[02:20] <lifeless> SpamapS: we're finally using rabbit btw :)
[02:21] <SpamapS> lifeless: \o/
[02:21] <SpamapS> lifeless: how's it going?
[02:21] <lifeless> SpamapS: slow progress, good performance
[02:21] <SpamapS> lifeless: I've heard a few grumbles from rabbit users about it becoming a SPoF even with replication/failover
[02:22] <lifeless> it is
[02:22] <lifeless> so e.g. https://code.launchpad.net/~lifeless/python-oops-datedir-repo/less-rsync/+merge/80861
[02:22] <lifeless> where we have a backup system
[02:23] <SpamapS> lifeless: I missed your answer if you gave it (or I forgot to ask) are you in Orlando or stil attending events virtually?
[02:23] <lifeless> virt
[02:23] <lifeless> SpamapS: OTOH, if you get an oops on qastaging, it gets injected into oops-tools in realtime
[02:23] <lifeless> SpamapS: faster than the time to get the page to you :)
[02:24] <lifeless> SpamapS: prod will have that come online late this week / next week I think
[02:46] <shang> RoAkSoAx: no, I am not in UDS this time
[02:47] <shang> RoAkSoAx: but yes, I am would like to do some demo on orchestra, but I am having some issues
[03:21] <shang> RoAkSoAx: https://answers.launchpad.net/ubuntu/+source/orchestra/+question/177059
[03:36] <hydromet> hello - if I want to upgrade a 9.10 amd64 server to an 11.04 amd64 server, do I first need to step to a 10.04 (or 10.10) amd64 server or is it possible to do a direct jump from 9.10 to 11.04?
[03:37] <lifeless> hydromet: the rule is that LTS->LTS is possible, and anything->anything+1 is possible.
[03:37] <philipballew> hydromet, you have to upgrade to each one
[03:38] <hydromet> lifeless thanks, that's very useful to know (hence a good reason to use LTS)
[03:39] <lifeless> hydromet: 10.04 is an LTS
[03:39] <hydromet> philipballew: thanks for reinforcing the rule ... so it looks like I'll need to stair step from 9.10 to 10.04 to 10.10 to 11.04. fun! :-)
[03:41] <SpamapS> just depends on how much you value your time.
[03:41] <hydromet> lifeless thanks ... looks like the next LTS (12.04) is quite some time away yet
[03:42] <lifeless> yes
[03:42] <lifeless> 5 months
[03:42] <mgw> How can I configure cobbler to *not* use its local mirrors (that haven't been populated yet)?
[03:42]  * SpamapS heads to the airport to fly to UDS.. woot. :)
[03:42] <lifeless> enjoyah
[03:43] <SpamapS> mgw: by default it uses archive.ubuntu.com through a squid proxy
[03:43]  * SpamapS drive-by's mgw with an answer then goes <POOF!>
[03:45] <mgw> hmm, odd… that's what I had sort of thought, but when I try to boot a new server with koan, i get this: HTTP hasFile: didn't find http://X.X.X.X/cblr/repo_mirror/oneiric-x86_64/.treeinfo: HTTP Error 404: Not Found
[04:02] <shang> SpamapS: a quick question about the orchestra+juju to deploy openstack
[04:03] <shang> SpamapS: I have the orchestra setup (successfully deploy server from it with the profile), but when I run juju bootstrap,  I got the error: Could not find any Cobbler systems marked as available and configured for network boot.
[04:03] <shang> 2011-11-01 11:49:16,389 ERROR Could not find any Cobbler systems marked as available and configured for network boot.
[04:03] <shang> SpamapS: from the profile, I re-enabled the netboot for one of the machine, but I stil l got that error
[04:26] <lucke> hey guys, we recently had a compromise on our server, a web site was hacked and they installed a porno forum on the site, it was taking up tons of bandwidth, I have since fixed the issue and removed the stuff they had there, I am still getting a ton of hits to that form they installed, what can I do to stop the hits?
[04:43] <qman__> lucke, in short, you can't
[04:44] <qman__> if you want to continue receiving requests from the general internet as a whole, there's no way to prevent those from coming in
[04:45] <qman__> you can rate limit connections if you're being brute forced, and you can remove the offending files/sites altogether (and should have already)
[04:45] <qman__> but you can't stop requests
[04:46] <lucke> Ah ok.
[04:46] <lucke> thanks qman__ I removed those a long time ago, i just keep seeing annoying connections coming into trying to access that forum, by the hundreds.
[04:47] <lucke> every night at this time of the night, I think they had some sort of scraper running or something
[04:47] <qman__> if they're coming from the same IP, you can block that IP
[04:47] <lucke> yea, I wish :(
[04:47] <qman__> but the requests will still come in, they'll just be blocked at the firewall instead of your web server sending a 404
[04:48] <qman__> not much difference between the two
[04:48] <lucke> yea, no big deal I guess, eventually they should go away.
