[00:36] <kirkland> hggdh: it's pretty much EoD ... want to just wait for tomorrow's?
[00:37] <kirkland> smoser: i don't really think it's worth waiting around for another two hours, then rerunning the tests (personally)
[01:08] <hggdh> kirkland: I will wait for tomorrow, and re-run the tests
[01:11] <kirkland> hggdh: ack
[03:31] <MTecknology> This may be a long stretch asking in here but... When I plug in an external drive, dmesg shows me that the drive and partitions are detected. I have pcmanfm installed and I want the new device to show up there so I can click and auto-mount the partition. The way I'm doing it now I have to drop to cli and manually mkdir && mount. What do I need to do for that to work?
[03:42] <thebwt> MTecknology: when you say "show up there" where is there?
[03:43] <MTecknology> thebwt: the file manager
[03:44] <thebwt> MTecknology: hmm to be more specific, do you want the filesystem to be mounted in the same place everytime or in a different place based on your firle manager's current working directory
[03:46] <MTecknology> thebwt: I have a very very slim version of Ubunut, I would love to have it working the same as it works in any other full version of ubuntu
[03:46] <thebwt> basically I'm seeing two paths, automount in fstab/mtab (not sure how it works exactly); or a script/plugin for pcmanfm (which I've never even heard of before)
[03:47] <thebwt> MTecknology: right now they auto mount in /media and then are sym linked to a "desktop" and added to the gnome "book marks"
[03:47] <MTecknology> ya - that
[03:47] <thebwt> does pcmanfm have a desktop manager?
[03:48] <MTecknology> no
[03:48] <thebwt> kk so don't have to worry about that
[03:48] <thebwt> you need to automount to /media. I don't know how, but I know where to find out how... let me look it up
[03:48] <MTecknology> I wonder what package gnome-volume-manager comes in...
[03:48] <MTecknology> maybe that could help :S
[03:49] <thebwt> not if you're trying to keep it tight
[03:49] <thebwt> the solution is probably super simple
[03:49] <MTecknology> OH!
[03:49] <MTecknology> pmount
[03:49] <MTecknology> !info pmount
[03:49] <thebwt> nice
[03:50] <MTecknology> thebwt: I just noticed this page saying pmount is hard coded in gnome-volume-manger - http://www.togaware.com/linux/survivor/Using_Gnome_Volume_Manager.html
[03:50] <MTecknology> thebwt: nwo how to use it :P
[03:50] <thebwt> indeed
[03:51] <thebwt> need to know how to have a script be called when a removable device is inserted.
[03:52] <MTecknology> pmount /dev/sdb1
[03:52] <MTecknology> mounted the drive to /media
[03:52] <thebwt> but first, go ahead and write a script that does what you need it to do. That is, mount it (and add to pcmanfm bookmarks?).
[03:52] <MTecknology> no permission or anything though
[03:52] <thebwt> neat
[03:53] <MTecknology> I'm thinking I need to go bck to something with a volume manager
[03:54] <thebwt> perhaps, but if you have the time to learn it, I'd do so.
[03:54] <MTecknology> thebwt: I mounted as root - probably wasn't supposed to :P
[03:54] <thebwt> hah
[03:55] <MTecknology> that's it...
[03:56] <MTecknology> thebwt: so... pmount is definitely the answer in some shape or another :D
[03:58] <MTecknology> thebwt: any idea how to figure out what package provides gnome-volume-manager?
[03:59] <MTecknology> !info gnome-volume-manager
[03:59] <MTecknology> !info gnome-volume-manager lucid
[03:59] <MTecknology> hrm...
[03:59] <thebwt> http://manpages.ubuntu.com/manpages/hardy/man1/gnome-volume-manager.1.html
[04:00] <thebwt> package seems to be gnome-colume-manager
[04:00] <thebwt> volume*
[04:00] <MTecknology> thebwt: seems to have gone away in lucid
[04:01] <thebwt> look at teh package halevt ...
[04:01] <thebwt> looks right up your alley
[04:02] <MTecknology> :D
[04:02] <MTecknology> yay - broke installing
[04:04] <MTecknology> thebwt: http://dpaste.com/184401/
[04:05] <thebwt> eww root :p
[04:05] <MTecknology> thebwt: hm?
[04:05] <MTecknology> thebwt: sudo -s
[04:06] <thebwt> *nods*
[04:06] <MTecknology> I do have a root pass on here - but good reasons for it :P
[04:06] <thebwt> http://www.mail-archive.com/debian-bugs-closed@lists.debian.org/msg214174.html
[04:06] <MTecknology> extremely long, rarely ever used
[04:08] <thebwt> ah
[04:08] <thebwt> it won't start without a configuration script
[04:09] <thebwt> fixed upstream it seems
[04:09] <MTecknology> this version is upstream
[04:09] <MTecknology> !info halevt
[04:09] <MTecknology> !info halevt lucid
[04:10] <thebwt> debian has 0.1.5-4
[04:11] <thebwt> http://packages.debian.org/unstable/main/halevt
[04:11] <thebwt> .debs are there
[04:12] <MTecknology> the bug was fixed before that last sync to lucid it seems
[04:13] <MTecknology> I guess I don't need that
[04:15] <thebwt> http://packages.debian.org/changelogs/pool/main/h/halevt/halevt_0.1.5-4/changelog
[04:15] <thebwt> bug was fixed in -4
[04:15] <thebwt> lucid has http://packages.ubuntu.com/it/lucid/admin/halevt
[04:16] <MTecknology> oh..
[04:24] <MTecknology> thebwt: there's a usbmount package....
[04:25] <thebwt> hal events really seems like what you need though
[04:25] <thebwt> *looks up usbmount*
[04:25] <MTecknology> thebwt: usbmount seems to jsut toss scripts in for udev
[04:25] <thebwt> *nods*
[04:26] <thebwt> I rescend my commit
[04:26] <MTecknology> maybe not exactly what I want - but it's hacky enough to work
[04:28] <thebwt> what other functionality do you need?
[04:28] <MTecknology> it would be nice if it were a little cleaner
[04:31] <MTecknology> thebwt: beggers can't be choosers - I know there's a better way but right now - it works :D
[04:33] <MTecknology> thebwt: if it could let me unmount them that would be great too but- oh well
[05:31] <arthurjohnson> clamav ruined me
[08:15] <laen> Is there a script, similar to Redhat's /etc/rc.sysinit, which is run once at boot?
[08:16] <Airells> hi , dpkg-reconfigure slapd  dont ask me about new password , what is the easiest way to change password ?
[08:16] <Airells> ubu 9.10
[08:20] <ttx> laen: /etc/init.d/rc.local ?
[08:21] <ttx> hm /etc/rc.local, sorry
[08:22] <laen> Hm, okay.
[09:39] <e-DIO-t> Yo!
[10:15] <binBASH> ttx: those images from ubuntu uec have no password?
[10:17] <ttx> binBASH: they use public key auth through SSH
[10:17] <binBASH> oh, ok
[10:18] <binBASH> so I have to create own because I don't get ips automatically assigned
[10:18] <ttx> binBASH: you're supposed to spin them up with some -k or userdata that will install a key in ~ubuntu/.ssh/authorized_keys
[10:19] <ttx> you can even use Launchpad SSh keys if you have one
[10:19] <ttx> uec-run-instances -l ttx $EMI
[10:39] <maxagaz> hi
[10:39] <maxagaz> I've got this error message on a desktop machine : Use of uninitialized value $item in hash element at /usr/share/perl5/Debconf/DbDriver/File.pm
[10:40] <maxagaz> when I do "aptitude install"
[10:40] <maxagaz> someone knows how to fix it ?
