOfTime: yeah, no output at all
[17:51] <TheLordOfTime> i assume 'restart' gives output though?
[17:51] <TheLordOfTime> 'start' 'stop' give none, 'restart' gives output
[17:51] <TheLordOfTime> on my tests at least
[17:53] <hallyn> TheLordOfTime: the checks are inverted i think:
[17:53] <hallyn> i.e.     [ "$VERBOSE" != no ] && log_daemon_msg "Starting $DESC" "$NAME"
[17:54] <hallyn> this is the problem with using double negatives.  the human brain does poorly with "if x != no" :)
[17:54] <TheLordOfTime> lol
[17:54] <TheLordOfTime> hallyn, i'd love to see a "CONFIRM" on that bug, as well as your musings attached
[17:54] <hallyn> TheLordOfTime: just inverting those to '= no' works
[17:54] <TheLordOfTime> otherwise i'll lose logs :P
[17:54] <hallyn> will comment
[17:55] <TheLordOfTime> see? wasn't that an easy "confirm"?  :P  init.d scripts not returning output is easy to confirm/refute :P
[17:55] <TheLordOfTime> ... i abuse the :P emoticon, i think...
[17:58] <TheLordOfTime> SpamapS, one old bug down, thousands more to go xD
[17:58] <SpamapS> TheLordOfTime: thank you for working on that! :)
[17:59] <TheLordOfTime> SpamapS, meh, was bored...
[17:59] <TheLordOfTime> when you've finished your exams for the day, and you have nothing else to do....
[17:59] <TheLordOfTime> what better way to spend time than to sift through ancient bugs and close the ones that must be closed
[17:59] <TheLordOfTime> :P
[17:59] <SpamapS> TheLordOfTime: only 87,000 more to go
[17:59] <TheLordOfTime> SpamapS, wait, there's 87k open bugs in php5 o.O
[17:59] <TheLordOfTime> or 87,000 open bugs in ubuntu :P
[18:01] <SpamapS> in ubuntu
[18:01] <TheLordOfTime> ... oh dear GOD, this one's old...
[18:01] <TheLordOfTime> https://bugs.launchpad.net/ubuntu/+source/php5/+bug/183933
[18:01] <SpamapS> roughly
[18:01] <TheLordOfTime> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=492442  <-- apparently part of 5.3.x?
[18:02] <SpamapS> 5.3.0-1 yeah
[18:02] <TheLordOfTime> (that's the relevant upstream bug)
[18:02] <SpamapS> TheLordOfTime: same result
[18:02] <TheLordOfTime> ... why wasnt this autohandled... *grumbles something abotu incompetence*
[18:02] <SpamapS> TheLordOfTime: Fix Released
[18:02] <TheLordOfTime> will do, have to print out an updated syllabus for a class first
[18:02] <TheLordOfTime> apparently, the instructor is lazy and compiled three weeks of changes into one document
[18:03] <SpamapS> TheLordOfTime: because php is merged. We're supposed to go through merged bugs and attach LP # headers in the merge process.. but.. that is not easy
[18:03] <TheLordOfTime> indeed.  :P
[18:03] <TheLordOfTime> oh, and don't forget about the SRU'd fixes that i occasionally get included... ;P
[18:06] <soren> zul: Thanks.
[18:08] <TheLordOfTime> SpamapS, i don't see a changelog entry for that...
[18:08] <TheLordOfTime> SpamapS, for fileinfo
[18:08] <SpamapS> TheLordOfTime: it was just part of 5.3 upstream
[18:09] <TheLordOfTime> as part of core?
[18:09] <SpamapS> TheLordOfTime: so yeah, you can just reference the Debian bug report
[18:09] <TheLordOfTime> so for that one, rather than pulling changelogs, i just quote the debian bug report, stating that upstream included it in 5.3.x as part of the package by default, and not a separate package?
[18:10] <SpamapS> TheLordOfTime: right
[18:10] <SpamapS> TheLordOfTime: we just want people who stumble upon the report to see when it was fixed
[18:10] <SpamapS> which is to say... a long long time ago :0
[18:12] <TheLordOfTime> :P
[18:12] <TheLordOfTime> indeed.