[04:48] <lucke> should i so a permanent redirect (302) to somewhere or just leave it 404?
[04:48] <lucke> s/so/do
[04:49] <qman__> I don't know if it would have any real effect
[04:49] <qman__> it'd only help if it was a legit search engine or something
[04:49] <qman__> as in, some program that respects that type of thing instead of blindly requesting
[04:50] <lucke> ya
[04:50] <lucke> ill just leave it I guess
[04:50] <lucke> thanks
[05:10] <mgw> in regard to cobbler… any ideas why squid seems to not be proxying?  I'm getting this error when invoking koan:  HTTP hasFile: didn't find http://X.X.X.X/cblr/repo_mirror/oneiric-x86_64/: HTTP Error 404: Not Found
[07:42] <Error404NotFound> What options other than compiling from source do i have to upgrade apache from 2.2.14 to 2.2.21?
[07:51] <Error404NotFound> If i use precise's main repo to just upgrade apache2, and later delete it, on my 10.10 server is it likely to cause package upgrade issues?
[10:02] <freakynl> Hi, is anyone aware of a comparison between the new in-kernel iscsi stack (lio iirc) vs iet (iscsi enterprise target)? New being the one in the 3.1 kernel btw
[10:07] <ikonia> don't know of any ubuntu version using 3.1 currently
[10:16]  * _ruben is likely to stick to the scst stack
[10:53] <_ruben> aww bugger .. keepalived's homepage seems "broken" .. only shows a login form :/
[11:21] <iclebyte> anyone found heartbeat to be unreliable for DRBD / IP Addr resource management?
[11:33] <_ruben> iclebyte: plain heartbeat (ancient) or combined with pacemaker/corosync/etc ?
[11:39] <freakynl> _ruben: any reason in particular? we're currently on iet and kinda curious on how interesting it is to change
[11:40] <_ruben> freakynl: i switched quite some time ago .. iet did have some issues which i dont recall anymore though .. also scst is said to have better performance (depending on config) .. and scst's config file is much cleaner imo
[11:41] <_ruben> never really looked back at iet, tho scst did base its iscsi part on iet
[12:39] <freakynl> _ruben: hmm scst is now dropped from kernel in favor of lio... then again, lio is sgst based/compatible (which is scsi stack in kernel) and scst has it's own scsi stack (which unfortunately isn't ever going to make it into the kernel by the looks of it (it has/had some more advanced options and fancy stuff)
[12:40] <_ruben> freakynl: the lio vs scst vs whatever inclusion in mainline kernels was a nasty political battle from what i've seen about it .. scst compiles just fine as a dkms module, you only lack a few performance enhancements which require actual kernel patches
[12:41] <_ruben> isn't sgst mostly userland? or am i mixing stacks up?
[12:42] <SpamapS> shang: did you figure out your juju issues?
[12:42] <shang> SpamapS: no, unfortunately...
[12:43] <shang> SpamapS: but the issues is a bit different now. I can't run bootstrap the
[12:43] <shang> SpamapS: I have the first system deployed with oneiricx86_64-juju profile
[12:44] <mgw> SpamapS : I'm still having trouble with squid (or something related)
[12:44] <shang> but when I run juju bootstrap, it says the system is not ready
[12:45] <SpamapS> shang: pastebin? also run it with 'juju -v bootstrap'
[12:47] <shang> SpamapS: http://paste.ubuntu.com/725249/
[12:48] <shang> SpamapS: is there a difference to have the bootstrap node using the oneiricx86_64 profile instead of the oneiricx86_64-juju profile?
[12:52] <SpamapS> shang: can you ssh to the machine that was started as part of the bootstrap?
[12:55] <shang> SpamapS: yes, using password authentication
[12:56] <SpamapS> shang: but not the key? check the console.. I bet cloud-init had problems.
[12:58] <shang> SpamapS: so we should just login automatically, right?
[13:03] <shang> SpamapS: and here is the cloud-init.log: http://paste.ubuntu.com/725269/
[13:04] <SpamapS> shang: uds sessions have started, will get back to you later today
[13:04] <shang> SpamapS: ok, thanks
[13:06] <mgw> Anybody know why I'm getting a 404 rather than hitting the squid proxy: https://gist.github.com/2dfbcf76c6fd5d3b1c52
[13:09] <RoAkSoAx> shang: yeah there's some issues with cobbler-enlist, could you ping Daviey or adam_g as they definitely know what's failing
[13:10] <shang> RoAkSoAx: ah... if u get the chance, can you help me take a look at the cloud-init ? http://paste.ubuntu.com/725269/
[13:11] <RoAkSoAx> shang: sure, what's up with it, I don;t see anything significant in the log
[13:11] <shang> RoAkSoAx: I tried to use juju bootstrap, but I can't get it working: http://paste.ubuntu.com/725249/
[13:12] <RoAkSoAx> shang: could you also post /var/log/cloud-init-(output i think).log from the bootstrap node
[13:13] <RoAkSoAx> shang: then do: sudo cobbler system getks --name <what ever name of the node selected to be bootstrap>
[13:14] <mgw> RoAkSoAx: would you have a second to look at my issue as well?