[11:16] <binBASH> ttx: Ok, will try because I have an ip assigning problem ;)
[11:52] <\sh> zul, when will  mysql-5.0 be removed from universe (lucid) ? looks like it's still in there ;)
[12:32] <mbiebl> hi, is this the right channel to for EC2 related questions?
[12:32] <mbiebl> I was wondering, why I have 5 different kernels within my karmic AMI
[12:32] <binBASH> This is the channel for Ubuntu Server Edition including Ubuntu Enterprise Cloud
[12:33] <mbiebl> binBASH: I have 2.6.31-14-server, 2.6.31-17-server, 2.6.31-20-server, 2.6.31-302-ec2 and 2.6.31-305-ec2 installed
[12:33] <mbiebl> but only 2.6.31-302-ec2 seems to be used?
[12:35] <binBASH> this is not really amazon support channel ;)
[12:35] <binBASH> I don't have amazon ec2, too
[12:39] <mbiebl> binBASH: Well, the images are from Ubuntu/Canonical
[12:40] <J_P> why ubuntu server 9.10 don't have nessus?
[12:40] <binBASH> mbiebl: Think those are from Amazon
[12:40] <binBASH> ;)
[12:40] <mbiebl> binBASH: they are from http://uec-images.ubuntu.com/releases/karmic/release/
[12:41] <mbiebl> I used ami-05c2e971
[12:41] <binBASH> ahh, these are for usage in Ubuntu Enterprise Cloud.
[12:41] <binBASH> if your run private cloud
[12:44] <ragi2010> I'm using the latest ubuntu server ami ebs boot from alestic.com, when I terminate the instance, the ebs also  gets deleted. How do I change this behaviour
[12:45] <J_P> anyone?
[12:46] <binBASH> J_P: Maybe because of licensing issues.
[12:46] <mbiebl> binBASH: are those the official,supported images from Canonical for EC2 or not?
[12:47] <J_P> binBASH: humm ok
[12:47] <binBASH> mbiebl: Like I said they're for usage in Ubuntu Server cloud. If you run a ubuntu enterprise cloud installation. It's a private version compatible to Amazon.
[12:47] <binBASH> J_P: I think nessus has .deb packages for download
[12:48] <J_P> binBASH: yes, I'm doing that.. :-)
[12:48] <binBASH> J_P: Ubuntu should have openvas
[12:49] <binBASH> an opensource alternative to nessus
[12:50] <J_P> binBASH: yes, I install openvas, but that is strange.. say like as nessus download app
[12:50] <binBASH> http://packages.ubuntu.com/de/source/karmic/openvas-client
[12:51] <drbobb> hey, how do i install and activate the git-daemon service on ubuntu?
[12:52] <drbobb> it doesn't appear to work seamlessly like just about all service daemons on debian-derived systems
[12:52] <drbobb> in other words, I think it's somehow foobared ;-/
[12:57] <zul> \sh: the request has been filed with the archive admin
[13:05] <drbobb> again, does anybody know how to achieve a working git-daemon on ubuntu?
[13:13] <drbobb> ok there is a package called `git-daemon-run' but it depends on some weirdness called `runit' which I never heard of before, and which is not used by anything else on my system
[13:13] <drbobb> oh and btw it's broken anyway, the git-daemon service fails to run with some cryptic error
[13:19] <drbobb> wow, it was known already in 2008 that git-daemon-run is broken
[13:20] <drbobb> and apparently it still hasn't been fixed
[13:31] <arthurjohnson> drbobb:  you should fix it.
[13:37] <hazmat> i'm having an issue adding a custom apt repository to karmic installations, using add-apt-repository or manually by hand into /etc/sources.d.list along with the key... after i'll do an apt-update and see a fetch for the package manifests. but doing apt-cache search will still show the package version from the dist repository instead of the new repository.. is there something obvious i'm missing?
[13:48] <TeTeT> hazmat: try apt-cache policy <package> to get more information
[13:51] <stephank> I'm reading about NSS and PAM. I've read that pam_unix uses standard C-library calls. Does that mean it uses NSS? And would I thus be able to set up an LDAP server with anonymous access somewhere, configure nss-ldap to use it (for passwd and shadow), and continue using pam_unix for authentication?
[13:52] <ahasenack> stephank: it's possible, but that would mean every user would have access to the hashes of the other users (shadow)
[13:53] <ahasenack> stephank: unless playing some tricks with root_bind_dn in nss_ldap
[13:53] <ahasenack> stephank: i don't remember all the implementation details, but you should watch out for that
[13:53] <stephank> ahasenack: ah, okay. that makes sense
[14:04] <AlexC_> morning
[14:05] <AlexC_> I've got a weird issue whereby 'aliases' added in /etc/network/interfaces via 'up ip addr add ....' do not always get added when I do a '/etc/init.d/networking restart' - some do, but not all - what is going on?
[14:12] <oru_work> urgent: mail stopped working because i'm running clamav and the new version was released
[14:13] <oru_work> i really need to know what is a proper way of upgrading to the new version
[14:13] <pmatulis> oru_work: mail stopped working?  what do you mean?
[14:14] <oru_work> pmatulis, none of my employees received mail this morning
[14:14] <oru_work> i'm checking mail.log and its really complaining about the outdated anti-virus
[14:15] <pmatulis> oru_work: outdated doesn't mean it's not working
[14:15] <pmatulis> oru_work: are you receiving mail or not?
[14:16] <oru_work> pmatulis, no i'm not
[14:17] <pmatulis> oru_work: so there must be an MTA error.  what MTA are you using anyway?
[14:17] <oru_work> postfix
[14:19] <pmatulis> oru_work: so pastebin the error(s)
[14:20] <oru_work> pmatulis, i just upgraded clamav antivirus and mail started working again, but its really messed up that all of mail had been lost since last night
[14:20] <oru_work> or not
[14:20] <oru_work> hang on
[14:20] <pmatulis> oru_work: standard MTAs will send mail for 5 days
[14:20] <bogeyd7> it should start processingthe que
[14:21] <bogeyd7> pmatulis, that doesnt mean anything when it is actually delivered and is just awaiting lmtp
[14:21] <reya276> morning
[14:21] <pmatulis> oru_work: or your deffered queue will be processed
[14:21] <PazDog> hey, how do i make sshd block multiple logins for a user. Like if the account user1 logged in twice, both sessions would be closed
[14:21] <pmatulis> bogeyd7: he didn't say if the mail was rejected or not, but i could have guessed not
[14:21] <reya276> does anyone know how to fix the Bad request in apache if you have SSL on with multiple sites?
[14:22] <bogeyd7> reya276,  https://help.ubuntu.com/community/forum/server/apache2/SSL
[14:22] <oru_work> pmatulis, as soon as i updated clamav i'm seeing lots of activitiy in tail -f mail.log , so I guess all the mail that was in que is getting processed atm
[14:22] <oru_work> i'm happy :)
[14:22] <reya276> right now if you try to hit this website http://www.accureports.com it will give you a bad request in firefox?
