=== mrjazzcat is now known as mrjazzcat-afk [00:42] :S [00:42] is kill built into the kernel? [00:42] it can't be.. but I can't figure out how the command is available in my chroot [00:43] MTecknology: procps [00:44] jpds: thanks [00:59] thesheff17: https://code.edge.launchpad.net/~jailkit/+archive/ppa/+builds?build_state=pending [01:00] thesheff17: it progressed to the next step... already.... [01:00] MTecknology: nice [01:01] MTecknology: I'm really new to the launch pad stuff...how did you get that in there? or can you link me to how to? [01:03] thesheff17: made a team, project, registered an upstream import, created a recipe, told the recipe to build [01:04] thesheff17: kind of a lot of stuff - you'd want to start with building your own packages in your own ppa - #ubuntu-packaging can help with that [01:04] MTecknology: ok cool...I will have to take a look thx [01:05] thesheff17: if the package works nice and perfect - I'll look into getting it into debian/ubuntu [01:06] MTecknology: excellent I think a chroot implementation through apt-get would be excellent. [01:07] thesheff17: this thing is already amazingly easy.. my how to is going to make it even easier :) [01:07] MTecknology: nice [01:07] thesheff17: I'll add two things that were missing and expand. I feel like I have a clue now. :) [01:08] I'm going to start on homework thouhg. :'( [01:08] This'll be the worst 5 hours of the day [01:08] probably longer actually [01:08] MTecknology: yea I'm in class right now :( [01:08] :P [01:08] paying attention I see [01:08] MTecknology: hehe yea [01:08] thesheff17: thanks again - I'll ping you when I get somewhere [01:09] * MTecknology runs off now.. [01:09] MTecknology: np ttyl [01:09] ttyl === bastidra1or is now known as bastid_raZor === _Techie_ is now known as _Techie_|_Away_ [01:34] iwconfig === jjohansen is now known as jj-afk === _Techie_|_Away_ is now known as _Techie_ === _Techie_|_Away_ is now known as _Techie_ [02:58] hi there, somebody know how to stop tcpdump automatically? when i do crl -c it stop but the process still appear in the system [03:02] Roxyhart0: can you get it's pid and `kill -9` it? [03:29] yes, but is not the idea as when i try then to merge files that i got with tcpdump give me errors as the file was cut [03:30] Roxyhart0: tcpdump is staying open? [03:31] when i do crt C, still is running... [03:31] i woudl like to stop with some condition [03:31] Roxyhart0: are you giving it an option to run in the background at all [03:33] no, i just do tcpdump -i van2 scr portrange 2000 -65535 [03:35] you have an interface named van2? [03:35] yes [03:35] it is running ok [03:38] Roxyhart0: interesting so when you run that and you hit ctrl c it continues to run === _Techie_ is now known as _Techie_|_Away_ [03:39] yes [03:39] Roxyhart0: how are you checking for it? ps aux | grep tcpdump [03:39] i do ps -fea |grep tcpdump and still ahow me the pdid [03:40] ahh [03:40] do a ps aux | grep tcpdump [03:41] i did and still shows me the pdid from tcpdump [03:41] is it there? [03:41] yes [03:41] pastebin the output to me [03:43] ps aux | grep tcpdump | grep -v grep [03:43] youre not seeing the grep output and thinking its tcpdump are you [03:44] i know was i did wrong i did cr z and not cr c, but still i would like to stop it automatically as i am running the comand with cron each 5 minutes for 1 minute [03:44] and i just can stp with killall -9 tcpdump [03:45] yeah ctrl z will let you type bg to background it [03:46] or fg to forground it [03:47] tcpdump -G will let you rotate files [03:49] i tried to use that, i will check a little bit more [03:49] thanks [03:49] Roxyhart0: if not the best bet it to run it from a script that will run it record the pid and kill it after a min [03:49] easiest way [03:50] yes, i did it [03:51] okay sweet [03:56] thanks :) [03:58] If i wanted to setup up central authentication for a small company network from scratch, which way should I go, openldap, samba, kerberos, or samba 4 ? === erichammond1 is now known as erichammond [04:24] Daviey, i dont see a eucalyptus release or build, were you still expecting one ? [04:39] How can I chain SSH logins? Like.. ssh user1@domain1.com ssh user2@domain2.com [04:39] I do it - but then I can't type anything [04:39] I can see I get logged in through the chain === erichammond1 is now known as erichammond [04:42] heh.. got it [04:50] MTecknology, what did you find ? [04:50] you