[18:13] <TheLordOfTime> cool, two bugs down, 86,999 more to go!  :P
[18:32] <frojnd> Is this bug if I change /etc/motd and then try to login and it's the old motd?
[18:34] <sarnold> frojnd: man 5 update-motd
[18:35] <frojnd> sarnold: thank you! That's exactly what I needed
[18:36] <sarnold> :)
[18:36] <frojnd> What does 5 mean?
[18:37] <frojnd> man man? :P
[19:25] <f00bar80> how to copy/mirror a ubunut lucid install to a new machine ?
[19:48] <TJ-> f00bar80: To mirror it; use 'dd' to clone the disks. As a follow-up you may want to change the UUID of the file-systems on the clone so they are different to the original, in case you want to mount both hard-drives in the same system at some point
[19:51] <ScottK> SpamapS: Go on juju
[19:51] <hallyn> utlemming: ubuntu-cloudimg-query on precise doesn't know about quantal.  Is there a ppa i can nab to make it do?
[19:52] <hallyn> well hm, here it does.  not on ec2 though
[19:52] <SpamapS> ScottK: thank you!
[19:53] <SpamapS> hmm
[19:53] <hallyn> smoser: ^  (ubuntu-cloudimg-query quantal on a precise ec2 host seems to not know about quantal?)
[19:53] <SpamapS> utlemming: we need to get ubuntu-cloudimg-query into Debian so lxc can spin up ubuntu-cloud based containers
[19:54] <SpamapS> smoser: ^^
[19:54] <hallyn> SpamapS: what does getting into debian do to help?
[19:55] <smoser> apt-get install distro-info
[19:55] <smoser> hallyn, SpamapS ^
[19:56]  * hallyn tries
[19:56] <hallyn> smoser: thanks!
[19:57] <smoser> it is a 'recommends'
[19:59] <smoser> hallyn, but i just now added it to cloud-utils trunk too in the hard coded list.
[19:59] <smoser> just in time for 'r'
[19:59] <smoser> :
[19:59] <SpamapS> hallyn: getting it into debian helps debian users use juju local provider
[20:03] <hallyn> SpamapS: what are you goin gto do there for networking?  (they don't get lxcbr0)
[20:04] <SpamapS> hallyn: pray? ;)
[20:04]  * hallyn is trying to charm openvswitch-linked lxc compute nodes
[20:08] <smoser> utlemming, https://code.launchpad.net/~smoser/maas/cleaner-shutdown/+merge/128807
[20:08] <smoser> might be of interest to you.
[20:08] <smoser> the cloudimg builders could do that.
[20:08] <roniez> anybody know what channel i can go for wubi problems?
[20:10] <zul> hallyn: ping
[20:16] <hallyn> zul: .
[20:17] <zul> hallyn: is there a way to determine the child process through /proc/<pid>
[20:18] <hallyn> zul: what is "the child process"?  you mean task 1000 just did a fork, and you want to find out the new pid through /proc/1000?  If so, no
[20:18] <hallyn> you can look through all tasks to see which have 1000 as parent...
[20:18] <zul> k thats what i thought
[20:18] <zul> ill have to do it the hardway then
[20:18] <RoyK> hallyn: ps axf
[20:19] <RoyK> f shows child processes
[20:19] <hallyn> RoyK: by tracing parent processes :)
[20:19] <hallyn> s/tracing/following
[20:20] <hallyn> zul: now if it's a thread, you can see it in /proc/$$/task/ fwiw.  but i'm sure that's not helpful
[20:20] <zul> hallyn: well its helpful for my education
[20:23] <f00bar80> TJ-, what tools i need , as do i need another HD, to clone the disks to or what, i mean how and what to use, you already mentioned dd and UUID , but what do i need as hardware ??
[20:25] <Daviey> roaksoax: Hey, so should i finally NEW review openstack-resource-agents ( ?