[14:22] <pmatulis> oru_work: ok, good
[14:22] <bogeyd7> oru_work, now you need to setup your freshclam for hourly updates
[14:23] <bogeyd7> cron.hourly is where you set it up
[14:23] <oru_work> bogeyd7, any specifics or details on that please ?
[14:24] <bogeyd7> oru_work, sudo su root , then crontab -e, then make a crontab entry for the freshclam to run hourly
[14:24] <bogeyd7> that is easier than trying to make a cron.hourly script
[14:31] <AlexC_> I've got a weird issue whereby 'aliases' added in /etc/network/interfaces via 'up ip addr add ....' do not always get added when I do a '/etc/init.d/networking restart' - some do, but not all - what is going on?
[14:33] <electro_> I am trying to install Ubuntu on a sunfire x4100 and I'm getting an error about no disks.  I see that this issue existed in dapper, but I am still running into the issue.  Has anyone found a workaround?
[14:34] <electro_> https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.15/+bug/37452
[14:35] <reya276> bogeyd7: would this tutorial apply to 9.10?
[14:35] <reya276> bogeyd7: is says for 7.10
[15:16] <oru_work> bogeyd6, around ? check your pm please :)
[15:18] <oru_work> i'm getting the following error ./freshclam
[15:18] <oru_work> ERROR: /var/log/clamav/freshclam.log is locked by another process
[15:18] <oru_work> ERROR: Problem with internal logger (UpdateLogFile = /var/log/clamav/freshclam.log).
[15:21] <smoser> mathiaz_, ping me when you have a minute
[15:30] <ttx> zul, smoser, kirkland, mathiaz: could one of you cover the following beta2 testcase today ? http://iso.qa.ubuntu.com/qatracker/result/3919/480
[15:31] <ttx> ara wants all optional cases covered before we start RC testing... and will cover iSCSI ones
[15:31] <zul> ttx: i can do it
[15:31] <ttx> zul: thanks !
[15:34] <kirkland> ttx: do you mind if it's done in a KVM ?
[15:35] <kirkland> ttx: ah, i see zul is the man ;-)
[15:35] <ttx> zul, smoser, kirkland, mathiaz: if you run out of ideas on bugs to fix, there are some unassigned bugs in the list @ https://wiki.ubuntu.com/ServerTeam/ReleaseStatus :)
[15:36] <ttx> that's mostly for smoser a,d mathiaz, since the others already have some :)
[15:36] <zul> ttx: ill never run out of ideas ;)
[15:38] <smoser> ok. thanks ttx
[15:43] <ttx> mathiaz: let me know if you received my "Server Team 20100414 meeting minutes" recent email
[15:49] <electro_> I am trying to install Ubuntu on a sunfire x4100 and I'm getting an error about no disks.  I see that this issue existed in dapper, but I am still running into the issue.  Has anyone found a workaround?
[15:50] <ivoks> which version do you use?
[16:22] <ivoks> mathiaz: hey
[16:23] <ivoks> mathiaz: i have a working solution for dovecot-postfix
[16:23] <mathiaz> ivoks: glad to hear that!
[16:23] <mathiaz> ivoks: have you tested upgrade as well?
[16:24] <ivoks> no, i just created a setup that works with default (upstream supported) config location
[16:24] <ivoks> i'll work on upgrade paths now
[16:25] <ivoks> # Config files can also be included. deliver doesn't support them currently.
[16:25] <ivoks> #!include /etc/dovecot/conf.d/*.conf
[16:25] <ivoks> # Optional configurations, don't give an error if it's not found:
[16:25] <ivoks> #!include_try /etc/dovecot/extra.conf
[16:25] <zul> ttx: iso test passed
[16:26] <ivoks> break and than wrap up this dovecot-postfix once for all
You searched for ivoks did you mean "Ewoks"?</italic>
[16:27] <ivoks> new config syntax allows us to do wonders in maverick
[16:27] <mathiaz> ivoks: maverick?
[16:27] <mathiaz> ivoks: what's the state for lucid then?
[16:27] <ivoks> for lucid i have solution
[16:28] <ivoks> for maverick we can split dovecot-postfix
[16:28] <ivoks> and use a tool to setup mail server
[16:28] <ivoks> that would copy config files to conf.d
[16:28] <ivoks> instead of using separate package
[16:28] <mathiaz> ivoks: what's the solution for lucid then?
[16:29] <ivoks> mathiaz: dovecot-postfix stays and i'll send a patch before 8AM CET
[16:29] <mathiaz> ivoks: could you give me a quick (2 sentence) overview of the solution?
[16:30] <ttx> kirkland/mathiaz: Foundations just asked that we help them on bug 557429, since they are getting their hands full with installer/mountall/plymouth
[16:30] <ivoks> mathiaz: put dovecot-postfix.conf in /etc/dovecot/conf.d/, enable include_try for that directory in /etc/dovecot/dovecot.conf
[16:30] <ttx> kirkland/mathiaz: I know it's not necessarily your area of expertise, but your combined looks is probably our best chance to help them
[16:30] <ivoks> mathiaz: add /etc/dovecot/auth.d and put auth part of dovecot-postfix.conf into /etc/dovecot/auth.d and add include_try under auth section of /etc/dovecot/dovecot.conf
[16:31] <ttx> kirkland/mathiaz: could you add it to your agenda ? jdstrand submitted it, so he can give you any missing detail
[16:31] <ivoks> so it works in the same way as apache, amavis...
[16:31] <mathiaz> ivoks: ok - sounds like a good plan to me
[16:31] <mathiaz> ivoks: how about upgrades?
[16:32] <ivoks> mathiaz: that could be tricky
[16:32] <ivoks> mathiaz: i'll look at the possible issues
[16:32] <kirkland> ttx: ugh, okay
[16:32] <ivoks> mathiaz: i'll work it out
[16:33] <kirkland> ttx: assign to me, i'll fight it
[16:33] <philsturgeon> i have a user csahost who can log into FTP, but is not allowed permssion to write files. the user csahost does not show when i do $ cat /etc/passwd
[16:34] <ttx> kirkland: we at least need to see if that's not a known issue, since apparently the package lags with upstream and this is *not* a lucid regression
[16:34] <ttx> kirkland: thanks !
[16:41] <skrite99> having trouble with dnsmasq, i can use it on a local machine, but other computers on the lan seem not to be working with it
[16:42] <persia> Do you have libvirt installed?
[16:43] <skrite99> i may not. checking
[16:44] <persia> If you do, there's some interaction issues that mean you have to configure dnsmasq in a special manner.
[16:44] <skrite99> install by apt  package libvirt0 ?
[16:44] <persia> If not, then I have no idea about your issue.
[16:47] <skrite99> persia, hey, found the problem. hosts file was nuked
[16:48] <persia> skrite99: Aha!  That would do it.
[16:49] <bogeyd6> I am thinking of getting a Ruckus 7731 wireless bridge for two buildings. Does anyone have any thoughts they would like to share if they have used them before?
[16:50] <skrite99> persia, yup, built a hosts file and what do you know? it works.    thanks
[16:51] <philsturgeon> stupid question, but how can i find out what FTP daemon is running?
