[00:20] <MFSOT> Hi all, intermediate linux user here, have my home boxes running linux mint and just switched from win2k8 to ubuntu server with a webmin gui, I have an external hdd on my server that I would like to be able to access/share/write to and it's proving to be difficult via the webmin, I've read I should SSH into the server, this that and the other thing, just looking for opinions and some help
[00:27] <ScottK> http://www.catb.org/esr/faqs/smart-questions.html is not a bad place to start.
[00:27] <MFSOT> thanks bud
[00:28] <MFSOT> took a lot of effort to be a dick
[00:31] <qman__> MFSOT, firstly, that was a perfectly valid response, and secondly
[00:31] <qman__> !webmin
[00:32] <qman__> all of that said, you need to get a bit more specific in what it is you want to accomplish
[00:32] <histo> Why would outlook completely fail at sending mail from my server when other clients work fine
[00:32] <histo> ughhh
[00:33] <histo> I'd assume it has to do with starttls or something goofy. Anyone experience this?
[00:33] <histo> Running postfix/dovecot setup with virtual users controleld by text files
[00:33] <qman__> because microsoft follows standards about as well as square pegs fit in round holes
[00:33] <qman__> outlook, and only outlook, requires a non-standard auth line to work on your server
[00:34] <histo> qman__: what is the non-standard line and where in the config do I put it?
[00:34] <qman__> everybody else uses "AUTH PLAIN", but outlook requires "AUTH=PLAIN" to be accepted
[00:34] <qman__> there's a bit in the default config file regarding this
[00:35] <qman__> you need to make sure it's enabled
[00:35] <histo> qman__: in postfix's config?
[00:36] <qman__> yes
[00:36] <qman__> this may be helpful: http://holdenweb.blogspot.com/2008/04/outlook-is-driving-me-nuts.html
[00:36] <histo> qman__: I don't see it anywhere
[00:36] <histo> k will read
[00:37] <histo> i've added both those changes
[00:37] <histo> already
[00:37] <histo> qman__: With no luck
[00:37] <MFSOT> qman__ I appreciate your response as at least you asked for something and didn't point me to a useless page to try to make me feel stupid, there's too much of that shit on irc - I know webmin isn't supported any longer but I've read that it works just fine and to be honest it's a lot cleaner and easier to use than the other gui's I've been messing around with. I ultimately want to be able to share files on my home network and pipe
[00:37] <MFSOT> in via my vpn that I have set up and browse and use my home file system, I also want to either have an FTP site or SSL site, a way to share larger files with friends basically
[00:38] <qman__> "works just fine" is anecdotal, I've seen it break plenty of stuff, and if you come here with broken stuff and webmin, we really can't help
[00:38] <bobbyz> histo: broken_sasl_auth_clients = yes
[00:38] <qman__> but if you want some basic instructions, try the server guide
[00:39] <MFSOT> I'm not opposed to giving webmin the axe, this is a setup I just started today, I was looking at using ISPConfig, or just putting an gui on until I got things where I wanted them
[00:39] <histo> bobbyz: yes I have that set
[00:39] <ScottK> MFSOT: If you think that page is useless, go back and reconsider.  No one here is under any obligation to help you at all.  Personally, I'm done with the prospect.
[00:39] <histo> bobbyz: also have plan login cram-md5 in my dovecot.conf as well
[00:40] <histo> I noticed though when I telnet to my server port 25 it doesnt' display auth anything
[00:40] <MFSOT> well you haven't helped in anyway, considering you never started, glad you're done
[00:40] <bobbyz> histo: this might be a dumb question, but you issued a 'postfix reload' after you set that in /etc/postfix/main.cf right?
[00:40] <histo> bobbyz: It's been set that way the entire time i've been trying let me restart just to be sure
[00:40] <ScottK> Also look in /var/log/mail.log for information about what's going on.  Both postfix and dovecot log there.
[00:42] <histo> bobbyz: qman__ still getting relay access denied from windows live clients
[00:43] <histo> ScottK: yeah i'm tailing the log but it just rejects it and says relay access denied
[00:43] <qman__> histo, are the clients attempting to log in? you have to specifically tell them to do SMTP auth, they try anonymous by default
[00:43] <histo> Thunderbird works perfectly with my setup
[00:43] <qman__> even if you have pop3/imap configured
[00:43] <histo> qman__: yes
[00:44] <histo> I'm testing on a windows 7 box with windows live mail. And it will not send friggen mail.
[00:44] <histo> Just 554 relay denied
[00:44] <histo> So it has to be something witht he AUTH
[00:44] <ScottK> Dunno if it's still the case, but for a long time Microsoft clients didn't support TLS properly and you had to use SMTPS on port 465.
[00:44] <qman__> also, with many typical virtual user configs, the username is "user+domain.com"
[00:44] <ScottK> There's a commented out entry in master.cf for that service.
[00:44] <qman__> rather than "user" or "user@domain.com"
[00:45] <qman__> but that should show against any client
[00:45] <ScottK> IIRC the server guide has a decent description of setting up dovecot for SASL auth with postfix.