[13:14] <RoAkSoAx> mgw: sure
[13:15] <shang> RoAkSoAx: the log that I sent was from the bootstrap node.
[13:15] <RoAkSoAx> shang: there's 2 cloud-init logs
[13:15] <RoAkSoAx> shang: cloud-init.log and cloud-init-(something i cant remember :$).log
[13:16] <RoAkSoAx> mgw: what are your issues
[13:16] <mgw> RoAkSoAx: thanks… when I run koan, I'm gettin a 404
[13:16] <mgw> https://gist.github.com/2dfbcf76c6fd5d3b1c52
[13:16] <shang> RoAkSoAx: here is the getks log:  http://pastebin.ubuntu.com/725281/
[13:16] <RoAkSoAx> shang: oh btw... what preseed are you using, orchestra.preseed or juju.preseed?
[13:17] <shang> RoAkSoAx: the juju.preseed
[13:17] <shang> RoAkSoAx: should the bootstrap node using the orchestra.preseed?
[13:17] <RoAkSoAx> shang: nope
[13:17] <mgw> From what I understand, it should be hitting the squid proxy, but it seems to be trying to access the non-existent mirror
[13:17] <RoAkSoAx> shang: it's fine using the juju one
[13:17] <RoAkSoAx> mgw: oneiric?
[13:18] <mgw> yes
[13:18] <RoAkSoAx> mgw: can you file a bug with an step by step to reproduce so I can track it there
[13:19] <mgw> RoAkSoAx: Sure, but is there a way to get orchestra back to a pristine state?
[13:19] <RoAkSoAx> shang: so is zookeeper running on the bootstrap?
[13:19] <mgw> So I can retrace my steps
[13:19] <RoAkSoAx> mgw: i don't think it's orchestra itself but rather a bug in koan
[13:20] <mgw> really? ok… any ideas how I could get around it?
[13:20] <mgw> What URL *should* it be accessing/
[13:21] <mgw> In order to got through squid
[13:21] <shang> RoAkSoAx: should it? I didn't see it in any of the doc
[13:21] <shang> RoAkSoAx: sudo apt-get install zookeeperd ?
[13:22] <RoAkSoAx> shang: yes it should, could you pastebin /root/user-data-late-command /root/meta-data-late-command
[13:22] <RoAkSoAx> shang: cloud init should have taken care of that automatically
[13:22] <RoAkSoAx> mgw: i would have to reproduce it to know what is wrong with it
[13:23] <RoAkSoAx> mgw: but for what I can see, if you manyually try to browse http://10.1.1.1/cblr/repo_mirror/oneiric-x86_64/ it should be accessible to it, is it?
[13:23] <mgw> RoAkSoAx: no, I get a 404 there as well
[13:23] <RoAkSoAx> mgw: ahh i know what's wrong
[13:23] <RoAkSoAx> :)
[13:23] <mgw> :-)
[13:24] <mgw> You have no idea how many hours I spent last night trying to figure this out....
[13:24] <RoAkSoAx> mgw: its using http://192.168.123.2/cblr/repo_mirror instead of http://192.168.123.2/cblr/ks_mirror
[13:24] <RoAkSoAx> mgw: could you please just file a bug and I'll take care of it
[13:24] <mgw> ok
[13:25] <RoAkSoAx> mgw: and what you could try is to just import a full ubuntu-server ISO
[13:25] <mgw> can I hack koan to fix it for the moment?
[13:25] <RoAkSoAx> mgw: instead of the mini iso
[13:25] <RoAkSoAx> mgw: yeah but it is really in cobbler as the install_tree is obtained with information from cobbler
[13:26] <RoAkSoAx> mgw: so maybe a patch is not being applied correctly or similar
[13:26] <RoAkSoAx> mgw: what you could do is edit the profile and change the tree to http://192.168.123.2/cblr/ks_mirror/ instead
[13:26] <RoAkSoAx> mgw: that should be the fix, just edit the profile and change the repo_mirror to ks_mirror
[13:27] <shang> RoAkSoAx: http://paste.ubuntu.com/725287/
[13:27] <shang>  http://paste.ubuntu.com/725290/
[13:27] <freakynl> _ruben: afaik it's kernel, it has been in-kernel for a very long time. But I'm no expert either :). SGST is the entire scsi stack in the kernel afaik. SCST seems to have a lot of advantages, but yes it is a political war
[13:28] <RoAkSoAx> mgw: the reason why it is done like that is because ks_mirror is used when you use a full ISO because all the debs are there, while with a mini ISO, there are no *.debs so it assumes that we can create a repository and then copy the packages in repo_mirror/<profile> but the way we use it in orchestra and juju is really differently and does not use the concept of using repositories managed by cobbler
[13:28] <mgw> i see
[13:28] <RoAkSoAx> shang: that's the problem: http://paste.ubuntu.com/725290/ for some reason there's not user-data
[13:28] <mgw> trying to find where to edit the profile
[13:28] <RoAkSoAx> shang: are you usign juju from branch?