[16:51] <ivoks> netstatp natp | grep 21
[16:52] <ivoks> s/p//
[16:52] <ivoks> netstat -natp | grep 21
[17:04] <hggdh> smoser: I am using today's UEC image, but the booted instance still shows 2.6.32-20
[17:04]  * hggdh is sorta confused
[17:04] <smoser> doen'st sound right
[17:05] <smoser> $ ls /boot/vmlinuz-2.6.32-21-generic-pae
[17:05] <smoser> t/vmlinuz-2.6.32-21-generic-pae
[17:05] <smoser> that is from today's image on ec2
[17:08] <hggdh> smoser: this is what I see, both from the instance and from cempedak: http://pastebin.ubuntu.com/415637/
[17:09] <smoser> in the instance, cat /etc/cloud/build.info
[17:10] <hggdh> I think I got it -- a stale ISO in the way
[17:20] <hggdh> ooooh this is not god: kernel oops in a instance
[17:22] <jpds> Hmm, god.
[17:25] <hggdh> er. s/god/good/
[17:25] <RoyK> the panic god?
[17:25] <RoyK> which one is that? :)
[17:26]  * RoyK votes for Loki
[17:29] <hggdh> in kernel we trust
[17:43] <smoser> hggdh, did you figure this out above ?
[17:43] <smoser> i think you must have registered an old image
[17:43] <smoser> ah, never mind. i see that you did.
[18:15] <shennyg> hey guys, I am trying to change the port ssh listens on... I edited /etc/ssh/sshd_config and then restarted ssh w/ /etc/init.d/ssh restart and it didn't work
[18:15] <shennyg> I noticed that the ssh I was running is /usr/sbin/sshd am I editing the wrong config file?
[18:15] <shennyg> just mentions my nick and I will be back in here :) thanks
[18:16] <au> can we see your /etc/ssh/sshd_config?
[18:17] <shennyg> au: sure https://gist.github.com/04181e137eb1c447cce9
[18:18] <au> # What ports, IPs and protocols we listen for
[18:18] <au> Port 22
[18:18] <au> change that
[18:19] <shennyg> yeah, I had that on 2222 and it didn't work
[18:19] <shennyg> I just changed it b/c I gave up
[18:19] <au> hehe, then I'm not sure what's wrong
[18:19] <shennyg> I just get connection refused
[18:19] <shennyg> it is an amazon ec2...
[18:19] <au> no idea
[18:19] <au> sorry that I can't help you
[18:20] <shennyg> ok thanks, do you know why when I do a "which ssh" I get /usr/bin/ssh but when I do a ps ax |grep sshd I see the process as /usr/sbin/sshd
[18:21] <au> no idea (I am a bit of a ubuntu newbie)
[18:21] <shennyg> gotcha, thx
[18:21] <au> probably because there are two versions
[18:21] <au> and your init.d script uses the sbin one
[18:22] <shennyg> yeah...
[18:22] <au> that's all I can think of
[18:22] <shennyg> lol, if anyone else has any clues just mention shennyg
[18:22] <au>  2349 ?        Ss     0:00 /usr/sbin/sshd
[18:22] <au> which sshd returns sbin, which ssh returns bin
[18:39] <shennyg> au: I figured it out... I was juggling 4 servers and had two of them mixed up :)
[18:39] <shennyg> thanks for your input
[18:39] <jpds> shennyg: ssh is the client and sshd is the server?
[18:40] <shennyg> gotcha
[18:40] <hggdh> smoser: yes, it was indeed a stale ISO in the path
[18:41] <hggdh> smoser: but I got about 50% success anyways :-(
[18:41] <smoser> hm.. how do you fail ?
[18:41] <hggdh> smoser: many different ways... kernel OOPS (1), euca tracebacks (most) ssh no route to host(same)
[18:42] <hggdh> smoser: I will email you with the results. Repeating, now, wth m1.small
[18:42] <smoser> hggdh, so this is just massive horrible regression ?
[18:43] <hggdh> smoser: sounds like
[18:43] <hggdh> kirkland: ^
[18:43] <smoser> hggdh, verified outside of you and data center ?
[18:44] <kirkland> hggdh: log?
[18:45] <hggdh> kirkland: emailed to you
[18:45] <hggdh> smoser: of course not... I can only test in the data centre
[18:46] <smoser> :)
[18:46] <hggdh> smoser: but I have been running stress tests, I do not know if this is done outside the rig
[18:47] <jcastro> kirkland, https://wiki.ubuntu.com/UbuntuOpenWeek
[18:47] <jcastro> kirkland, I've put you down for two sessions, let me know if this isn't doable
[18:47] <kirkland> jcastro: d00d
[18:47] <jcastro> kirkland, feel free to sell out any other person for the Q+A
[18:47] <kirkland> jcastro: wed may 5 i'm at somehands
[18:48] <jcastro> ok
[18:48] <kirkland> jcastro: move byobu to the monday session and i'm happy to do that one
[18:48] <jcastro> ok
[18:48] <kirkland> jcastro: grab another server dude for the q&a on wednesday
[18:48] <jcastro> ok
[18:48] <kirkland> jcastro: maybe mathiaz ;-)
[18:48] <kirkland> jcastro: thanks for the recruitment
[18:49] <jcastro> amber told me to just volunteer you so I JFDIed you. :D
[19:10] <kirkland> hggdh: still around?
[19:10] <hggdh> kirkland: aye
[19:10] <kirkland> hggdh: have you run these stress tests before (with complete success)?
[19:10] <hggdh> kirkland: yes, up to beta2
[19:10] <kirkland> hggdh: so this is a regression since beta2?
[19:11] <hggdh> kirkland: I am not able to state so with certainty, but looks like
[19:11] <kirkland> hggdh: there has been a major kernel kvm change since beta2
[19:11] <kirkland> hggdh: i was a little nervous about that
[19:12] <hggdh> kirkland: certainly: (1) the kernel oops is new and ugly; (2) I used to have some few instances not tested; (3) we did not have the console output before
[19:12] <hggdh> kirkland: this is why I am reserving a final position here
[19:13] <kirkland> hggdh: and libvirt has changed a few times too
[19:13] <kirkland> hggdh: understood
[19:13] <hggdh> kirkland: not counting that, yes
[19:13] <kirkland> hggdh: what state is the test rig in?
[19:13] <kirkland> hggdh: ideally the same state as the tests you just ran and emailed to me ...
[19:13] <hggdh> kirkland: finishing a 100-instance run on m1.small
[19:13] <kirkland> hggdh: how long (ish) does that take to run?
[19:14] <hggdh> kirkland: indeed it is. We can always restart the cloud, anyways
[19:14] <kirkland> hggdh: here's what i'm thinking ....
[19:14] <kirkland> hggdh: right, i just want quick turnaround on this, without reinstalling
[19:14] <hggdh> kirkland: finished it
[19:14] <kirkland> hggdh: these are all guest oopses, right?
[19:14] <hggdh> kirkland: correct
[19:15] <kirkland> hggdh: dmesg in the host is clean of oopses?
[19:15] <kirkland> hggdh: in the NCs ?
[19:15] <hggdh> kirkland: hum. Checking
[19:16] <kirkland> hggdh: and can you confirm for me the kernel version in both host and guest?
[19:18] <kirkland> jjohansen: around?
[19:18] <jjohansen> yep
[19:18] <hggdh> kirkland: no OOPSes on the NCs, and both NCs and guests are running 2.6.32-21.31
[19:19] <kirkland> jjohansen: we have another OOPs in guest VMs in UEC
[19:19] <kirkland> jjohansen: not sure if it's new or not, actually
[19:19] <hggdh> kirkland: but many audit messages with denied for libvirt
[19:19] <kirkland> jjohansen: but its an oops that we apparently didn't have in beta2
[19:19] <jjohansen> kirkland: paste bin or bug#?