can do it 2 ways. [04:51] robably by forcing the terminal on the first 'ssh -t user1@domain1.com ssh user2@domain2.com' [04:51] smoser: -t is what I found [04:51] smoser: is there something better? [04:51] probably [04:51] hold on [04:52] in .ssh/config [04:52] Host domain2.com [04:52] ProxyCommand ssh user1@domain1.com nc -q0 %h %p [04:53] that Proxies you through the domain1 to get to domain2 [04:54] cool [04:54] .ssh/config has a lot of little gems, huh? [04:54] I am having a problem configuring my interfaces, i have a block of static ip's and 2 netowrk cards in my server. one has a internal ip in the 192.168.2. subnet and the other is in my public pool of addresses. the problem is i cannot access my interface with the public ip outside of my local network. any ideas ? === _Techie_|_Away_ is now known as _Techie_ [05:04] taget: sounds like a route problem [05:04] thats kind of what i thought also, but i am unsure of how to fix it [05:04] taget: can you server ping the outside? === erichammond1 is now known as erichammond [05:05] taget: you can adjust your route by doing "route del default gw x.x.x.x" and "route add default gw x.x.x.x" [05:06] taget: first get your box to reach the internet. [05:06] taget: and the private network shouldn't have any routes. [05:06] found TraceEnable Off in security.dpkg-dist but I think TRACE is still enabled [05:07] sorry, i was disconnected [05:07] i can ping the outside, but i hace not tried disabling the internal network interface [05:08] taget: you shouldn't have to disable the private network interface [05:08] how can i be sure i am accessing outside of my network with the correct interface ? [05:09] taget: as root type route and is that IP in the same subnet as the public network? [05:10] thesheff17: the first destination is in the correct subnet, but the gateway is an * [05:11] after the first entry i have a 192.168.2.0 dest to * entry and a default to 192.168.2.10 entry [05:12] taget: can you paste all the output to pastebin [05:12] sure [05:14] http://pastebin.com/WFaU1V6v [05:14] thesheff17: oh ya.. it built - https://code.edge.launchpad.net/~jailkit/+archive/ppa [05:14] MTecknology: nice [05:14] taget: so this is the private network? 192.168.2.100 [05:15] correct [05:15] taget: your route should be something on the public side. [05:15] taget: which is the last entry [05:16] thats kind of what i was thinking but i am unsure of how to change it [05:16] taget: /etc/network/interfaces controls your interfaces [05:17] yup [05:17] but just adding the interface will not adjust the routing, correct ? [05:17] taget: correct [05:17] taget: I will pastebin my dual network /etc/network/interface [05:18] i have eth0 configured for static with all of the correct info [05:18] ty [05:20] http://pastebin.com/Vp2BekR9 [05:20] taget: so all you should need for the private network is ip and subnet [05:24] taget: and when you type route the only entry in gateway should be the IP of your public gateway. [05:24] thanks for the help, i will make some corrections and try it. i am looking at my interfaces file right now and for some reason while using nano it removed all the text that wasnt currently in view of the terminal [05:25] and one of the lines it cut out was the gateway [05:26] taget: yea be careful if you only have remote access. also vim is a great text editor [05:27] that is what i have heard, i just have never taken the time to learn how to use it. also the server is only 15 minutes away from my home :) [05:27] so i ma fortunate in this case [05:31] taget: cool yea most of the machines I manage are hours away if not states away :) [05:34] thesheff17: i hear you on that, i am fortunate in this case, id hate to have to jump in the car to go and reset a server in this instance. [05:36] On another topic, i am looking for a solution for a web based file access with a nice login and ui for downloading and uploading files to a server. (family file server) [05:36] any ideas ? [05:37] taget: what I do is create a cronjob that runs every hour and restores the original network config regardless of what I do.....so if and when I do break it I know within the hour it will be fixed. Just remember though when the hour comes around your config file will change regardless of what you do. [05:37] taget: I would use apache with php [05:38] thesheff17: good idea. [05:40] taget: apache can easily do https/login/download... php is