[13:29] <RoAkSoAx> mgw: either in the profile or in the distro
[13:29] <shang> RoAkSoAx: no,
[13:29] <RoAkSoAx> mgw: i dont remember right now
[13:29] <RoAkSoAx> shang: what version of juju are you using?
[13:29] <shang> RoAkSoAx: from the CD. apt-get install juju
[13:29] <RoAkSoAx> shang: uhmmmmmmmmmmmmm
[13:29] <RoAkSoAx> shang: that might be a problem with juju instead of orchestra
[13:29] <shang> RoAkSoAx: 0.5+bzr398-0ubuntu1
[13:30] <RoAkSoAx> shang: let me verify with someone from juju
[13:30] <RoAkSoAx> might be juju
[13:30] <RoAkSoAx> might be that orchestra is not doing what's suppose to
[13:30] <mgw> RoAKSoAx: cblr/ks_mirror/oneiric-x86_64/.treeinfo: HTTP Error 404: Not Found
[13:30] <mgw> :-/
[13:30] <shang> RoAkSoAx: ummm...
[13:30] <shang> RoAkSoAx: ok, let me try to install the zookeeper manually
[13:31] <RoAkSoAx> mgw: arrgh.. now it looks like the ubuntu patch I had is not correct :S:/
[13:31] <RoAkSoAx> mgw: just file a bug I'll track it today :)
[13:33] <RoAkSoAx> mgw: nah, the .treeinfo is from fedora/rhel
[13:33] <mgw> is there something missing in /cblr/ks_mirror/oneiric-x86_64/ ?
[13:33] <RoAkSoAx> mgw: so maybe the ubuntu patch i had for that no longer working
[13:33] <mgw> or maybe .treeinfo isn't what it should be looking for?
[13:33] <RoAkSoAx> mgw: but that might be virt-inst now :)
[13:34] <RoAkSoAx> mgw: nah .treeinfo is fedora/rhel specific, but I patched virt-inst for that
[13:34] <RoAkSoAx> so the issue might be in virt-inst
[13:34] <RoAkSoAx> mgw: please, file the bug I'll track it as soon as I can get a test environment
[13:34] <mgw> virt-inst on the guest system (where I'm running koan), right?
[13:35] <RoAkSoAx> shang: weird... in my case the user-data gets filled correctly
[13:35] <RoAkSoAx> shang: do this instead:
[13:35] <RoAkSoAx> shang: create the file user-data-late-command with this: http://paste.ubuntu.com/725299/
[13:35] <RoAkSoAx> shang: then excecute it
[13:36] <RoAkSoAx> shang: then reboot, and cloud-init should handle the rest
[13:36] <RoAkSoAx> mgw: yeah
[13:37] <mgw> How can I tell what version of virtinst I have installed (for the bug report)
[13:37] <RoAkSoAx> virt-install --version
[13:38] <mgw> ok, 0.600.0 is correct, right?
[13:38] <RoAkSoAx> mgw: yes
[13:39] <mgw> what *should* it be looking for, rather than .treeinfo?
[13:40] <_ruben> freakynl: according to http://scst.sourceforge.net/comparison.html : STGT has all SCSI target processing and memory management in the user space. In the kernel it has only a small library to help in-kernel target drivers to interact with it. As a direct consequence, neither iSCSI, nor iSER as well as any other user space STGT target need any STGT kernel modules loaded to run.
[13:44] <RoAkSoAx> mgw: don't remember.. and yeah it seems that the patch has been dropped
[13:44] <RoAkSoAx> mgw: gonna have to write a new patch
[13:48] <shang> RoAkSoAx: no, unfortunately, that does not work...
[13:48] <RoAkSoAx> shang: rm -rf /var/lib/cloud
[13:48] <RoAkSoAx> shang: then sh user-data-late-command
[13:49] <RoAkSoAx> sh meta-data-late-command
[13:49] <RoAkSoAx> shang: reboot and cloud-init should rerun
[13:49] <shang> RoAkSoAx: ok
[13:53] <shang> RoAkSoAx: wow, that's it!
[13:53] <RoAkSoAx> shang: cool, not the zookeeper should be installed, initialized, same as juju
[13:53] <RoAkSoAx> and you should be able to status to it
[13:53] <shang> RoAkSoAx: right, but what happen? is it bug or ?
[13:54] <RoAkSoAx> shang: to tell you the truth I have no idea because the user-data is being feed correclty into cobbler
[13:55] <RoAkSoAx> shang: and been feed correctly into the late command that creates that user-data-late-command
[13:55] <RoAkSoAx> shang: maybe something to do with localization?