[19:19] <jjohansen> oh!
[19:20] <kirkland> jjohansen: pasting
[19:20] <jjohansen> hggdh: so 32-21.31 cleanedup your issues then
[19:20] <hggdh> kirkland: oooh this is good. apparmor is in enforce mode for libvirt
[19:20] <hggdh> jjohansen: not on my laptop, no
[19:21] <jjohansen> okay, one can always hope /me reading backscroll
[19:22] <kirkland> jjohansen: scp chinstrap.canonical.com:~kirkland/Results_from_a_stress_test_using_the_ISO_from_20100416.bz2 .
[19:22] <kirkland> hggdh: i put your results there for jjohansen, internal server
[19:24] <smoser> hggdh, did you see this sort of failure when you were (mistakingly) running guests from yesterday ?
[19:26] <kirkland> hggdh: i have another suggestion ....
[19:26] <hggdh> smoser: yes
[19:26] <hggdh> kirkland: shoot
[19:26] <kirkland> hggdh: can you register a karmic 9.10 image in your uec, and run your stress test against that?
[19:26] <kirkland> hggdh: let's take this all the way back to that released image
[19:26] <hggdh> kirkland: no prob
[19:27] <kirkland> hggdh: take the image/instance out of the equation
[19:27] <kirkland> hggdh: and see if we see issues there
[19:27] <hggdh> kirkland: roj. It takes about 20 minutes for a 100-instance tun
[19:27] <hggdh> run
[19:27] <kirkland> smoser: what's your confidence in the 9.10 uec-image as a guest?  pretty rock solid?
[19:27] <kirkland> hggdh: great, then we can iterate over that a few times per hour
[19:28] <smoser> it should be, yes.
[19:28] <kirkland> smoser: ie, if we need to track a problem down to being host or guest, could we lean on the 9.10 uec image as a guest, and with a high degree of confidence, expect that image to work?
[19:28] <kirkland> smoser: it's been my experience so far, anyway ^^^
[19:28] <smoser> well, i dont think you are that lucky really
[19:28] <kirkland> smoser: ?
[19:28] <smoser> its certainlyi something worth trying
[19:28] <kirkland> smoser: care to explain that?
[19:29] <kirkland> hggdh: okay, i think we have an approach; let's run the stress test with the karmic image and see what happens
[19:29] <smoser> well, i guess i retract the comment.
[19:29] <kirkland> jjohansen: were you able to grab that log?
[19:29] <smoser> but it would be more informative to back off to beta2 guest image
[19:30] <smoser> as that is "known working" and would give you a bisect path on lucid
[19:30] <kirkland> smoser: right, that's my next suggestion, pending the results of using 9.10 as the baseline guest
[19:30] <kirkland> hggdh: got that?  after running stress test against 9.10, then bump to beta2
[19:30] <jjohansen> kirkland: I copied the file, I haven't unecoded the log yet
[19:30] <kirkland> hggdh: let's compare both of those against today's image
[19:31] <hggdh> kirkland: just registered current karmic, will start run now
[19:31] <kirkland> hggdh: rock on
[19:31] <smoser> for what its worth, i've run a couple guests in the 0413 range that were fine. i didn't stress, but 50% failure rate doesn't require stress to run into
[19:31] <hggdh> kirkland: then, back to beta2.
[19:31] <kirkland> hggdh: you got it
[19:31] <kirkland> hggdh: poke me as you get results
[19:32] <hggdh> kirkland: will do
[19:33] <hggdh> kirkland: backtracking to beta2 will take significantly longer, though -- I will have to reinstall all
[19:34] <jdstrand> hggdh: hey, what is it that apparmor is denying?
[19:34] <jdstrand> hggdh: (in libvirt)
[19:34] <jdstrand> hggdh: and do you have ubuntu24 of libvirt?
[19:35] <hggdh> jdstrand: just a sec
[19:35] <hggdh> jdstrand: libvirt\* is 0.7.5-5ubuntu24
[19:36] <hggdh> jdstrand:  type=1503 audit(1271441213.099:434):  operation="open" pid=31521 parent=1333 profile="/usr/lib/libvirt/virt-aa-helper" reques
[19:36] <hggdh> ted_mask="::r" denied_mask="::r" fsuid=0 ouid=107 name="/var/lib/eucalyptus/instances/admin/i-465B0851/disk"
[19:36] <jdstrand> hmmm
[19:36] <smoser> hggdh, no wait.
[19:36] <jdstrand> hggdh: those are non-fatal btw, but I'd rather get rid of them
[19:36] <smoser> do not (please) back the system back to beta2
[19:36] <smoser> lets do one component at a time
[19:36] <hggdh> smoser: roger
[19:36] <jdstrand> hggdh, smoser: what is the pattern for disks in euca?
[19:37] <hggdh> smoser: right now it is running on karmic UEC current
[19:37] <smoser> pattern ?
[19:37] <jdstrand> yes-- the name of the disks? eg disk.img, disk.gcow2, just plain 'disk'
[19:37] <smoser> hggdh, you mean lucid-current euca running karmic guest, right?
[19:37] <smoser> jdstrand, let me look
[19:37] <hggdh> smoser correct
[19:37] <smoser> i actually think it is 'disk', but i will have to verify
[19:38] <jdstrand> I'd like to get rid of those apparmor denials-- they will confuse people
[19:38] <jdstrand> oh, I could acutally just add /var/lib/eucalyptus/instances/...
[19:39] <smoser> jdstrand,
[19:39] <smoser>         <disk type='file'>
[19:39] <smoser>             <source file='BASEPATH/disk'/>
[19:39] <smoser>             <target dev='sda'/>
[19:39] <smoser>         </disk>
[19:39] <jdstrand> smoser: cool, thanks
[19:39] <jdstrand> I'll get that fixed up
[19:39] <jdstrand> (it is just the backingstore checking-- totally non-fatal, but confusing for people to see in the kern.log)
[19:41] <smoser> jdstrand, let me make absolute sure, launching an instance and then will look at virsh dumpxml
[19:41] <jjohansen> kirkland: I don't think I got the entire log
[19:42] <smoser> jdstrand, xml: http://pastebin.com/3QYBaPGW
[19:42] <jdstrand> well, between what you just said and hggdh's denied message, I should have enough. if you want to chek feel free, but I think I've got it
[19:42] <jdstrand> cool, yeah
[19:42] <jdstrand> smoser: you were fast! :)
[19:42] <jdstrand> fixing now
[19:42] <jjohansen> kirkland: I get an "invalid input" when decoding, though I do get a fairly decent size decoded log file
[19:45] <hggdh> jjohansen: the whole log is available at http://people.canonical.com/~cerdea
[19:45] <jjohansen> hggdh: thanks
[19:45] <hggdh> jjohansen: it is the single* file
[19:46] <jjohansen> hrmm, so I pretty much got the whole log, I'm not sure why the decode is complaining
[19:46] <smoser> jdstrand, for completeness : http://pastebin.com/VkfrCL8k
[19:46] <hggdh> ugh. the karmic UEc does not even start
[19:46] <smoser> that one has a second disk (ebs volume) attached
[19:47] <smoser> hggdh, console messages?
[19:47] <jjohansen> hggdh: do we have a bug open for this yet?