good for uploads and moving the files around where you want them.....though there is a couple of places to adjust php in order to upload big files but it is well documented. === dendro-afk is now known as dendrobates [05:40] MTecknology: so that built package is that a manually approved process or is that automatic? [05:41] thesheff17: i will look into that in the morning. thank you for the help [05:41] taget: no problem I should be around if you need more help [05:42] somebody know any good solution to block P2P traffic? [05:42] Roxyhart0: I have create a linux based router and then you can forward all the traffic through that box and filter what you want and don't want. [05:43] thesheff17: that's just a normal build. It could keep building there forever and never make it into ubuntu. There's a couple different processes that you need to go through to get it included [05:45] thesheff17, do you post that solution in somewhere? [05:45] did [05:45] Roxyhart0: http://www.gentoo.org/doc/en/home-router-howto.xml this one is based on iptables that I have used and you have to adjust for ubuntu but it is pretty straight forward [05:46] thanks a lot [05:46] Roxyhart0: np [05:47] Roxyhart0: you don't have to anything with the kernel...that is all built into ubuntu [05:47] ok, thanks ..i will have a look [05:50] Roxyhart0: I actually just let all the traffic through on the router and then install a monitoring tool on the private interface to see what clowns where using P2P stuff...because sometimes I would use bittorent stuff to download ISO and other important data. [05:52] MTecknology: well I hope it makes it into universe...but I can tell you I have no clue how all this works :) [05:52] thesheff17: me either :P [05:52] thesheff17: I think only about 1 or 2 actually do :P (or a few more-but it's not a lot that really understand it all) [05:53] thesheff17: 1) if it gets into debian, it'll be in ubuntu 2) if not, then you need to get it into ubuntu directy 3) sometimes debian realizes it's good and then brings it in [05:54] MTecknology: ah ok [05:54] thesheff17: all three ways are a different process and have pros/cons [05:55] thesheff17: I'm actually considering just helping the developer get it into debian - they're so close to being able to do it easily [05:56] thesheff17: for now.. You can install from that PPA I linked you to [05:56] MTecknology: excellent yea I feel like I haven't contributed much to ubuntu yet...hopefully that will change :) [05:57] thesheff17: I feel the same :P [05:59] thesheff17: You offer support in here which already helps a lot. You could also check out https://answers.launchpad.net/ubuntu (and also https://answers.launchpad.net/launchpad - a lot get incorrectly put here) [06:00] MTecknology: I haven't been in a irc chat room since the early 90's of the aol days for who knows what. Its great to be back. I also feel like the forums for ubuntu aren't great...most of the questions go un answered which is disappointing. When I was big into gentoo (before I found ubuntu) I would answer as much as I could on the forums. [06:01] MTecknology: it could though just be the sheer volume of ubuntu users vs gentoo users. [06:02] thesheff17: it is indeed. Ubuntu is geared toward easing the enterance barrier where Gentoo is absolutely not out to do that. [06:03] MTecknology: haha I know...gentoo was the first linux distro I found and I thought all linux distos where like that :) how wrong I was. [06:05] thesheff17: Ubuntu and Fedora are usually the first distros a person tries. That comes with "Where did my Start button go?" and "Where is my Outlooks?" which are so obviou to us, but not someone who's never tried something different. (I hate the pluraization of outlook) I don't think it's possible to eliminate most of the support without looking and acting exactly like windows. Problem is that the efforts to do that wound [06:06] thesheff17: #ubuntu, #ubuntu-server, answers.launchpad.net/ubuntu, and ubuntuforums.com and the bestest best places to help new users. [06:07] thesheff17: You helped me out a massive amount today. Where did that go? The tool you showed me to may now wind up in the repositories. :D [06:07] today and yesterday* [06:08] MTecknology: yea I have no problem helping as much as I can...because I wouldn't know anything without the help of everyone else....I never touched a linux machine in college and I graduated with a comp