[00:45] <histo> ScottK: I've been following the server guide
[00:46] <ScottK> OK.
[00:46] <ScottK> If it works for non-MS clients, I'd try SMTPS on port 465 then.
[00:46] <histo> qman__: I know as I can configure thunderbird to login and send mail just fine. Windows live mail is the problem
[00:46] <qman__> just thinking aloud, my go-to ideas have been exhausted
[00:48] <histo> ScottK: If I enable ssl it still errors denying the relay I haven't tried port 25 although I don't see how that matters
[00:49] <ScottK> port 465/SMTPS is a different technology than TLS, so it was worth a shot.
[00:49] <ScottK> pastebin the log entry for the entire transaction.
[00:49] <histo> ScottK: which the mail.log?
[00:49] <ScottK> yes
[00:49] <ScottK> Also pastebin the output of postconf -n.
[00:52] <bobbyz> histo: share your dovecot.conf too?
[00:52] <histo> ScottK: http://paste.ubuntu.com/5675238/ is the mail.log http://paste.ubuntu.com/5675239/ is the postconf
[00:54] <histo> bobbyz: ScottK http://paste.ubuntu.com/5675244/ is my dovecot.conf
[00:55] <histo> This setup is from the wiki postfix virtual users and domains with clamav
[00:55] <histo> I believe the dovecot.conf has some obsolete settings from the wiki however there are only warnings in the dovecot logs
[00:57] <histo> I added the disable_plaintext_auth = no to the dovecot.conf to get outlook to login for pop and imap
[00:58] <ScottK> So there's no indication of SSL/TLS or SASL in that log snippet.
[00:59] <histo> ScottK: No that's the problem outlook isn't using either. Thunderbird does starttls and just works
[00:59] <histo> I don't have enough experience with these setups to know what needs to be enabled for outlook to just work.
[00:59] <ScottK> What if you enable SMTPS and set outlook to connect via port 465?
[00:59] <histo> I have pop working
[00:59] <histo> ScottK: How would one do that?
[01:00] <histo> I'm kind of confused by the whole listener thing in the dovecot.conf so is postfix actually listening or dovecot??? I'd assume it's just for internal communicaiton between dovecot and postfix but I don't know why they would even need to communicate
[01:01] <ScottK> histo: Look in /etc/postfix/master.cf and you'll find a commented out entry that starts: smtps     inet  n       -       -       -       -       smtpd
[01:01] <ScottK> Uncomment that and the succeeding lines then stop/start postfix.
[01:01] <ScottK> (not reload, for a new master.cf service to be recognized, you need to stop then start.
[01:03] <ScottK> Dovecot only talks to the outside world via pop3 and imap.  Any mail sending/delivery between other mail servers is handled by postfix.  So when postfix receives mail that is to be delivered locally, it has to give it to dovecot (thus communication required)
[01:04] <histo> ahh
[01:04] <histo> ScottK: what would the service imap and service pop lines in the dovecot.conf be for. I don't believe they are even needed
[01:05] <ScottK> I believe they aren't in the default setup if you have the right packages installed, but I mess with dovecot config rarely enough that I tend to go look things up to be sure.  Don't quote me.
[01:06] <ScottK> One problem at a time though.
[01:06] <histo> It's still denying it
[01:06] <histo> New error though client host regjected
[01:07] <ScottK> Progress.  Paste the log again
[01:08] <histo> Apr  3 21:07:45 bremecmail postfix/smtps/smtpd[7524]: connect from h230.171.189.173.dynamic.ip.windstream.net[173.189.171.230]
[01:08] <histo> Apr  3 21:07:46 bremecmail postfix/smtps/smtpd[7524]: NOQUEUE: reject: RCPT from h230.171.189.173.dynamic.ip.windstream.net[173.189.171.230]: 554 5.7.1 <h230.171.189.173.dynamic.ip.windstream.net[173.189.171.230]>: Client host rejected: Access denied; from=<admin@bremecgroup.com> to=<histoplasmosis@gmail.com> proto=ESMTP helo=<KimPC>
[01:08] <histo> Apr  3 21:07:46 bremecmail postfix/smtps/smtpd[7524]: disconnect from h230.171.189.173.dynamic.ip.windstream.net[173.189.171.230]
[01:08] <histo> Only 3 lines in log there
[01:10] <histo> ScottK: Those are the new entries in the log. Same generic errors
[01:10] <ScottK> Oh.
[01:10] <ScottK> I thought you said there was a new one.
[01:10] <histo> ScottK: On the client side there is a new error
[01:10] <ScottK> Oh.
[01:10] <ScottK> What's the exact error there?
[01:10] <histo> let me restart postfix to be sure
[01:11] <histo> ScottK: Server response 554 5.7.1 IPblah...: Client host rejected: Access denied
[01:11] <ScottK> Actually the error is different.
[01:11] <histo> Gawd I hate microsoft
[01:11] <ScottK> Before it said "Relay access denied"  Now is says " Access denied"
[01:12] <histo> Yes and keep in mind thunderbid on my linux laptop works perfectly fine
[01:12] <ScottK> So that's an indication you've switched from refusal to relay unauthenticated mail to an authentication failure.