[19:48] <hggdh> jjohansen: no, not yet
[19:48] <hggdh> smoser: heh. The karmic image needs 3G of disk for small, and lucid has 2...
[19:48] <hggdh> restarting the whole thing, using c1.medium
[19:49] <smoser> hggdh, yes, known issue
[19:50] <hggdh> smoser: known, and forgotten my me :-(
[19:51] <hggdh> jjohansen: I am opening a new bug now
[19:52] <hggdh> jdstrand: thank you for the bug
[19:53] <jdstrand> hggdh: sure. it'll be fixed momentarily
[19:57] <hggdh> jjohansen: bug 564924
[19:58] <jjohansen> hggdh: thanks
[19:59] <hggdh> jjohansen: just attached the dmesg
[20:05] <hggdh> smoser: I am considering cycling all servers, and redoing the karmic UEC run. Success rate is at .08 right now
[20:05] <hggdh> kirkland: ^
[20:05] <kirkland> hggdh: ?
[20:06] <smoser> here i have success rate of 100% reaching guests of lucid 20100416 (granted i've launched 5 of them)
[20:06] <hggdh> kirkland: just in case. It seems most instances are not reachable by ssh
[20:06] <kirkland> hggdh: any oopses?
[20:07] <hggdh> smoser: kirkland yes, at least one
[20:07] <smoser> i'm at 1.6.2-0ubuntu29 in eucalyptus. i tihnk i'm booted into andy's (apw) test kernel from a wekek ago or so.
[20:07] <kirkland> hggdh: hmm, okay so the karmic kernel is oopsing too?
[20:07] <smoser> but as a reference everything seems generall working here.
[20:08] <hggdh> kirkland: correct
[20:08] <hggdh> smoser: darn! So what is different?
[20:08] <tesseracter> running 9.10, installed memcached, tried to start it with start memcached, and it didnt have a clue what I was talking about.
[20:09] <kirkland> hggdh: okay, next, let's replace the host kernel on the NCs
[20:09] <kirkland> hggdh: with the beta2 kernel
[20:09] <kirkland> hggdh: can you do that, and rerun your tests against the karmic image?
[20:10] <smoser> i can dis-upgrade and see if the world fails for me
[20:10] <hggdh> kirkland: only on the NCs?
[20:10] <kirkland> hggdh: yeah
[20:10] <kirkland> hggdh: shouldn't matter elsewhere
[20:11] <hggdh> kirkland: this will take a while. I will have to find out how to get it done remotely
[20:11] <kirkland> ?
[20:11] <kirkland> hggdh: can you scp to the nc's?
[20:11] <kirkland> hggdh: it will just take a couple of hops
[20:11] <hggdh> kirkland: what you propose?
[20:12] <kirkland> hggdh: 1) mount -o loop the beta2 server iso, grab the server kernel
[20:12] <kirkland> hggdh: 2) scp that to the nc's
[20:12] <kirkland> hggdh: 3) dpkg -i it
[20:12] <kirkland> hggdh: you might have to edit /etc/default/grub to make sure that kernel boots
[20:12] <hggdh> kirkland: oh, you just want the kernel
[20:12] <kirkland> hggdh: yeah
[20:13] <kirkland> hggdh: let's start there
[20:13]  * hggdh was thinking of a whole install...
[20:13] <kirkland> hggdh: no no no
[20:13] <kirkland> hggdh: let's downgrade piece by piece
[20:13] <kirkland> hggdh: a) kernel, b) qemu-kvm, c) libvirt
[20:13] <hggdh> kirkland: k. Give me 15m to grab a lunch
[20:13] <kirkland> hggdh: ack
[20:14] <leonel> ScottK: Thank you for your work on clamav for keep the updated version and keep my servers running with the clamav change yesterday  Thank you very much
[20:14] <ScottK> leonel: You're welcome.
[20:14] <ScottK> I don't see any sign that other distros were similarly prepared.
[20:15] <ScottK> #clamav has been totally insane.
[20:16] <leonel> ScottK: even twitter and all arround is crazy but  ubuntu servers just working   THANK YOU !
[20:18] <tesseracter> anyone? i want memcached to start and stay running - was gunna use upstart, but i dont see much on google.
[20:19] <hggdh> kirkland: er. Where can I find a beta2 ISO for the server?
[20:19] <kirkland> releases.ubuntu.com
[20:19] <kirkland> hggdh: or there should be one in the lab
[20:19] <hggdh> dammit
[20:19] <kirkland> hggdh: let me grab the kernel for you
[20:21]  * RoAkSoAx feels that this release cycle has been more problematic than others
[20:22] <kirkland> hggdh: okay, here's the quickets way to get this
[20:22] <kirkland> hggdh: start here: http://releases.ubuntu.com/10.04/
[20:23] <kirkland> hggdh: where you can find this: http://releases.ubuntu.com/10.04/ubuntu-10.04-beta2-server-amd64.list
[20:23] <JanC> tesseracter: it uses a sysvinit-style init script by default, and you might have to configure it in /etc/memcached.conf and then enable it in /etc/default/memcached
[20:23] <kirkland> hggdh: in that file, you should see: /pool/main/l/linux-meta/linux-image-server_2.6.32.19.20_amd64.deb
[20:23] <kirkland> hggdh: then go to Launchpad to find that deb
[20:24] <kirkland> hggdh: https://edge.launchpad.net/ubuntu/+source/linux/+publishinghistory
[20:26] <kirkland> hggdh: a few clicks later, i find http://launchpadlibrarian.net/42795268/linux-image-2.6.32-19-server_2.6.32-19.28_amd64.deb
[20:27] <hggdh> kirkland: removed :-(
[20:27] <kirkland> hggdh: i can download http://launchpadlibrarian.net/42795268/linux-image-2.6.32-19-server_2.6.32-19.28_amd64.deb
[20:28] <smoser> kirkland, so your plan here is to back out to the older kernel and see if that fixes the guests ?
[20:29] <smoser> older kernel on host, that is.
[20:29] <hggdh> kirkland: my mistake, fat-fingered the keyboard
[20:29] <kirkland> smoser: that's the current plan
[20:29] <kirkland> smoser: then back down kvm, libvirt, and eucalyptus (in that order)
[20:29] <kirkland> smoser: until we get back to the state at which everything worked
[20:30] <kirkland> smoser: the kvm changes in the kernel were pretty massive
[20:30] <smoser> well, for some point of reference:
[20:30] <smoser> http://pastebin.com/KsQXF7kx
[20:30] <kirkland> smoser: i acked them after a sniff test on my local hardware
[20:30] <smoser> is currently working as well as anything for me.
[20:31] <kirkland> smoser: running the security kernel?
[20:31] <kirkland> smoser: can you apt-get dist-upgrade and see if everythign is still working?
[20:31] <smoser> uname -r shows: 2.6.32-21-server
[20:32] <smoser> so, yeah, 2.6.32-21.31~security201004122115
[20:35] <leonel> what ?
[20:35] <smoser> kirkland, i'm in the process. i'm waiting on
[20:35] <smoser> Setting up eucalyptus-walrus (1.6.2-0ubuntu30)
[20:35] <smoser> from the dist-upgrade of the CC
[20:35] <smoser> i really should have just gotten the new kernel
[20:40] <smoser> hggdh, for the record, you can get karmic images to fit into m1.small fairly easily, just register them as:
[20:41] <smoser> uec-publish-tarball --resize 1408M karmic-*.tar.gz my-karmic
[20:44] <smoser> well, kirkland i'm 3 for 3 in ssh reaches at the moment.