sci degree which is sad [06:08] thesheff17: oh.. and if you really want to get ambitious... https://answers.edge.launchpad.net/ubuntu/+questions?field.search_text=&field.sort=RELEVANCY&field.sort-empty-marker=1&field.actions.search=Search&field.language=en&field.language-empty-marker=1&field.status=OPEN&field.status-empty-marker=1 [06:08] hi thesheff17, sorry my question i did capture package with tcpdump and then analyze with wireshark and i got some pakages that are recognized by wireshark as bittorrent, but the port that shows me for that oacjages are > 50000 usually, and the rule that your shows me iptables -t nat -A PREROUTING -p tcp --dport 6881:6889 -i ${WAN} -j DNAT --to 192.168.0.2 [06:09] the port say 6881:6889, maybe i am confuse? [06:09] thesheff17: that's about how my university is. <20 people actually understand linux at any depth in the entire university including both faculty and students [06:10] Roxyhart0: all ports over 1024 are unprivileged ports and P2P can run on all of them...so only enable what you want. [06:10] Roxyhart0: I think :) [06:11] yes, i think so...just was confuse with the rule that said for bittorent, why say port 6881:6889 [06:11] Roxyhart0: prob just what is standard but feel free to adjust. [06:16] ok, thanks a lot! [06:17] do you know ipp2p? [06:17] im wondering how as good it is? [06:18] Roxyhart0: I do not...reading a little about it now [06:23] Roxyhart0: sounds like that is still under massive development and has moved to http://www.ipp2p.org/ [06:24] so, still wait to use it [06:26] Roxyhart0: well I would just do extensive testing before moving into production where ever you are using it. Though that is pretty much anything I use on linux :) [06:26] thanks :) [06:37] ttx: Don't suppose you are around yet? :) [07:02] Roxyhart0: with my experience with ipp2p is that you just can't keep up with the tricks p2p software will throw at you [07:03] Roxyhart0: instead, go about it another way: use HTB+SFQ to throttle subscribers (tc) and keep track of heavy abusers by logging b/w per IP in some way. if someone is downloading (anything really) too much, blast them @ tc [07:04] Roxyhart0: p2p software, or any software for that matter, can't download through a pipe bigger than the one you give it, so its most reliable to manage your problem that way [07:13] Daviey: I'm here [07:35] sorry drw200169, i was away ...im interested in your sugestion, do you know some page how to? [07:53] does anyone know if its possible to │ [07:53] | connect to an ssl vpn on a headless │ [07:53] Does anyone know if it is possible to connect to a ssl vpn on a headless linux box? [08:01] * Roxyhart0 slaps Roxyhart0 around a bit with a large trout [08:02] Roxyhart0: haha, i'm looking for something for you;) [08:05] Roxyhart0: see, the problem with doing traffic shaping, is that there is no easy way to do it [08:07] Roxyhart0: i ran into the same problem(s) with p2p traffic in the past, and ended up using traffic shaping to do all my dirty work [08:16] Roxyhart0: you may be interested in reading http://voxel.dl.sourceforge.net/project/htbinit/HTB.init/0.8.5/htb.init-v0.8.5 , the man pages for tc, tc-tbf, tc-htb, tc-sfq, and of course the manual on http://lartc.org [08:18] Roxyhart0: note, also, that iptables and tc can work together using fw marks (which means tc filtering can be endless ...) === jj-afk is now known as jjohansen [08:41] somebdy know how i can copy from a server to another files and folder preservating permisions? [08:42] <\sh> rsync with preserve permissions [08:42] do i need to install that tool? [08:42] <\sh> if it's not already installed, then yes [08:43] thanks a lot [08:47] xserver on windows machine, or rdp server on the linux server? [08:56] thanks a lot sh..it work perfect.. [08:58] * cwillu_at_work feels an urge to introduce sh to \sh, and watch the fireworks fly [08:58] -> #lesswrong :) [08:59] although he doesn't seem to be around at the moment :/ [09:06] morning [09:49] Hi. Is there a difference between doing "ifconfig ... up" in /etc/rc.local, and having "auto eth0..." in /etc/network/interfaces? [09:51] <\sh> cwillu_at_work: hmm? [10:01] StefanMonov: Yes. [10:02] StefanMonov: Numerous differences. [10:09] \sh, other sh [10:12] <_ruben> ifconfig .. yuck [10:14] supposedly-final Maverick alpha3 server ISO candidates posted at http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [10:14] let's test the heck