[13:55] <shang> RoAkSoAx: sorry, my bad, same thing: the bootstrap works, but still not able to run the status command
[13:55]  * shang spoke too fast
[13:56] <RoAkSoAx> shang: yeah zookeeper needs to be running as well as juju
[13:56] <RoAkSoAx> shang: so just let cloud-init do its thing
[13:56] <RoAkSoAx> it might be just taking a while
[13:56] <RoAkSoAx> shang: and to figure out what could be wrong, manually run this: sh -c 'f=$1; shift; echo $0 | base64 --decode | gunzip > $f && chmod u+x $f && $f $*' H4sIAPHwr04C/5VU23LiRhB911dMOaniAYSQZAxxLalwlQFL3NmFrS3XoBndECOhmcFISf49I4HxOuU8hCd19+me06dPIVGM0QtqKSeYKKG/U+ww4kjJswqJikAmmEmHPfITIMfg7tdLx51kQwZ+f48VTnEiI8gg+PLlrj8Z3Em/XPrtiDi+K8GYyTwWAPwIWMLxNeEmEL1lDtD2fIKLKY/gz4AHXI6T6OQjMZqlscBFie1hyhJYAUU5i6I9xrGoexFl9BGUwsiGYR48ampTLVUkAN7m+kjUa6W/pYizmD
[13:57] <RoAkSoAx> shang: errr, this: http://pastebin.ubuntu.com/725324/
[13:57] <RoAkSoAx> shang: and see if the file is correctly created
[13:58] <mgw> RoAkSoAx: so the issue is the contents of cblr/ks_mirror/oneiric-x86_64/ — or virtinst?
[13:59] <mgw> (sorry, just trying to hack this to get it working…)
[13:59] <mgw> I see virtinst is not detecting the system as ubuntu
[14:00] <RoAkSoAx> mgw: try accessing directly to cblr/ks_mirror/oneiric-x86_64/
[14:00] <shang> RoAkSoAx: zookeeper seems to be running, I see the port 2181 is open
[14:00] <RoAkSoAx> mgw: yeah that's the problem
[14:00] <RoAkSoAx> shang: what about juju
[14:01] <shang> RoAkSoAx: on the bootstrap node?
[14:01] <RoAkSoAx> mgw: so I need to re-write whatever is needed to make it happen
[14:01] <RoAkSoAx> shang: yep
[14:01] <mgw> RoAkSoAx : i can access ks_mirror
[14:01] <mgw> but it doesn't have any of the files virtinst is looking for
[14:02] <shang> RoAkSoAx: so the cloud init suppose to install juju and the zookeeper, correct?
[14:02] <RoAkSoAx> shang: yes
[14:02] <RoAkSoAx> mgw: yeah there was a patch for it
[14:02] <RoAkSoAx> mgw: and I need to invastigate why it got dropped
[14:03] <RoAkSoAx> mgw: cause i also forwarded a patch upstream that was supposely taken
[14:03] <mgw> do you have a link to the patch?
[14:03] <mgw> opening a ticket now
[14:04] <RoAkSoAx> mgw: need to run for a bit, at the Ubuntu developer summit atm
[14:06] <mgw> ok
[14:06] <mgw> ty
[14:13] <RoyK> hm.. I have a client (or a few) using no_NO.ISO8859-1@bokmal as their locale, which seems to be a new name for nb_NO.ISO8859-1 - any idea if it's possible to add this as an alias somehow without overriding LC_* etc?
[14:14] <shang> RoAkSoAx: i install juju on the bootstrap node manually, rm the /var/lib/cloud, reboot, try bootstrap again, still can't see the status
[14:18] <RoAkSoAx> shang: run this: http://pastebin.ubuntu.com/725348/
[14:27] <shang> RoAkSoAx: on the orchestra node (where I deploy juju from)? or on the bootstrap node?
[14:27] <RoAkSoAx> shang: bootstrap
[14:28] <shang> RoAkSoAx: 2011-11-01 22:25:45,404 INFO Initializing zookeeper hierarchy
[14:28] <shang> node exists
[14:28] <shang> RoAkSoAx: 2011-11-01 22:25:45,404 INFO Initializing zookeeper hierarchy
[14:28] <shang> node exists
[14:28] <shang> 2011-11-01 22:25:45,417 ERROR node exists
[14:32] <RoAkSoAx> shang: drop the admin stuff
[14:32] <RoAkSoAx> JUJU_MACHINE_ID=0 JUJU_ZOOKEEPER=localhost:2181 python -m juju.agents.machine -n --logfile=/var/log/juju/machine-agent.log --pidfile=/var/run/juju/machine-agent.pid && JUJU_ZOOKEEPER=localhost:2181 python -m juju.agents.provision -n --logfile=/var/log/juju/provision-agent.log --pidfile=/var/run/juju/provision-agent.pid
[14:32] <RoAkSoAx> just use that
[14:36] <shang> RoAkSoAx: Another twistd server is running, PID 1529
[14:36] <shang> This could either be a previously started instance of your application or a
[14:36] <shang> different application entirely. To start a new one, either run it in some other
[14:36] <shang> directory, or use the --pidfile and --logfile parameters to avoid clashes.