[20:26] <roaksoax> Daviey: TBH i haven't tested the ra's but are simply shell scripts to be run by pacemaker
[20:26] <roaksoax> and very well tested upstream
[20:26] <roaksoax> and with openstackk
[20:26] <roaksoax> so we shouldn't have any problmes with them
[20:26] <roaksoax> as the deabin maintainer is the actual packager
[20:26] <roaksoax> and uipstream
[20:26] <TJ-> f00bar80: I'd hook the 'target' drive up to the 'source' PC, boot it using a USB or CD (e.g. liveISO image) and simply use the form "sudo dd if=/dev/sdX of=/dev/sdY" to clone sdX to sdY. It is important the drive being read from isn't also hosting the running OS, since you don't want the file-systems to change whilst the cloning is happening
[20:27] <Daviey> roaksoax: Hmm, has it actually been uploaded to debian stable?
[20:27] <Daviey> err unstable/sid
[20:29] <roaksoax> Daviey: no and I don't know what's the hold up
[20:29] <Daviey> roaksoax: ok, thanks
[20:31] <hallyn> SpamapS: so if i fire up 3 ceph instances with your new juju charm (when available), is the idea that i log into the ceph nodes and mount it by hand there, or is there some client i install on a different ec2 node, to connect to the ceph cluster and provide me with something i can mount as an fs?
[20:31] <hallyn> or should i wait until you blog about it with a clear example :)
[20:32] <TheLordOfTime> would you guys say that upstreaming LP Bug 1059272 to Debian is a good idea?
[20:35] <hallyn> TheLordOfTime: assuming their script is the same, yeah
[20:35] <TheLordOfTime> hallyn, want to run a diff on our debian/php5-fpm.init against theirs?
[20:35] <TheLordOfTime> :P
[20:35]  * TheLordOfTime can't from Windows
[20:36] <f00bar80> TJ-, sorry also i didn't get what i mean by or what i have to change the UUID of the file-systems?
[20:38] <TJ-> f00bar80: Once the cloning is done, if you boot the system the file-systems will be identical. Each has a Universally Unique IDentifier (UUID). So, if you want to run the clone (as opposed to simply keeping it as a back-up spare) it makes sense to alter the UUID if each file-system on the clone, so if you connect that drive to the original system in future, the file-system IDs don't confuse the operating system
[20:38] <hallyn> utlemming: lxc-create -t ubuntu-cloud -n q1 -- -r quantal -a i386 from a precise amd64 host is not giving me a 'ubuntu' user.  i thought all ubuntu-cloud images had an ubuntu user?
[20:39] <utlemming> hallyn: that'd be a bug. There is no default user for the Quantal. Cloud-init creates it on first boot.
[20:40] <hallyn> utlemming: ah, that explains it then, thx
[20:41] <f00bar80> TJ-, got you, what if i'll put it on a new machine "drivers wise" ?
[20:41] <hallyn> TheLordOfTime: network is slow here, but at least i can confirm all of debian is at a newer version than our precise
[20:41] <TheLordOfTime> so... SRU fix for precise separately from upstream?
[20:42] <hallyn> but no, quantal still has the same init script
[20:42] <TheLordOfTime> hallyn, is our quantal significantly older than Debian?
[20:42] <TheLordOfTime> !info php5 sid
[20:43] <TheLordOfTime> !info php5 unstable
[20:43] <TheLordOfTime> ....
[20:43] <TJ-> f00bar80: kernel modules are all there waiting. The one thing to be aware of is that udev hard-codes the interface name (e.g. eth0) to a MAC address so you'll need to edit /etc/udev/rules.d/70-persistent-net.rules to remove the interface-name assignments, if you want the 'new' system to also have an eth0 (rather than picking a higher number)
[20:43] <TheLordOfTime> !info php5-fpm testing
[20:43] <hallyn> TheLordOfTime: no debian sid has the same debian/php5-fpm.init as quantal and precise (or at least both have the 'VERBOSE != no &&' bug)
[20:43] <TheLordOfTime> okay, then that'd need upstreaming to debian.
[20:43] <TheLordOfTime> anyone want to do that faster than me?
[20:43]  * TheLordOfTime would have to use the email system to file the bug, and that could take some time
[20:44] <hallyn> theres another way? :)
[20:44] <TheLordOfTime> apparently 'reportbug' within debian
[20:44] <hallyn> i can do it tonight or tomorrow, but not right now.  (woudl just use submitttodebian)
[20:44] <TheLordOfTime> but i'm not IN debian
[20:46] <hallyn> TheLordOfTime: write it down to do later, thanks.  ttyl
[20:46] <TheLordOfTime> might preempt you, trying to file via email now :P:
[20:46] <f00bar80> TJ-, sounds good , thx cheers
[20:47] <f00bar80> TJ-, sorry a very strange question, is there a way from command line to know if the motherboard fan is working or not :) ?