[01:13] <ScottK> I don't have any MS clients here to mess with, so I'm not sure exactly what to do next and I can't mess with it.
[01:13] <ScottK> You're making progress though.
[01:13] <histo> It's still a 554 error though.
[01:14] <histo> This can't be that random of a question getting MS clients to work with a linux mailserver.
[01:14] <qman__> you're sure you have SMTP authentication set up on the client? from what I see it looks like it's not even trying
[01:14] <histo> qman__: I'm ticking my server requires authenticaiton in the windows live config
[01:15] <histo> and it's set to use the same settings as my incoming mail server
[01:15] <qman__> they bury it, http://blog.arvixe.com/how-to-turn-on-smtp-authentication-on-windows-live-mail/
[01:15] <qman__> ok
[01:16] <qman__> just checking, I've seen that problem derail lots of people
[01:18] <histo> qman__: yes I have those settings there. If I try port 465 and requires ssl I get that client host rejected. If I try port 25 without ssl I get the relay denied
[01:18] <ScottK> Does that user name and password work from thunderbird?
[01:19] <histo> in the mail.err log there is an error fatal: no SASL authentication mechanisms
[01:19] <histo> ScottK: yes
[01:19] <ScottK> Ah.
[01:19] <histo> ScottK: nvm that is an old error
[01:19] <ScottK> OK
[01:19] <histo> from hours ago when I was mucking around in configs
[01:19] <histo> It's fixed now
[01:19] <ScottK> Anything in mail.err should also be in mail.log, so that would have been weird.
[01:19] <histo> well the error is fixed but the problem persists from outlook
[01:20] <ScottK> What SASL mechanisms do you have enabled?
[01:21] <histo> ScottK: dovecot
[01:21] <histo> ScottK: and the path is private/auth-client
[01:21] <ScottK> No, I mean like PLAIN, LOGIN, etc.
[01:21] <histo> ScottK: plain login cram-md5
[01:21] <histo> in the dovecot.conf
[01:21] <ScottK> OK  That should be fine since login is what MS clients like.
[01:22] <histo> ScottK: in the postfix main.cf I see smtpd_sals_security_options noplaintext is that potentially the issue?
[01:22] <ScottK> Yes
[01:22] <ScottK> Good point.  Since you're using SSL/TLS you don't need that.
[01:23] <ScottK> LOGIN is a plain text mechansim.
[01:23] <qman__> yeah
[01:23] <qman__> I forget the names but the way it works is, you have to allow plaintext logins, but then there's another setting that says only allow plain over SSL/TLS
[01:23] <histo> jesus christ it works
[01:23] <histo> stupid outlook
[01:24] <histo> okay so now to enable pop3s imaps and disable the others
[01:24] <histo> That way everythign is ssl and outlook can be happy plaintexting passwords
[01:24] <ScottK> Good catch.
[01:24] <histo> ScottK: qman__ TY guys for helping me think through this
[01:24] <ScottK> You're welcome.
[01:25] <histo> Nother notch in my belt for reasons to never us M$
[01:25] <qman__> your linux client must have been using cram-md5
[01:25] <qman__> because that setting in particular is not outlook specific
[01:25] <histo> qman__: yea it was
[01:25] <histo> looking back at old linux client logs
[01:26] <histo> ScottK: I noticed in my master.cf smptd isn't enabled but bsmtp is that normal?
[01:27] <histo> nvm i'm seeing stuff now
[01:27] <ScottK> No
[01:27] <ScottK> OK
[01:27] <histo> Been at this awhile now i'm seeing things
[01:30] <histo> Now trying to enable imaps and pop3s
[01:31] <histo> actually dont' care about those nvm
[03:25] <hachre> I've just installed a fresh server and for some reason it is settings some of my vars like LC_ADDRESS to de_DE while the german language pack isnt even installed. I have LANG set to en_US.UTF-8 in /etc/default/locale but that doesn't seem enough.. What's missing here?
[03:28] <hachre> I also keep seeing this error message on many things apt does: "locale: Cannot set LC_ALL to default locale: No such file or directory", I'm not sure what it's referring to when it speaks of no such file or directly though since locale-gen works fine
[03:32] <sarnold> hachre: 'locale' will report the values of all the locale variables..
[03:33] <sarnold> hachre: maybe grep -r de_DE /etc/ would be useful?
[03:35] <hachre> sarnold: locale reports en_US only for LANG and the rest is on de_DE... grep shows only the normal aliases in /etc/locale.aliases
[03:35] <sarnold> hachre: how about your ~/.??* files?
[03:35] <hachre> sarnold: I can solve the issue by setting LC_ALL in /etc/default/locale but I'm wondering where it's getting the idea for de_DE from
[03:36] <sarnold> hachre: I think _some_ environment variables can come through via ssh; any chance you've ssh'd in from a de_DE terminal?