[14:53] <mgw1> RoAkSoAx: I've opened the bug (#884765)
[15:03] <p1ruj3> hey, so had a entry in deny.hosts I needed to remove (local printer who freaked out) how do i rehash that file so the printer is no longer banned?
[15:05] <mgw> I also found the original patch (http://www.redhat.com/archives/virt-tools-list/2011-April/msg00146.html) — but I see it is not looking for any of the files in /cblr/ks_mirror/
[15:07] <RoAkSoAx> shang: kill it and re run it
[15:07] <RoAkSoAx> and you should be able to get the status
[15:08] <RoAkSoAx> mgw: that's passed by koan
[15:08] <RoAkSoAx> mgw: that';s why i suggested to change that manually for mini iso's
[15:08] <RoAkSoAx> mgw: but there was a newer patch for that
[15:08] <mgw> where would I change it?
[15:09] <mgw> I koan?
[15:09] <RoAkSoAx> mgw: cobbler profile or distro
[15:09] <mgw> oh, i changed from repo_mirror to ks_mirror
[15:09] <mgw> in distro
[15:09] <mgw> was that incorrect?
[15:09] <RoAkSoAx> mgw: that's correct
[15:10] <RoAkSoAx> mgw: now that patch should work
[15:10] <mgw> but it's not… it's not finding any of the files (MANIFEST, install, etc) in /cblr/ks_mirror/oneiric-x86_64/
[15:11] <mgw> … /cblr/ks_mirror/oneiric-x86_64/ has initrd.gz, linux, etc
[15:12] <mgw> i.e., isValidStore is returning False
[15:12] <mgw> For UbuntuDistro
[15:13] <mgw> If I import a full install iso, that might fix it though, right?
[15:14] <RoAkSoAx> mgw: it should
[15:17] <shang> RoAkSoAx: re run the bootstrap command?
[15:18] <RoAkSoAx> shang: yeah destroy the environment and rerun it
[15:18] <RoAkSoAx> mgw: http://paste.ubuntu.com/725405/ http://paste.ubuntu.com/725406/ http://paste.ubuntu.com/725408/
[15:18] <RoAkSoAx> that's agains the older verrsion
[15:20] <mgw> RoakSoAx: I have the patched version
[15:21] <mgw> but install/netboot/version.info is not in ks_mirror/*/
[15:21] <mgw> either
[15:21] <mgw> maybe it's just my iso import is bad
[15:22] <RoAkSoAx> mgw: yeah it will detect it one way or another
[15:22] <RoAkSoAx> mgw: not all ISO's have version.info
[15:22] <mgw> mine seems to have neither version.info nor MANIFEST
[15:23] <mgw> FYI, I have whatever got automatically pulled in by ubuntu-orchestra-server
[15:23] <mgw> I've not done any cobbler import
[15:23] <RoAkSoAx> mgw: i knoe :)
[15:23] <RoAkSoAx> let me check the patches
[15:23] <mgw> ok
[15:23] <mgw> thank you
[15:28] <RoAkSoAx> mgw: yeah i think something might nbe missing in those patches
[15:28] <RoAkSoAx> mgw: i don't have the latest patch I had for it in this computer
[15:28] <RoAkSoAx> so I will have to re-test re-write
[15:29] <mgw> I'm guessing it should check for initrd.gz
[15:29] <mgw> ?
[15:29] <mgw> or linux
[15:29] <RoAkSoAx> mgw: both
[15:29] <mgw> but then what?
[15:30] <RoAkSoAx> mgw: i think that's the fix I did. I will try to search for the patch whenever I can access my home computer and provide a fix for it soon
[15:35] <RoyK> any idea where to change things to make oneiric start with the xen kernel by default? grub2 is confusing...