out of them ! [10:21] * Daviey grabs his heck removal tool. [10:22] ttx: do you have a netboot environment too (pxe stuff), then I might try them. [10:23] I have a few lucid annoyances that I wouldn't mind testing to see if they are gone [10:23] We have some netboot tests in http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [10:24] they are not that well documented though [10:24] ah, excellent. let me see if I can get the time this afternoon to do an intsall test, or not. [10:25] maswan: We are all very keen to get PXE booting solid, so if there are some quirks you have discovered - do please raise them :) [10:26] <\sh> hmm...why don't we install ifenslave-2.6 during server install? bonding is not that uncommon for servers... [10:29] \sh: I think we'd need to drop something else to get that on there.. but you could raise a blueprint for maverick+1, "Making bonding rock" meaning it's at least proposed for discussion. [10:29] the most annoying lucid (and karmic before that) bug for intsalling for me is certainly 415353/571872. [10:29] <\sh> bug #415353 [10:29] Launchpad bug 415353 in linux "karmic installation slow on "detecting network hardware" with bnx2x" [Undecided,Incomplete] https://launchpad.net/bugs/415353 [10:29] bug #415353 [10:30] hah, \sh you are as lazy as me :) [10:30] <\sh> bug #571872 [10:30] Launchpad bug 571872 in debian-installer "bnx2x drivers take 100 seconds to initialize" [Undecided,New] https://launchpad.net/bugs/571872 [10:30] <\sh> Daviey: I'm sysadmin, I have to be lazy ;) [10:30] It makes install's "detecting network hardware" step for me take ~20 minutes. [10:31] \sh: http://twitter.com/Daviey/status/15926972131 [10:31] <\sh> maswan: these are not those new HP Flex10 NICs for the new blade server series? [10:31] \sh: uip [10:31] yup even [10:31] <\sh> maswan: oh shit [10:32] \sh: booting them when installed isn't that bad, but the installer part is rather annoying. [10:32] Hmm [10:32] anyway, I was going to relocate myself. bbiab. [10:33] <\sh> maswan: also during PXE boot and nfs root mounts? I mean, ipconfig inside initramfs is hell alone... [10:33] \sh: dunno, haven't tried that. but it takes a couple of minutes to get the interfaces up during a normal boot. [10:34] (after the console gets to the login: prompt you have a minute or two before networking) [10:34] ugh, really should run, back in 30 [10:36] <\sh> maswan: lovely...that sounds like fun for me during the next couple of months of HW testing [10:41] Daviey: starting a UEC/amd64/topo1 test [10:42] ttx: Same here.. having to use a CD-R - as my usual pendrive for this is U/S [11:00] Daviey: arh, its broken [11:01] tgt deps not on CD [11:06] ISO testing news: looks like we'll have to respin again [11:08] ETA for new spin is ~2 hours [11:10] \sh: yup. it's "slightly" annoying when you run into disk numbering issues and grubs that fail to boot, which you "just" need to boot a rescue environment to fix. [11:34] Hi. I'm porting my mailserver from centos to ubuntu server, but one of my scripts use 'passwd --stdin' functionality which ubuntu's passwd command doesn't have. Is there any alternatives? / A way to get the same passwd functionality? [11:36] psteyn: Ehm, usermod, IIRC [11:37] ahh [11:37] But that only takes the crypted user [11:37] But that only takes the crypted password [11:37] Just a sec, I should have it somewhere [11:38] I only have versions where a crypted password is used [11:39] chpasswd worked for older Ubuntu's [11:39] usermod -p is needed for > karmic [11:45] I see, how would I encrypt the password? I tried with mcrypt..but it doesn't seem right. [11:49] nm [11:49] usermod -p `mkpasswd -H md5 newpass` username [11:49] :") [11:50] :) [11:53] err! [11:56] Hmm.. i had a script for just that purpose.. just looked it up.. and it's not as clean as i had hoped either :) [11:56] http://pb.daviey.com/JvD4/ [11:58] Daviey: That looks like a wrapper around passwd, what's the use? [12:01] Jeeves_: being able to set the passwd from a script === _Techie_ is now known as _Techie_|_Away_ [12:05] Hm. Maverick looks much better, only 5 seconds per interface. Still takes a bit over a minute for the installer to cycle through them all, but much better. [12:07] maswan: The installer? [12:07] maswan: Can you mention that on the bug report please? [12:07] Yeah, I am going to [12:08] I'm just busy runnign through a full install first. The boot screen was weird, I