[20:49] <f00bar80> TJ- or even the system temperature and get mail notification if any overheating issues
[20:52] <sarnold> f00bar80: install lm-sensors package; 'sensors' command shows a good deal of data
[20:53] <f00bar80> sarnold, can send notification email messages if over-heating issues ?
[20:54] <sarnold> f00bar80: I don't know if the lm-sensors package itself has that ability; maybe nagios or similar tool would do that. but they probably rely on lm-sensors being installed and configured
[20:57] <Daviey> SpamapS: Plans to upload juju?
[20:58] <TheLordOfTime> hallyn, http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=690091  <-- w00t, it worked!
[20:58] <TheLordOfTime> hallyn, i filed the bug :
[20:58] <TheLordOfTime> :P *
[20:58] <TJ-> f00bar80: There is; see the package "lm-sensors"
[21:04] <TheLordOfTime> hallyn, can you email 690091@bugs.debian.org and attach to that the description / possiblesolution you posed on LP?
[21:11] <SpamapS> Daviey: yes, running tests and builds now, will upload in the next hour
[21:15] <Daviey> SpamapS: Rocking.  That includes sought after MAAS fixes :)
[21:19] <TheLordOfTime> https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/997978  <-- thought this had an SRU fix committed?
[21:19] <hallyn> TheLordOfTime: no wait, my mind was triple-inverting or something.  VERBOSe is no by default in /lib/init/vars.sh
[21:20] <TheLordOfTime> hallyn, i upstreamed anyways?  its still an issue that'll occur in both places
[21:20] <hallyn> TheLordOfTime: yeah i'm just saying that the test isn't technically wrong.  rather i don't think that's what we want to check.
[21:21] <TheLordOfTime> hallyn, any idea what the turnaround time'll be on LP Bug 997978 for precise, for when it'll actually get pushed into the repos?
[21:21] <TheLordOfTime> (its affecting a critical service)
[21:22] <hallyn> TheLordOfTime: no, SpamapS may know.  it's been verified, the fix is in precise-proposed.  when it gets promoted to precise-updates, i don't know
[21:23]  * TheLordOfTime repeatedly pokes SpamapS
[21:23] <TheLordOfTime> xD
[21:23] <TheLordOfTime> lol, i kid
[21:23] <hallyn> TheLordOfTime: try '/etc/init.d/lprng stop', as an example.  it's also quiet
[21:23] <hallyn> TheLordOfTime: so, i personally disagree with VERBOSE by default, bit it seems to be the norm ...  <shrug>
[21:24] <TheLordOfTime> perhaps for the php5-fpm init, it can actually be coded so it always posts it
[21:24] <hallyn> TheLordOfTime: it certainly can, but is ignoring the system default the right thing?
[21:24] <hallyn> anyway letting the debian maintainers decide is the right thing
[21:25] <TheLordOfTime> indeed
[21:25] <TheLordOfTime> that's why i upstreamed ;P
[21:25] <TheLordOfTime> besides, the php package maintainers are up there, so if they change it, good.  if not, i'll just fork the package :P
[21:25] <hallyn> yup - thanks
[21:26] <hallyn> utlemming: ok, so yes, even on my local precise system, 'lxc-create -t ubuntu-cloud -n q1 -- -r quantal' fails bc of user ubuntu not existing
[21:27] <utlemming> hallyn: checking...
[21:27] <hallyn> utlemming: so should the usermod for user ubuntu not be done when not -C ?
[21:27] <hallyn> utlemming: but it does work for precise and lucid guests :)
[21:27] <hallyn> so either way the cloud images are inconsistent.
[21:28] <utlemming> hallyn: so the problem is likely that the script....there is no ubuntu user pre-baked in Quantal.