[15:40] <mgw> RoAkSoAx: I imported the full server image, now it keeps wanted to import the kernel for i386:
[15:40] <mgw> … /cblr/ks_mirror/ubuntu-oneiric-x86_64/install/netboot/ubuntu-installer/i386/linux
[15:40] <mgw> Which of course is not there
[15:41] <mgw> I've double checked the distro kernel setting
[15:41] <mgw> … /var/www/cobbler/ks_mirror/ubuntu-oneiric-x86_64/install/netboot/ubuntu-installer/amd64/linux
[15:43] <RoAkSoAx> mgw: i'll have to test it myself, dont have the enivronment to this right now
[15:45] <mgw> ok, sorry
[15:45] <mgw> thanks for all the help
[15:45] <RoAkSoAx> mgw: no worries, you dont have to apologize i just done have the test environment
[15:45] <RoAkSoAx> mgw: will try to tackle it down tomorrow though
[15:45] <RoAkSoAx> mgw: thanks for making me aware of this otherwise I probabloy wouldn't have
[15:45] <RoAkSoAx> lol
[15:45] <RoAkSoAx> :)
[15:49] <dsirijus> hey, this is my iptables -L output
[15:49] <dsirijus> http://pastie.org/2793990
[15:50] <dsirijus> but i cannot telnet to either 8080, 9933, nor 9399
[15:50] <mgw> RoAkSoAx: the current problem seems to be with this line in OSDistro.py
[15:51] <p1ruj3> I removed an entry from hosts.deny, restarted denyhosts, and the removed entry still cannot connect, what am I missing? If I reboot it will work.... How can I rehash this entry removal without rebooting?
[15:51] <mgw> 897:         elif uri.count("installer-amd64"):
[15:52] <mgw> changing the else: a few lines later to amd64 fixes that problem, seemingly
[15:58] <mgw> ok.. got a vm running in virsh (via koan)
[16:00] <mgw> But it's not showing in cobbler system list
[16:00] <shang> RoAkSoAx: still the same error... thanks for the help BTW, I really appreciated this!!
[16:01] <shang> RoAkSoAx: do we need to reinstall the sytstem everytime we destroy the environment?
[16:02] <RoAkSoAx> shang: yes
[16:02] <RoAkSoAx> shang: every time you bootstrap it should be a new installation
[16:02] <mgw> RoAkSoAx: does koan do everything, A-Z? that is, does it register it in cobbler as a system?
[16:03] <RoAkSoAx> mgw: cobbler just uses cobbler to deploy
[16:04] <shang> RoAkSoAx: um.... so if we modify anything on the node1 (zookeeper) or even reboot, we don't need to destroy-environment?
[16:06] <RoAkSoAx> shang: destory environment will technically destroy the system within juju but wont uninstall the machines obviously, because on ec2 it just destroys the instances
[16:06] <RoAkSoAx> shang: so destroying the environment is just telling that the machine is available now again to be deployed with something else
[16:07] <RoAkSoAx> which will mean it will have to be PXE booted again and ubuntu installed agian
[16:07] <RoAkSoAx> if you just reboot the machine just because
[16:07] <RoAkSoAx> you need tomanually start juju
[16:07] <mgw> ok, so set up the system first in cobbler, then use koan —system ?
[16:07] <RoAkSoAx> becaues they still not have upstart jobs
[16:08] <RoAkSoAx> mgw: you want to register a system into cobbler with koan? you usually register the system first
[16:08] <shang> RoAkSoAx: so we should setup the profile or juju deploy xxx before we turn on the system?
[16:09] <Potatoe> in the package description that contains Package, Priorirty, section, Maintainer, etc.... Is their a field that indicates a security update ?
[16:09] <Potatoe> or that it is a stability fix, feature fix etc... ?
[16:09] <RoAkSoAx> shang: yes, before you run juju deploy you should have available systems. Then when you should run juju deploy, and right after that, you should turn on the system
[16:10] <RoAkSoAx> shang: if you configure power management for the machine, i.e. wkae on lan
[16:10] <RoAkSoAx> the machine should be turned on automatically
[16:10] <RoAkSoAx> right after juju deploy
[16:10] <RoAkSoAx> shang: or even, right after juju bootstrap
[16:10] <mgw> RoAkSoAx: i've tried using koan with both —system and —profile
[16:10] <mgw> they both seem to be working
[16:10] <RoAkSoAx> mgw: if a system exist, --system will work otherwise profile
[16:10] <mgw> but profile does not register as far as i can tell
[16:11] <mgw> but now juju bootstrap is telling me i have no systems
[16:11] <shang> RoAkSoAx: so we should have the bootstrap run first, then turn on the system?
[16:11] <RoAkSoAx> shang: so basically, 1. install orchestra 2. add systems 3. juju bootstrap (3.1. turn on machine if not done automatically) 4. juju deploy (4.1 turn on machine if not done automatically)
[16:11] <RoAkSoAx> mgw: you need to add systems to cobbler
[16:12] <shang> RoAkSoAx: ok
[16:12] <RoAkSoAx> mgw: you need to add systems to cobbler with the management classes you define or the defaults installed by orchestra
[16:12] <shang> RoAkSoAx: unfortunaltey, I wont be able to test it out today... I will definitely give that a try tomorrow
[16:12] <shang> RoAkSoAx: thank you very much for all the help
[16:13] <RoAkSoAx> shang: you're welcome
[16:14] <freakynl> _ruben: lio != sgst's iscsi target :) lio is also on the comparison page, afaik it just uses / is more compatible with sgst (which is why they ditched scst iscsi target from kernel in favor of this one)
[16:18] <mgw> still getting the problem with juju bootstrap
[16:18] <mgw> I added my test2 system to juju-orechestra-available
[16:20] <mgw> but maybe my problem now is networking
[16:22] <dsirijus> here's my iptables -L http://pastie.org/2793990
[16:22] <dsirijus> i still cannot telnet to either of the ports, how come?