had to hit F1 for help before I could hit enter to boot the installer, and there were no "Advanced" choices either [12:09] hmm.. interesting [12:11] Yeah, I was wondering if I had messed up my download of the ubuntu-installer directory, but I could hit F1 for the normal help and then enter to install from that screen.. [12:13] \sh: so, good news, assuming you're planning on running the maverick instead of the lts. ;) [12:14] oh, wait, my mistake [12:14] I had messed up, the wrong pxelinux.cfg file. [12:19] hello [12:19] is there a way to force postfix to use only TLS in his communications ? [12:41] ETA for new Maverick Alpha3 ISO candidates for server: 30min [12:52] New bug: #613414 in dhcp3 (main) "package dhcp3-server 3.1.3-2ubuntu3 failed to install/upgrade: subprocess installed pre-removal script returned error exit status 2" [Undecided,New] https://launchpad.net/bugs/613414 === oubiwann is now known as oubiwann-away [12:58] <\sh> maswan: no...it has to be lucid for the next 2 years until next LTS release ;) === oubiwann-away is now known as oubiwann [13:08] ttx: goody i can slack off then :) [13:09] zul: for 5 more minutes :) [13:13] Hm. Is there a reason why vmbuilder wgets files one by one instead of grabbing a big set of them at once? The second is much likely to get large tcp windows and throughput... [13:18] ISO testing update: supposedly-final Alpha3 candidates posted on http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [13:18] let's get crackin === RoyK^ is now known as RoyK [14:03] Daviey: about the registration failure [14:03] you reported one test ok and one test failing [14:03] ttx: Reproducing here [14:03] was it under the same conditions ? [14:03] identical. [14:03] or one was netbootpreseeded and one was ISo ? [14:04] it's quite steadily failing now [14:04] so I wonder if there is not something else taht regressed [14:05] hmm [14:06] the web UI says it's already registered [14:07] Daviey: then I deregister them in the web UI and restart... [14:07] and now it says it's already registered in the euca_conf calls [14:07] * ttx sighs [14:08] ttx, oh joy [14:08] * ttx logs in again on the web UI for fun [14:08] and sure, it's registered again [14:10] Daviey: so actually it looks like though there are errors in registration.log, the registration succeeds [14:11] ... as long as you log in on the web UI ? [14:11] * ttx tries to make some sense out of that [14:12] heh [14:12] Daviey: could you try adding a node ? [14:13] * ttx will restart to test [14:13] reinstall [14:13] * Daviey notes that CD's are much slower than USB sticks. [14:16] how do I prevent a process such as apache or mysql from starting next time I reboot my computer? [14:17] anyone have experience with DoS on ubuntu server? === jjohansen is now known as jj-afk [14:20] ttx, Daviey: so auto-registration is failing? It failed on my all-in-one install [14:21] hggdh, Will confirm in a few mins my results [14:21] hggdh: well "failing" might not be the right word [14:21] hggdh: euca_conf seems completely borked and reports crap [14:21] and since we trust it to assess success/failure... [14:22] ttx, Have you succeed in registering an image before there is an NC avaliable? [14:22] I did nothing and it fell into place [14:22] Daviey: hav,'t tried [14:22] oh joy, joy, joy [14:23] unfortunately I'm completely useless due to major headache today [14:23] http://pb.daviey.com/7B0J/raw/ [14:23] I had that yesterday aswell.. but it fixed itself [14:23] ttx: in my case it actually failed to register on install (registration.log shows the failure). manually --register-* succeeded [14:23] weirdest thing is the SCS *did* get registered, even without a cluster [14:24] of course, it was not reported registered until I registered the cluster [14:24] I'm thinking it's pot luck, at the time you fire the command :( [14:24] hggdh: in my case registration.log shows a failure, euca_conf says "can't register, login on web UI, web UI says "it's registered" [14:24] oh different [14:25] Daviey: pot luck, with replacements, and many options [14:25] I'm not that sure. If you consider euca_conf reports incorrectly success/failure, it makes (some) sense [14:25] I agree we cannot trust euca_conf right now, but the registration log *does* show the failure [14:25] oh [14:25] if the reg.log maintained by euca_conf? [14:25] hggdh: I'm starting to suspect it actually registers, despite registration.log