[03:36] <hachre> sarnold: nope
[03:36] <hachre> sarnold: oh wait... wow you're right
[03:36] <hachre> sarnold: my machine here has the same issue
[03:36] <sarnold> hahaha
[03:36] <hachre> sarnold: but it isnt complaining about it
[03:37] <hachre> running raring here on the desktop
[03:37] <sarnold> I wonder why (both counts...)
[03:38] <hachre> and I wonder why my desktop doesnt complain about it like the server does
[03:38] <hachre> I have no german language pack installed here either
[03:38] <hachre> youre probaly right and the server takes it from my ssh
[03:39] <hachre> im gonna set all my LC variables right and ssh back in to check
[03:39] <sarnold> at least the desktop I can flummox my way to changing the setting system-wide.. in the settings, language support, there's an "apply system-wide" button
[03:41] <hachre> ok
[03:41] <hachre> the server is fine
[03:42] <hachre> its really coming from my ssh
[03:42] <hachre> nice catch :)
[03:43] <sarnold> woo :)
[03:43] <sarnold> now you get to answer the same question on your desktop. hehe.
[03:43] <hachre> on my desktop the /etc/default/locale has this crazyness set up
[03:43] <hachre> was easy to find :)
[03:44] <hachre> who knows how that happened
[03:45] <sarnold> I've got a vague feeling I've seen some geolocation code in the installer to pre-select "the right timezone" and stuff, perhaps it also automatically picks locales, too? (though that'd put more faith in the geolocate stuff than I'd care to use myself...)
[03:45] <hachre> yea I've been thinking that too
[03:45] <hachre> and it detected rubbish for me when i installed
[03:45] <hachre> but i didnt care
[03:45] <hachre> maybe thats why
[04:42] <gazoombo> utlemming: I'm still interested in learning more about the build process for the Vagrant boxes, if you have time.
[04:46] <gazoombo> utlemming: CentOS is also looking to start releasing official Vagrant boxes: http://bugs.centos.org/view.php?id=6365
[08:26] <yousaf> Could do with some help
[08:27] <yousaf> dmesg output shows multiple instances of [    0.000000] *BAD*gran_size: 128M 	chunk_size: 2G 	num_reg: 8  	lose cover RAM: -904M
[08:27] <yousaf> its a dedicated unmanaged server, the hosts says they have done CPU/MEM test and are confident that its fine
[08:27] <yousaf> so what does those msgs in the dmesg output mean then?
[08:34] <R1ck> hard to tell from a single line, there are probably alot more
[08:35] <R1ck> could you pastebin the entire dmesg?
[08:49] <Arun_GP> Hello
[08:49] <Arun_GP> Is someone listening?
[08:49] <R1ck> yes
[08:50] <Arun_GP> :)
[08:50] <Arun_GP> I am writing a tutorial on LAMP stack
[08:50] <Arun_GP> My public_html has permissions 750. The problem is when my webserver uploads a fil/folder it takes ownership of that file/folder. So I can't modify them.
[08:50] <Arun_GP> Does you know a method in which, a newly created file/directory(by other users) inherits ownership of that directory.
[08:50] <Arun_GP> I know ACL won't work, because it inherits POSIX permissions only
[08:51] <Arun_GP> SGID inherits GID of directory
[08:51] <R1ck> Arun_GP: you mean when a file gets uploaded the owner is whatever Apache is running as (ie www-data) ?
[08:52] <Arun_GP> and POSIX states newly created file/directory inherit ownership from UID os that process
[08:52] <R1ck> don't pm me.
[08:52] <Arun_GP> Pardon me...new to IRC
[08:53] <Arun_GP> And, Yes that is what I meant
[08:53] <R1ck> ask your questions in here so multiple people can answer them or learn from the answers to your questions
[08:53] <Arun_GP> ok
[08:53] <R1ck> so, when you say "So I can't modify them.", what do you mean exactly, in what way are you trying to modify them?
[08:54] <Arun_GP> R1ck I went to stackoverflow and saw many webmasters are confused by this
[08:54] <Arun_GP> And some of them uses scripts to periodically check the permissions
[08:55] <Arun_GP> But I got another solution fsniper that uses inotify instead of continuous polling - http://freecode.com/projects/fsniper
[08:55] <R1ck> Just answer my question please
[08:56] <Arun_GP> Suppose Apache creates a directory
[08:56] <R1ck> Just answer my question please
[08:56] <Arun_GP> I do not have write access to that directory
[08:56] <Arun_GP> Because Pache takes ownership of that directory
[08:57] <Arun_GP> Did you get it R1ck?
[08:58] <R1ck> answer my question and I'll try to help you.
[09:00] <Arun_GP> Creating a subdirectory inside a directory can also be considered as 'modifying'
[09:00] <Arun_GP> If you knew how to solve this, you would have already did
[09:00] <Arun_GP> DOn't be a D1ck R1ck
[09:01] <Arun_GP> Bye
[09:01] <R1ck> *how* are you trying to create a subdirectory
[09:01] <R1ck> in a shell? ftp? website? rsync?
[09:01] <R1ck> ugh.
[11:21] <shafox> i added the ppa:nathan-renniewaldock/ppa for the latest php and mysql-server , but when i did sudo apt-get install php5 it installed the php fine but couldnt install the mysql-server5.5 showing me this error. https://gist.github.com/anonymous/5309328 ... any solutions for this one ??