[16:24] <dsirijus> the ports i'm interested in are 8080, 9933 and 9399
[16:35] <Potatoe> I am doing an audit of my servers and I notice that security updates are on lucid-updates and not lucid-security. Am I doing something wrong ?
[16:36] <Potatoe> ah ha! The order of the repository matters!
[16:41] <Potatoe> Looks like the update comes from the first repository in the list, so if I put security before updates, I see my updates are coming from the security repository!
[16:56] <b0gatyr> hi folks, is there something like "periodic" in freebsd in ubuntu?
[16:57] <sevist> ïðèâåò íàðîä
[16:59] <Pici> b0gatyr: What does periodic do?
[17:03] <b0gatyr> Pici: just compiles different security, system reports on a daily/weekly/monthly basis and sends it by email.. but now that I think about it, it can be done just by creating a script with the output I want then mail it.
[17:04] <b0gatyr> just thought linux had something similar..
[17:04] <Pici> b0gatyr: logwatch might do what you're describing.
[17:04] <b0gatyr> logcheck seems cool too
[17:04] <Pici> If its just collect system logs, email to somewhere ton a regular basis, thats what logwatch does.
[17:04] <Pici> Also has different levels of verbosity.
[17:05] <b0gatyr> nice, i'll give it a test run
[17:05] <b0gatyr> thanks
[17:05] <Pici> np
[22:16] <puppy3ghz> hello room.
[22:16] <squidly> I just updated my server to 11.10, and I'm having an issue starting up my network bonding and bonding aliases
[22:19] <squidly> I keep getting RTNETLINK answers: File exists
[22:19] <squidly> Failed to bring up bond0:1.
[22:20] <squidly> does anyone have an idea what that would happen?
[22:26] <puppy3ghz> what is the usual way to share a folder using permissions across the internet? simplest would be the best.
[22:27] <puppy3ghz> read and write and only read. per user.
[22:47] <virusuy> puppy3ghz: share a folder across the internet ? over VPN ?
[22:48] <qman__> puppy3ghz, SFTP, and HTTP, respectively
[22:57] <puppy3ghz> i have a http server (xp) but dont want to use that anymore.
[22:58] <qman__> I simply chose what is in my opinion the best of the most common methods
[22:58] <qman__> there are literally thousands of options
[22:59] <puppy3ghz> i was thinking ubuntu with hiawutha or maybe try samba. i want to have folder permissions
[22:59] <qman__> samba is for local sharing
[22:59] <qman__> it's possible but incredibly unwise to use it over the internet
[22:59] <puppy3ghz> i want to have some folders to read and write from across the inet
[22:59] <qman__> you can, however, use a VPN, or webDAV
[23:00] <qman__> all file sharing solutions offer that option
[23:00] <qman__> they're called permissions
[23:00] <puppy3ghz> all he client machines are xp or vista.
[23:01] <qman__> your options without installing extra software are HTTP, FTP, or webDAV
[23:01] <qman__> for reasons that should be obvious, I recommend you don't use FTP, for anything, ever
[23:01] <puppy3ghz> oops... ftp.
[23:01] <qman__> SFTP is easy to use with free programs like WinSCP or FileZilla
[23:02] <puppy3ghz> i have file zilla, but need the window to be more user freindly for the non techy folks.
[23:03] <puppy3ghz> but on filezilla i think i might have had issues with sftp. cant rember why
[23:03] <qman__> SFTP is the most secure and widespread
[23:03] <qman__> webDAV is easy on the users but really difficult to get set up
[23:03] <qman__> it's also had some security problems in the past
[23:04] <qman__> samba + openVPN is a good solution too, but the users have to deal with the VPN client
[23:04] <qman__> and some users are too thick to work it
[23:05] <puppy3ghz> i need something very simple that would appear as a mapped drive
[23:05] <qman__> only way to have a mapped drive is samba
[23:05] <qman__> webDAV appears similarly but not as mapped drives
[23:06] <puppy3ghz> can a http do a simular appearance to a maped drive for xp?
[23:07] <qman__> no
[23:07] <qman__> at least, not without extra software I have not seen
[23:07] <puppy3ghz> can i set filezillla to keep the connect open instead of timing out?
[23:07] <qman__> yes
[23:08] <qman__> I would also look into a possible sshfs client for windows
[23:08] <qman__> sshfs is a way to mount SFTP shares in linux
[23:08] <qman__> there's no technical reason it couldn't be done on windows, but I don't know of any programs that do it
[23:11] <puppy3ghz> let me hook up my server stuff..... 4port KVM, 6computers.