[20:44] <smoser> remember, my hardware isn't necissarily blazing
[20:44] <kirkland> smoser: bzr branch lp:~mathiaz/+junk/uec-testing-scripts
[20:45] <kirkland> smoser: edit config_single.yaml
[20:45] <kirkland> smoser: i changed id to my emi
[20:45] <kirkland> smoser: type: to m1.small
[20:45] <kirkland> smoser: max_instances_to_start: to my max instances
[20:46] <kirkland> smoser: and cred_dir: to my .euca dir
[20:47] <kirkland> smoser: or use a symlink
[20:48] <hggdh> kirkland: rebooting NCs
[20:49] <osmosis_> kirkland, hi dustin. I am the one who emailed you about volunteering.
[20:52] <kirkland> smoser: did you get all of that?
[20:52] <oru_work> how can I install 7zip, i tred apt-get install 7zip which didn't work :)
[20:52] <kirkland> smoser: my wifi seems to be a little lossy
[20:53] <kirkland> hggdh: i'm running latest/greatest lucid cloud here
[20:53] <kirkland> hggdh: and i'm stress testing via mathiaz' test suite
[20:53] <hggdh> kirkland: good
[20:53] <osmosis_> oru_work, apt-cache search 7zip
[20:54] <hggdh> kirkland: running another 100 instances on 2.6.32-19
[20:56] <smoser> kirkland, running now
[20:57] <smoser> kirkland, is this output written somewhere?
[20:58] <kirkland> smoser: stdout
[20:58] <smoser> oh. i see, it finished.
[20:58] <kirkland> smoser: if you're in screen, you can write your scrollback buffer to a file
[20:58] <smoser> well, i was 2 for 2 on it
[20:58] <barbarella> oru_work:7za
[20:59] <smoser> can i run multiple runs with a single (admin) user ?
[20:59] <kirkland> smoser: yes
[20:59] <kirkland> smoser: bump up max_instances_to_start:
[20:59] <kirkland> smoser: i just ran that with 60
[20:59] <kirkland> smoser: my cloud can accommodate 6 at a time
[20:59] <smoser> so it starts them serially ?
[20:59] <kirkland> smoser: yeah
[20:59] <smoser> or does it check the available
[20:59] <kirkland> smoser: checks available
[20:59] <kirkland> smoser: ask mathiaz for the algorithm
[21:00] <kirkland> smoser: or read the code
[21:00] <kirkland> smoser: but he says it monitors availability
[21:00] <hggdh> smoser, kirkland just edit the config_single.yaml, and uncomment the log file at the beginning
[21:01] <smoser> i dont seen in code that it launches more than one
[21:01] <kirkland> hggdh: heh, well there you go, smoser ^
[21:01] <hggdh> it is assyncrhonous
[21:01] <hggdh> and it does check availability
[21:01] <smoser> it does not pass '-n' to euca-run-instances
[21:01] <hggdh> mathiaz did a good job there ;-)
[21:01] <smoser> oh. i see.
[21:02] <smoser> but it will never invoke --instance-count=2
[21:02] <smoser> or N
[21:02] <smoser> that would be something to add, which would possibly stress in a different way
[21:02] <kirkland> smoser: hggdh: i gotta drop, but i'm going to run this over the weekend on my local hardware with some really, really, really large number
[21:03] <smoser> kirkland, are you seeing any errors ?
[21:03] <smoser> i see 100% success here.
[21:03] <kirkland> smoser: hggdh: i'll setup a cronjob to rsync that output file to my people.canonical.com pub html
[21:03] <kirkland> smoser: 100% success here too
[21:03] <kirkland> ubuntu@beagle:~$ grep -i oops out
[21:03] <kirkland> ubuntu@beagle:~$ grep -i error out
[21:03] <kirkland> ubuntu@beagle:~$
[21:04] <kirkland> smoser: success rate is 100%
[21:04] <kirkland> smoser: this is up-to-date lucid
[21:04] <smoser> same here now
[21:04] <kirkland> image form this week
[21:04] <smoser> well, yo ushould be using image from today
[21:04] <smoser> but i am
[21:04] <hggdh> kirkland, smoser not 100% here, but no console output
[21:05] <smoser> no console output where ?
[21:05] <kirkland> gotta run
[21:05] <hggdh> we changed the uec_test.py to output console on error
[21:05] <hggdh> kirkland: k
[21:07] <oru_work> which command is used to upgrade from 8.10 to 9.10 ?
[21:08] <guntbert> !upgrade | oru_work
[21:16] <hggdh> smoser: how did it go? How many total sessions?
[21:17] <smoser> well, my first run '2'
[21:17] <hggdh> oh
[21:17] <smoser> as i understood that to be concurrent
[21:17] <hggdh> try 100
[21:17] <jdstrand> fyi, I uploaded ubuntu25 for libvirt... needs to be approved of course
[21:17] <smoser> yeah, its doing 100 now
[21:17] <hggdh> jdstrand: thanks, I hope it is. These messages really suggest failures
[21:18] <hggdh> smoser: I got 32 succeeded, 68 not tested
[21:18] <jdstrand> hggdh: I'm confident it will be. I understand the confusion it could cause
[21:18] <hggdh> jdstrand: IOU
[21:18] <jdstrand> hggdh: thanks for noticing it! :)
[21:18] <hggdh> heh
[21:19] <ruben23> hi guys when i try to type in this cd /etc/network/interfaces ----> i get error  no such file or directory..
[21:19] <ruben23> im on root- i type in sudo bash
[21:19] <ruben23> any ideas
[21:19] <hggdh> ruben23: /etc/network/interfaces is a file, not a directory
[21:20] <smoser> hggdh, wait.
[21:20] <smoser> so now everything is happy for you?
[21:20] <hggdh> ruben23: and, BTW, it is better to run 'sudo vi' than 'sudo bash' -- safer to only be root when needed
[21:20] <smoser> you reverted to old kernel on nodes ?
[21:20] <ruben23> hggdh: even i do- i do this cd /etc/network/---same error
[21:20] <hggdh> smoser: yes, I am running kernel 2.6.32-19. No OOPSes so far
[21:20] <smoser> this is strange to me.
[21:21] <ruben23> hggdh: also cd /etc/-----no such file or directory
[21:21] <ruben23> what should be the problem iwth this
[21:21] <hggdh> ruben23: are you sure you are running Linux? ;-) if you do not have /etc, your system should be hosed (or on its way to)
[21:22] <ruben23>  hggdh: yeah im running ubuntu-all boots up and said OK. then i login my usernmae
[21:23] <hggdh> smoser: and you are all in current, including UEC?
[21:23] <hggdh> ruben23: and?
[21:24] <smoser> current as of my mirror
[21:24] <smoser> but let me check its currency
[21:24] <ruben23> then type in sudo bash
[21:24] <ruben23> input hte password then i type in cd /etc/---no such file or directory
[21:26] <hggdh> ruben23: weird. What version of Ubuntu, and is this a server edition?
[21:27] <guntbert> ruben23: no need for sudo bash  -- if you *really really* need a root shell you better use sudo -i
[21:27] <incorrect> how does Eucalyptus compare to kvm?