[11:21] <shafox> i am on lucid btw
[11:38] <R1ck> shafox: find out why it fails to start
[11:41] <roniez> anybody knows if its possible to move the /home folders to another disk by creating a new partition on a 2nd disk for only the /home what would be the best way to go by that.
[11:49] <R1ck> roniez: create the new partition, copy all files from the old /home to the new location, add the partition to /etc/fstab, clean the old /home and mount /dev/newpartition /home
[12:32] <halvors> system.sysLocation.0
[12:32] <halvors> obs.
[12:32] <halvors> May anyone please help me out with the Ubuntu nagios-snmp-plugins package? :)
[12:32] <halvors> I can't get snmp to work probatly :(
[12:35] <halvors> I get the error in Nagios "ERROR: Description table : No response from remote host "192.168.0.2". " Does that mean that it can't reach the snmp host or that no information could be pulled from it?
[12:35] <halvors> I'm using the check command "check_snmp_int_v1!"GigabitEthernet-[1-6]"" :)
[12:35] <R1ck> halvors: either snmp isnt running or the community is invalid
[12:35] <R1ck> halvors: very snmp is working correctly with snmpwalk first
[12:40] <halvors> R1ck: It is, but maybe my snmp community isn't set probatly?
[12:41] <halvors> I have set the snmp community using this command in /etc/nagios3/resource.cfg
[12:41] <zul> yolanda:  ping https://code.launchpad.net/~zulcss/python-novaclient/2.13.0/+merge/157095
[12:41] <yolanda> hi zul
[12:42] <halvors> $USER7$=-C public -2
[12:42] <halvors> R1ck: Any idea? Is it set probatly then? Any more detailed log files i can check for nagios?
[12:46] <halvors> Rick: ? :)
[12:47] <ndee> I have a fresh server and want to install all the packages from another server. This is what I found: http://askubuntu.com/questions/17823/how-to-list-all-installed-packages when I try this, I get following message: http://pastebin.com/D643sp9m <-- how can this be on a total new server?
[12:49] <patdk-lap> easily, likely it isn't a real server, but some openvz thing
[12:50] <ndee> patdk-lap, no, it's a physical server, that I'm sure.
[12:58] <halvors> I get the error "ERROR: Description table : No response from remote host "192.168.0.2"." when trying monitor a cisco switch using the nagios-snmp-plugins ubuntu package. May anyone please help me out?
[13:02] <rbasak> ndee: are both servers running the same release?
[13:04] <ndee> rbasak, yes
[13:07] <patdk-lap> you did do an, apt-get update, first?
[13:09] <ndee> patdk-lap, yes
[13:09] <ndee> I made apt-get update, apt-get upgrade on the new server
[13:10] <ndee> then got the packages from other server with dpkg --get-selections, then set the new selection with dpkg --set-selections < packages on the new server and now that error shows up.
[13:11] <zul> yolanda:  ping https://code.launchpad.net/~zulcss/python-glanceclient/0.9.0/+merge/157106
[13:15] <ndee> adduser, ron and upstart are also installed so I'm really baffled.
[13:21] <ndee> I cleared the selections again with: dpkg -l '*' | grep '^in ' | awk '{ print $2 " deinstall" }' | dpkg --set-selections so everything is now as before. Is there another way to get server foo to the same state as server bar?
[13:23] <patdk-lap> personally, I use aptitude search '~i !~M' -F "%p install"
[13:23] <patdk-lap> do a diff between the two systems
[13:23] <patdk-lap> then install the diff
[13:28] <halvors> How do i set the SNMP community in the nagios commands in nagios-snmp-plugins?
[13:39] <hallyn_> zul: !  uh, can i get you to consider packaging libvirt 1.0.4 for raring at the last minute? :)
[13:39] <hallyn_> zul: it fixes bug 1157626 without a custom hack in qemu
[13:39] <zul> hallyn_:  sure
[13:40] <hallyn_> i'm not opposed to fixing it in qemu, but that's going to end up more different from usptream and debian, which will hurt in the long run
[13:40] <hallyn_> of course, with raring htere isn't really a long run :)
[13:40] <hallyn_> zul: awesome, thanks
[13:42] <RoyK> any idea when this fix will make it to the repos? http://www.postgresql.org/about/news/1456/
[13:42] <RoyK> seems to be pretty critical
[13:48] <ScottK> RoyK: I know that the security team knew it was coming, so I'd imagine "soon".
[13:51] <RoyK> ScottK: I hear from #ubuntu-bugs that the fix was released minutes ago. guess it takes a while to reach my repos
[13:51] <ScottK> Yes.
[13:52] <ScottK> Also, since the release included more than just the security fixes, for -security, the Ubuntu security team will probably have to segregate out the security changes to apply separately.
[13:53] <ScottK> Actually I'm wrong.
[13:54] <ScottK> It looks like 9.1.9 is published to security already.
[13:54] <ScottK> RoyK: Published 20 minutes ago, so it should be hitting mirrors real soon now.