[21:28] <utlemming> hallyn: meaning that LXC cloud script is busted
[21:28] <hallyn> utlemming: you'll likely deny it, but i always claim you own ubuntu-cloud lxc template :)
[21:28] <hallyn> utlemming: woudl it be deemed bad to have the lxc template always create the ubuntu user?
[21:29] <utlemming> hallyn: actually, the fix should be really easy
[21:29] <utlemming> hallyn: I'll take a look right now
[21:30] <hallyn> utlemming: awesome, thanks.  lemme know if you want anything from me, else i'll leave it in your hands.
[21:30] <utlemming> hallyn: is there a bug number?
[21:30] <hallyn> i'm heading outside for a bit.  bbl - hoping to do some serious charming tonight.
[21:30] <hallyn> utlemming: not yet.
[21:30] <hallyn> wnat me to open one?
[21:32]  * hallyn files
[21:32] <utlemming> hallyn: oh, sure...assign it to me
[21:33] <hallyn> utlemming: bug 1064702
[21:33] <hallyn> thanks - ttyl
[21:57] <SpamapS> Daviey: so, I uploaded 0.6-1 to Debian experimental. Better to wait 1 day and sync or make a 0.6-0ubuntu1 ?
[22:05] <Daviey> SpamapS: I'll sync it in the morning.. sound ok?
[22:05] <virusuy> hi all
[22:06] <SpamapS> Daviey: sounds good. Beware, it may FTBFS a few times.. I'm finding that its about 33% fail rate (will investigate soon.. probably racey tests). Seems to be more reliable on faster boxes, so it might just be the PPA builders being slow.
[22:08] <Daviey> SpamapS: yeah, i found the same.. infact, your last upload i had to give back aswell :)
[22:08] <Daviey> (real archive)
[22:10] <SpamapS> Daviey: well, at least it doesn't have to run on the armel builders :)
[22:13] <Daviey> SpamapS: powerpc!
[23:05] <DarkStar1> hi. I'm having problems getting vsftpd to start syslog showe the following message: kernel: init: vsftpd respawning too fast, stopped
[23:05] <DarkStar1> nothing in vsftpd.log
[23:05] <hallyn> utlemming: lucid ubuntu-cloud containers are created fine, but don't start up.  haven't yet figured out why
[23:06] <utlemming> hallyn: I can't replicate the issue myself. But this might be a cloud-init bug.
[23:06] <utlemming> hallyn: do you have logs?
[23:06] <hallyn> oh, lxcguest is not installed i think
[23:07] <hallyn> utlemming: which issue is that?
[23:07] <sarnold> DarkStar1: anything in dmesg?
[23:07] <utlemming> hallyn: the usermod lxc quantal one
[23:07] <DarkStar1> sarnold: dmesg?
[23:07] <sarnold> DarkStar1: the kernel message buffer
[23:07] <DarkStar1> sec
[23:07] <hallyn> utlemming: you've done it in a fresh precise instance?
[23:08] <sarnold> DarkStar1: dmesg | tail ought to be sufficient
[23:08] <utlemming> hallyn: oh...no, I was using quantal on quantal
[23:08] <hallyn> utlemming: right, this is quantal on precise.  d'oh, did i not mention that in the bug?
[23:09] <DarkStar1> sarnold:  nothing there except for mdadm: sending ioctl 1261 to a partition!
[23:09] <sarnold> DarkStar1: seems a bit odd but probably unrelated. hrm.
[23:09] <hallyn> utlemming: ok, yeah, lucid cloud guests don't automatically install lxcguest (which is needed since lucid is not container-safe like precise is)
[23:10] <DarkStar1> sarnold: sorry I get the same message
[23:10] <DarkStar1> init: vsftpd main process (726) terminated with status 1
[23:10] <DarkStar1> init: vsftpd respawning too fast, stopped
[23:10] <utlemming> hallyn: at this point the game, I think disallowing Lucid cloud images is the right thing to do then
[23:10] <utlemming> unless the template install it during installation
[23:10] <hallyn> anyway i hit these bc creating a pristine lvm-backed continer for each release was part of my charm, but i guess i'll go to the ubuntu template, bc the cloud one just isn't quite as ideal for what i'm doing
[23:10] <DarkStar1> sarnold: didn't pipe it through tail
[23:10] <sarnold> DarkStar1: that's fine, tail just shows you only the last ten lines.