[21:27] <hggdh> incorrect: eucalyptus *uses* kvm
[21:27] <incorrect> hggdh, i had a feeling you were going to say that
[21:27] <incorrect> so Eucalyptus can run my windows vms?
[21:29] <gzmask> hello guys, question forcha: at https://help.ubuntu.com/community/UEC/CDInstall step 7 where I got lost: should I run the ssh key generation at node or at cluster?
[21:29] <ruben23> guntbert: im root
[21:30] <guntbert> !root | ruben23
[21:30] <leonel> jcastro: pong
[21:30] <hggdh> incorrect: if you want a cloud, yes
[21:30] <guntbert> ruben23: then why do you need sudo? (if you are root already)
[21:30] <ruben23> still i can do cd /etc/ or cd /home/ ---> no files or directory
[21:31] <hggdh> ruben23: what version of Ubuntu?
[21:31] <hggdh> second time I ask
[21:31] <ruben23> 8.04 server
[21:32] <incorrect> hggdh, what i want it to be able to run up vm's to compile my code, get a build and then shutdown, hudson seems to have plugins to do the control for me, i just don't see what the difference between Eucalyptus and just kvm is
[21:34] <ruben23> hggdh: ubuntu-server 8.04 LTS
[21:35] <hggdh> incorrect: eucalyptus will allow you to build a cloud of virtual machines (like Amazon EC2). KVM will allow you to run a virtual image
[21:36] <hggdh> ruben23: and this server was working perfectly before?
[21:36] <binBASH> smoser: My vms get ips assigned now!
[21:36] <incorrect> never used EC2
[21:36] <ruben23> hggdh: this is a new install
[21:36] <ruben23> fresh install
[21:36] <binBASH> just installed a local dhcp server on each node, where I specify the ip range.
[21:38] <ruben23> what could be wrong where i can download the officcial copy of ubuntu-server 8.04 torrent..?
[21:40] <hggdh> ruben23: try to run 'echo /etc'. I am curious about what you see
[21:41] <ruben23> hggdh: i guess i need to re install
[21:41] <ruben23> where can i download officeial copy of ubuntu server 8.04- torrent file
[21:42] <hggdh> smoser: the plot thickens. Now I see eucalyptus-cloud using 790M of memory
[21:42] <smoser> well, thats not too big a surprise
[21:42] <smoser> it is a massive java stack
[21:42] <hggdh> smoser: sorry. 749M
[21:42] <hggdh> and I am getting out-of-memory there...
[21:43] <ScottK> Massive Java stack is a bit redundant
[21:43] <hggdh>  java.lang.ClassCastException: java.lang.OutOfMemoryError cannot be cast to java.lang.RuntimeExce
[21:44] <hggdh> heh
[21:45] <osmosis_> doest ubuntu have an equivalent to  /etc/init.d/iptables save   ?
[21:45] <osmosis_> ruben23, google  ubuntu torrents
[21:48] <smoser> hggdh, ok, thats bad news.
[21:48] <smoser> hggdh, ok, so let me see if i understand where you are
[21:48] <smoser> right now, we had massive complete failure
[21:49] <smoser> then we backedleveled the host kernel on the node controllers
[21:49] <smoser> then we have much better results
[21:49] <smoser> is that correct ?
[21:52] <resno> is it possible to use the credentials a user uses to logon to log into their samba account?
[21:52] <tesseracter> so am I right to assume that email sending is not set up by default on 9.10 server(in this case, on amazon ec2)?
[21:53] <smoser> hggdh, ^^
[21:55] <hggdh> smoser: sounds correct, but I have not had any oops on 2.6.32-19
[21:56] <hggdh> smoser: also, after I bounced all of cloud/cluster/sc, I am now able to run imagtes again
[21:56] <smoser> jjohansen, it appears we're seeing regression on our data center hardware with the newer kernel
[21:57] <smoser> hggdh, i'd like to see verification that switching the NC to the current lucid kernel will cause catostrophic failure
[21:57] <smoser> note, that kirkland and I are running tests on our more modest hardware and do not see these issues.
[21:58] <hggdh> smoser: after I can complete a stress run without other issues, I will revert to current kernel, and try again
[21:58] <hggdh> smoser: I follow you. It bothers me that you and Dustin cannot get the same failures
[21:58] <resno> is there a way when i login to my machine when i access samba on the server, it will know my username and password?
[21:59] <resno> and allow me access to my directory only, etc
[22:02] <smoser> hggdh, i see 2 failures in 69 at the moment.
[22:03] <smoser> for which i just opened bug 565018
[22:04] <smoser> hggdh, yeah, i dont like that we don't see the failure either, but i have a puny little single cpu system
[22:04] <hggdh> you got access denied? That's new, I think
[22:04] <smoser> Intel(R) Core(TM)2 CPU          6300  @ 1.86GHz
[22:04] <smoser> and dustin is on laptop class hardware
[22:04] <smoser> i think it turns access denied
[22:04] <hggdh> yeah. This may be related to concurrency level
[22:04] <gzmask> on UEC cluster controller, when I am starting my VM, I got FinishedVerify not enough resources available: addresses error. How do I assign ip address to VMs? How do I register my VMs Mac address so that my DHCP server knows what to do?
[22:07] <smoser> hggdh, i updated it. it wasn't access denied.
[22:07] <smoser> hggdh, well, yeah, thats my thoughts is that its racy kernel multi cpu. i dont know. but thats all i can explain.
[22:07] <smoser> hggdh, i have to run. i'll check back in later.
[22:09] <hggdh> smoser: k
[22:09] <hggdh> smoser: I will revert the NC back to current kernel & run it again
[22:12] <guntbert> resno: sorry - my experience with samba/windows network is minimal - I couldn't tell
[22:12] <resno> ah ok
[22:14] <resno> guntbert: lol
[22:14] <guntbert> resno: but have a look at man smbpasswd
[22:16] <guntbert> resno: and if you want something to read : https://help.ubuntu.com/9.04/serverguide/C/windows-networking.html
[22:32] <binBASH> smoser: How to connect to the uec image from ubuntu?
[22:34] <gzmask> can I run Cloud controller, cluster controller and node controller on the same machine?
[22:41] <hggdh> gzmask: yes
[22:48] <CodPair> Does anyone know of any issues with the partition formatter in ubuntu server 9.10? Mine formated a few partitions but froze at 33% when formatting the / partition. Is this common?
[22:50] <CodPair> Specifically during installation.
[22:52] <gzmask> I use Gparted from desktop version to partition and format. then select "manul" in server version to use the formatted drive
[22:53] <JanC> CodPair: it's not common, but can happen...
[22:56] <Ng> how come we have an /etc/default/libvirt-bin when it's not used?
[23:01] <osmosis> Ng, what version?
[23:02] <Ng> osmosis: lucid
[23:07] <CodPair> How would you suggest I proceed with installation?
[23:09] <CodPair> Could my RAID card be incompatible with the partitioner?
[23:18] <JanC> CodPair: maybe
[23:20] <CodPair> JanC: I just tried manual partitioning as ext3 and it jumped right to 33% and froze
[23:21] <JanC> I guess it's some fake raid thing  ;)
[23:25]  * CodPair face-palms
[23:26] <CodPair> i figured it out
[23:26] <CodPair> i accidentally put the dead drive in the server.
[23:56] <orudie> should i upgrade 9.04 to 9.10 ?