[13:59] <zul> hallyn_:  do we need the enable qemu-spice patch anymore?
[14:01] <hallyn_> uh, lemme check
[14:02] <hallyn_> zul: no
[14:02] <hallyn_> zul: though, if a vm.xml specifies /usr/bin/kvm-spice, will that then still work?
[14:03] <zul> dunno
[14:03] <zul> ill leave it in then
[14:07] <mardraum> hallyn_: re 1157626 I'm here for a little while longer (1 hr?) until I need to sleep if you want me to try anything :)
[14:11] <ScottK> RoyK: http://www.ubuntu.com/usn/usn-1789-1
[14:12] <hallyn_> mardraum: after you migrate a vm with my patch applied in qemu, have you logged in and really used it?  Do you see any corruption with things coming from disk (which weren't yet in page cache)?
[14:13] <RoyK> ScottK: thansk
[14:13] <hallyn_> mardraum: basically there are 3 ways to go:  (1) my patch, (2) a probably better patchset from upstream, (3) newer libvirt.  so zul is packaging newer libvirt right now, andI'll run a testsuite on it
[14:13] <hallyn_> if it looks ok, we'll try to push that as a cleaner fix
[14:13] <mardraum> hallyn_: yes, I have logged in, run new binaries, rebooted etc
[14:13] <mardraum> all ok
[14:14] <halvors> Anyone know how to use the check commands in the nagios-snmp-plugins package?
[14:14] <mardraum> still using the 12.10 test vm
[14:14] <hallyn_> mardraum: awesome, thanks
[14:14] <mardraum> I'll install some new software now
[14:14] <hallyn_> mardraum: (just saw your email reply sorry :)
[14:14] <mardraum> no worries
[14:15] <mardraum> just installed nginx for example, and the depends all ok
[14:31] <zul> hallyn_:  just doing a local build now
[14:35] <hallyn_> mardraum: cool.
[14:36] <hallyn_> zul: ^ newer libvir tis the *cleaner* fix but after spending 1.5 days on that one-line qemu fix, i'm attached to it dammit :)
[14:37]  * hallyn_ biab
[14:38] <zul> hallyn_:  ftbfs :( ill fix it up
[14:39] <patdk-lap> :)
[14:52] <mardraum> hallyn_: thanks for your time on this one. bbl
[15:16] <zul> hallyn_:  around?
[15:20] <hallyn_> zul: yup
[15:21] <zul> hallyn_:  im just in the middle of uploading libvirt to http://people.canonical.com/~chucks/libvirt can you see why the testsuite fbtfs please
[15:23] <hallyn_> ok
[15:24] <zul> yolanda:  ping https://code.launchpad.net/~zulcss/nova/2013.1/+merge/157138
[15:24] <yolanda> ok
[15:39] <zul> yolanda: ping https://code.launchpad.net/~zulcss/swift/1.8.0/+merge/157142
[16:14] <hallyn_> zul: fwiw at least the first error appears to be a real bug in the test driver...  just doing '../tools/virsh -c test:///default list' should show the test domain, and doesn't.  <huh>
[16:16] <bobbyz> hallyn_: I haven't seen all of your above, but have you added '--all' flag to that command?  If the domain is not running, you won't see it in 'list' output unless you tag --all onto that
[16:27] <hallyn_> bobbyz: this is the test driver.  it does show up without --all in 1.0.2, and teh testcases (upstream) expect it
[16:39] <hallyn_> zul: i mean ffs, 1.0.2..1.0.4. has 729-line diff in src/test/test_driver.c.
[16:45] <yousaf> hi all
[16:45] <yousaf> I am having issues with twitter oauth, after debugging i found out that my server can not literally find twitter.com
[16:46] <yousaf> any reasons why?
[16:51] <Fieldy> i'm trying to get an ipv6 tunnel up. I have done this many times on many distributions and kernel versions, however for some reason on this ubuntu server system, i'm getting a problem. any ideas? commands and error output: https://gist.github.com/anonymous/5312013/raw/4ccccaba09acfaf5239bba852ed1031d52475d93/stdin.txt
[16:52] <Fieldy> i've done a lot of searching on that error, with nothing solid as a result
[16:52] <Daviey> roaksoax: I know you are committed with a tonne of stuff right now, but in the next few days.. can you work with plars to coordinate MAAS ISO tests for Raring?
[16:53] <roaksoax> Daviey: sure thing
[16:58] <sarnold> Fieldy: a few shots-in-the-dark -- does your kernel.modules_disabled sysctl prevent loading one of the ipv6 modules on demand?
[16:59] <Fieldy> sarnold: i'll have a look, that's actually what i was just now poking around in -- i had formerly disabled ipv6 (because I didn't need it yet) in 3 sysctl lines however I reversed that (by echoing the opposing value by hand to them)
[16:59] <Fieldy> and confirmed that worked by catting them
[17:00] <Fieldy> I don't have kernel.modules_disabled at all in /etc/sysctl.conf or /etc/sysctl.d/*
[17:00] <sarnold> Fieldy: okay, then it's probably not changed (you could check with sysctl directly) -- are the ipv6 modules blacklisted in /etc/modprobe.d/ ?