[23:11] <hallyn> utlemming: we should probably as the juju folks which they prefer.  ti's a two-step process to install lxcguest, so i don't know
[23:11] <DarkStar1> sarnold: I got this after enabling ssl in the conf
[23:11] <hallyn> if we're better off refusing to create them, or creating them and warning that it's not usable without lxcguest
[23:11] <sarnold> DarkStar1: if you back out your change does it work alright?
[23:11] <hallyn> utlemming: i'll leave you alone now and use lxc-ubuntu :)  have a good night
[23:11] <sarnold> DarkStar1: (I would hope that it would log config file errors before dying, but perhaps that's configured after the ssl...?)
[23:13] <DarkStar1> sarnold: I figured maybe it couldn't read the cert files in the /etc/ss/private and certs directory but I have since added the ftp user to the ssl-cert group which has read access to the files
[23:13] <DarkStar1> sarnold: I'll check on the logging but enabling ssl is the last thing in the conf file
[23:14] <sarnold> DarkStar1: hrm, are you sure you want your ftp server to have read access to /etc/ssl/private ? that seems like a high-value target to expose over .. ftp.
[23:14] <DarkStar1> sarnold: nope but I only did it for temp reasons
[23:14] <DarkStar1> just to see
[23:15] <sarnold> okay
[23:15] <DarkStar1> I'll change it back in a sec :)
[23:18] <DarkStar1> argh……
[23:20] <DarkStar1> sarnold: It seems it doesn't like the ssl_enable option
[23:20] <sarnold> DarkStar1: hrm :/
[23:22] <DarkStar1> I don't know why without no useful logging information
[23:23] <sarnold> DarkStar1: well, there's always strace. :/
[23:23] <sarnold> DarkStar1: strace -o /tmp/vsftpd -f /etc/init.d/vsftpd start
[23:23] <sarnold> then read /tmp/vsftpd and find out why it died...
[23:24] <DarkStar1> sarnold: thx. Never used strace before :)
[23:24] <sarnold> DarkStar1: it'll be like drinking from the firehose. awesome tool but it takes a bit of experience to get the hang of reading its output.
[23:24] <sarnold> DarkStar1: protip, head straight to the end of the file and look upwards. :)
[23:28] <DarkStar1> sarnold: Not cool man :) I'll have better luck reading Japanese
[23:29] <sarnold> ohayou gozaimasu! ogenki desu ka? :)
[23:29] <sarnold> DarkStar1: darn. Like I said, powerful tool, but it _is_ a handful. :(
[23:29] <DarkStar1> lol
[23:30] <DarkStar1> sarnold san. Wax on wax off
[23:32] <DarkStar1> sarnold: There's actually nothing useful in here. I could pastebin it for you to verify
[23:32] <SpamapS> DarkStar1: there's always something useful. Its just finding it inside all that useless crap
[23:32] <sarnold> DarkStar1: be careful there's no private data in there first. it scoops up everything in all system calls made along the way.
[23:33] <DarkStar1> I gathered. There is no private data in there far as I can see
[23:34] <blendedbychris> hey guys i have two servers and the times are out of sync by a few minutes
[23:34] <blendedbychris> i installed ntp
[23:34] <nsahoo> I recently installed ubuntu server 12.04 on a machine with both wired and wireless networking options. While installing I think it preferred to use the wired option. If I disconnect the ethernet cable and restart will it automatically use the wireless connection?
[23:35] <SpamapS> nsahoo: no, servers require some different configuration for wireless
[23:35] <SpamapS> nsahoo: wish I could help further but I'm sure google has info on it :)
[23:35]  * SpamapS signs off
[23:36] <nsahoo> SpamapS: thanks, can I do something now that'll allow it to catch the wireless connection when that is the only one available?
[23:36] <DarkStar1> sarnold: dbl checked but just in case I /msg'ed it to you
[23:36] <DarkStar1> brb in 30 secs
[23:48] <DarkStar1> sarnold: should I file it with ubuntu or with sftpd? because the latter might say "consult your distro"
[23:48] <sarnold> DarkStar1: I say distro..
[23:48] <DarkStar1> Ok