[17:01] <Fieldy> ooh, no i didn't. i'm tired and was doing crap like echo "0" > net.ipv6.conf.all.disable_ipv6   ... FAIL. let me do that right
[17:02] <Fieldy> that'll teach me to get with the times -- i'm so used to just echoing and catting
[17:04] <sarnold> Fieldy: heh, well... the kernel devs have wanted to kill off sysctl for years. I'm not sure they're ever going to do it, but it's been on their minds for ages. blah.
[17:04] <Fieldy> heh heh heh. that was it.
[17:05] <sarnold> Fieldy: nice :D
[17:05] <Fieldy> cough: ls                   bin  mysql-backups  net.ipv6.conf.all.disable_ipv6  net.ipv6.conf.default.disable_ipv6  net.ipv6.conf.lo.disable_ipv6
[17:05]  * Fieldy deletes those and pretends they never existed
[17:06] <Fieldy> sarnold: unless they have something that much better... leave it alone imo
[17:06] <sarnold> Fieldy: yeah. I've always liked sysctl. it's easier to sysctl -a | grep ... than it is to find /proc -name '*...*'  :)
[17:06] <Fieldy> no doubt
[17:08] <qhartman_> I'm trying to install ubuntu server 12.04.2 i386 on an older P4 IBM server. The installer gets to 77% complete and then asks for a disk change. Has anyone found a work around for this? I've checked all the MD5's and done a disk validation, so I know the media is good.
[17:08] <qhartman_> I also know the drive is good, I've used it to install other distros and versions on this machine w/o problems
[18:25] <zul> hallyn_:  sucks doesnt it have you figured out whats going on with libvirt?
[18:29] <Quest> how to know how much bandwidth is taken on a specific port?
[18:29] <hallyn_> zul: no i haven't
[18:29] <hallyn_> i'm buildign from git right now
[18:29] <hallyn_> zul: i mean yes - the test:///default is not properly registering domain test
[18:29] <Quest> rephrase:  I need to know what bandwidth is taken on trafic on a specific port. how it can be done?
[18:30] <rbasak> Quest: look into iftop
[18:30] <Quest> hm
[18:30] <rbasak> (though that's interface-based, not port-based)
[18:30] <zul> hallyn_:  awesome
[18:31] <zul> adam_g:  how do you want to review these 2013.1 merge requests do you want them one by one or do yout want me to queue them up for you
[18:31] <adam_g> zul, doesnt matter. im actually about to push a keystone bug fix that would be great to get out /w 2013.1
[18:32] <zul> adam_g: okies i havent gotten to keystone yet so that would be cool
[18:33] <zul> adam_g: btw https://code.launchpad.net/~zulcss/glance/2013.1/+merge/157196
[18:35] <adam_g> zul, https://code.launchpad.net/~gandelman-a/ubuntu/raring/keystone/lp1158563/+merge/157199
[18:41] <Quest> rbasak,  how to see traffic on a specific port by iftop?
[18:43] <plars> Anyone here who could help with the maas tests or iscsi tests on iso tracker?
[18:44] <Quest> rbasak,  ?
[18:46] <hallyn_> zul: well, with libvirt from git i don't get thta problem.  wtf?
[18:47] <zul> hallyn_:  maybe something is missing from the tarball (wouldnt surprise me)
[18:47] <hallyn_> true
[18:48] <hallyn_> but, 2 other possibilities:  1. i had to apt-get purge libcurl4-gnutls again to get git version to build.  maybe it built but wrongly bc of that from package.  (re-building to test)
[18:48] <hallyn_> 2. could be osme patch we have.
[18:48] <hallyn_> do you want to try building from the release tarball?
[18:48] <hallyn_> (else i'll try that third)
[19:05] <vedic> I have removed .bash_history file but I still see output of about 10 commands when I write 'history' on the terminal. How to clear history completely?
[19:07] <zul> adam_g:  https://code.launchpad.net/~zulcss/quantum/2013.1/+merge/157205
[19:08] <sarnold> vedic: see the HISTIGNORE and HISTFILE environment variables in the bash(1) manpage
[19:09] <hallyn_> zul: well huh, for some reason it does not get past
[19:09] <hallyn_>     if (!conn->uri->scheme || STRNEQ(conn->uri->scheme, "test"))
[19:09] <hallyn_>         return VIR_DRV_OPEN_DECLINED;
[19:09] <hallyn_> (it being testOpen())
[19:09] <zul> which file is that in?
[19:10] <hallyn_> zul: src/test/test_driver.c
[19:10] <hallyn_> zul: OMG conn->uri->scheme is 'qemu', despite "-c test:///default"
[19:11] <zul> hallyn_:  can you do me a favor? can you drop 9002-better_default_uri_virsh.patch
[19:11] <hallyn_> yup
[19:12] <hallyn_> (gonna take a few misn to build)
[19:13] <zul> k
[19:22] <hallyn_> zul: all tests pass with that patch dropped
[19:23] <zul> hallyn_: hah thought so
[19:23] <hallyn_> zul: odd, it seems to have been ported correctly at first glance
[19:23] <zul> i noticed debian isnt using that patch anymore either
[19:39] <hallyn_> zul: i think i see why
[19:39] <zul> oh?
[19:39] <hallyn_> zul: commit abfff210060625af8914e28601f1ec6ed96b05ae switched the order to first call vshInit(), then parse argv
[19:40] <zul> ah
[19:44] <jackln> I'm trying to setyp a PPTP VPN server, what IPs do I use?!
[19:44] <jackln> for remoteip and localip
[19:44] <hallyn_> zul: building with that reversed
[19:52] <red82> can anyone point me at a good resource that I can read up on the process of replicating a mysql database from server A to B such that server B can function as a failover in the even that A goes down?  Application is e-commerce.
[19:55] <plars> is http://iso.qa.ubuntu.com/qatracker/testcases/1463/info still valid? I'm not seeing these options listed in preferences
[19:58] <jackln> Do you know anything about VPNs?
[19:59] <hallyn_> zul: yeah with http://people.canonical.com/~serge/9002-better_default_uri_virsh.patch it works
[20:00] <hallyn_> aaaand.... the machine i was doing that just shut down for no reason
[20:00] <hallyn_> hardware sucks
[20:07] <hallyn_> zul: OTOH, virsh list now gives me segfault.  <fume>
[20:07] <hallyn_> oh, heh.
[20:07] <hallyn_> left some debug cruft in my pkg
[20:25] <hallyn_> zul: i dunno, pkg isn't working for me - won't show my installed vms
[20:26] <hallyn_> or have they been deleted
[20:27]  * hallyn_ slaps self
[20:28] <hallyn_> yeah nm it's working.  now to run the qa tests
[21:47] <cwinebrinner> echo $-
[21:47] <cwinebrinner> echo $T
[21:47] <cwinebrinner> woops >.>
[21:50] <adam_g> zul, not sure if its related to that keystone fix from earlier, but i can't the test suite to pass for the life of me.
[22:05] <hallyn_> zul: so the qa tests mostly pass, except there is some domxml-to-native weirdness
[22:10] <Quest> what is the package name to install GUI  on ubuntu server? (i think its unity or gnome for ubuntu?) (and i would prefer kubuntu GUI (kde) ?
[22:13] <sarnold> Quest: try: 'sudo apt-get install kubuntu-desktop^'
[22:13] <Quest> whats ^ for?
[22:13] <sarnold> (the ^ asks to install the _task_ named kubuntu-desktop, which is different from the package named kubuntu-desktop)
[22:14] <Quest> really
[22:15] <Quest> if i omit ^. will it install kde?
[22:16] <sarnold> Quest: it'll install the metapackage kubuntu-desktop. you might never know the difference between the two.
[22:19] <Quest>   Temporary failure resolving 'us.archive.ubuntu.com'
[22:21] <sarnold> worked for me locally, and with both 8.8.8.8 and 4.2.2.1 public recursors
[22:29] <Quest>  I just ssh through public ip to my server. but it seems i cannot ping -s 25 google.com or apt-get update.
[22:30] <Quest> are all ports open by default on ubuntu-server (which is contrary to ubuntu desktop)?
[22:32] <hallyn_> zul: ffs - it dies unless youhave a mem-balloon specified in the xml
[22:33] <Quest> sarnold,  ^?
[22:36] <sarnold> Quest: yes, all ports are open by default on both ubuntu-server and ubuntu-desktop. if you've configured iptables (or one of its frontends like ufw) then of course that'll be different.
[22:37] <Quest> sarnold,  i just installed ubuntu-server  . it updated fine during installation . but now its not
[22:37] <sarnold> Quest: check demsg. check 'ping 4.2.2.1'.
[22:40] <Quest> let me se
[23:22] <Quest> I just installed ubuntu-server . it boots with a black / blank screen and does not shows anything. what can be wrong?
[23:23] <patdk-lap> kernel graphics driver issue probably
[23:25] <Quest> patdk-lap,  what can be done?
[23:34] <Quest> whats the main difference b/w ubuntu server and ubuntu desktop. ? i am having problems installing server . can i go with desktop?
[23:36] <sarnold> Quest: server doesn't come with a gui and doesn't use NetworkManager to manage its interfaces. The desktop variant will use NetworkManager, and come with a GUI.
[23:37] <sarnold> Quest: you sure can use the desktop variant as a server platform if you wish -- I've done that for my pandaboard, which has no monitor connected
[23:41] <Quest> hm
[23:41] <Quest> one add is ssh
[23:41] <Quest>  i can install openssh my self. any thing else that is different?
[23:42] <patdk-lap> there is 0 difference between ubuntu server and desktop
[23:42] <patdk-lap> just what default packages are installed
[23:43] <sarnold> Quest: you can compare the package manifests at e.g. http://releases.ubuntu.com/precise/
[23:43] <sarnold> oh. uh. there's no server manifest there.
[23:57] <Quest> iam having a general error of mounting file systems. iam at shell while installing kubuntu. how to install from shel?