showing failure [14:26] ttx: ack [14:26] no, it's maintained by the uec-component-listener scripts [14:26] it just reports the return code of euca_conf [14:26] yeah, so it is unreliable as well [14:29] hggdh, I'm sure we'll get a fix :) [14:30] hggdh: apparently "registered" means different things to different pieces of the eucalyptus puzzle [14:30] Daviey: yes. Pigs *do* fly, only catch is when they land [14:30] the UI says "yes", euca_conf says "screw you" [14:30] heh [14:30] ttx, On your NC install... did you get grub asking you to check the "Linux command line" it apparently extracted from /etc/default/grub? [14:31] entry = blank [14:31] haven't installed NC yet, reinstalling all-in-one [14:31] Daviey: on install? [14:32] hggdh, yes, from the iso span today [14:32] (not preseeded) [14:32] starting one right now [14:32] will look at the serial cons of all NCs [14:32] hggdh, during debian-installer [14:33] This registering an image is interesting... [14:34] and then... registration works [14:34] ... [14:34] it's one of those days I should really consider working from hoime shouldn't prevent me from calling sick [14:35] ttx, ping. [14:35] weird. I am opening a console on all, all all show me at a d-i screen "Configuring grub-pc" [14:35] so you want some little blurb about kernel upgrades ? [14:35] ttx, Agreed - me and hggdh can keep cracking on. "Go Home" :) [14:35] hggdh, Daviey: so the workaround seems to be: keep trying [14:35] exactly! [14:35] smoser: can be done tomorrow [14:36] ... I think [14:36] ... I hope [14:36] dammit.. i intended to make the registration.log more useful in the last upload [14:36] ttx, well, http://bazaar.launchpad.net/~ubuntu-on-ec2/ubuntu-on-ec2/ami-pages/annotate/head%3A/maverick-i386.body.html [14:36] hggdh: so the onoly thing I did differently this time was to login immediately and keep doing netstat -tl to check services going up [14:36] has something (See 'Changelog') [14:37] Daviey, so what is up with grub causing problems ? [14:37] maybe it slowed down the stuff sufficiently to win whatever race [14:37] i'm thinking this is likely related to my addition of dependencies in eucalyptus-nc [14:37] ttx: I had a similar thingy yesterday -- of the 6 installs I did, 4 worked OK, two failed to register [14:38] hggdh: did they actually not work ? Or you supposed they were broken by looking at euca_conf output and ergistration.log ? [14:38] I trashed all my NCs installs, and I am reinstalling & monitoring the serial console. All of them got stuck in d-i on grub [14:38] smoser, not entirely sure.. doing a fresh install to check [14:38] smoser, certainly *not* the changes you introduced.. So don't worry [14:39] ok. [14:39] * ttx wonders if everything does not come from the new "I need walrus up first" type of constraints [14:39] hggdh, Oh good.. it's not only me [14:39] * ttx finishes his test and preps a long pause [14:40] I think grub is poorly. :( [14:40] ttx: they all worked after *manually* registering the cc, walrus, and sc [14:40] hggdh: but not before ? [14:41] smoser, Actually.. i take that back.. it could be the depends change.. i was thinking in code changes. [14:41] ttx: but not before (these two I pointed as failing) [14:41] the other fours worked perfectly with auto-reg [14:43] samba isn't in my /etc/init.d .. i selected to install samba from the ubuntu server install.. i'm wondering what i'm doing wrong. [14:43] ttx, Seems my walrus and cluster failed to auto register [14:44] Daviey: got that grub-pc question as well [14:44] ah is it called smbd ? [14:44] ttx, oh good [14:44] Daviey: see if they got registered in secret [14:44] ttx, In a few steps.. see if grub fails to install [14:44] DavidLevin: does entry=blank work ? [14:44] Daviey: ^ [14:45] ttx, I just pressed return [14:45] Daviey: ok [14:45] (which didn't work) [14:45] Daviey: please check netstat -tl [14:45] i dont understand though. [14:45] to see if you have 8773 8774 8443 and 9001 [14:45] wouldn't grub-pc have been installed before as the boot loader ? [14:45] dammit, i've just euca_conf --registered [14:45] i wouldn't have thought that would have been a new dependency there. [14:45] smoser, i agree [14:46] smoser, I don't know if it's a recent change with our stuff.. or foundations [14:46] smoser: any clue what I should answer to that boot line question ? [14:46] i didnt see it. [14:46] it wasn't there yesterday [14:46] oh. [14:46] wait. whats the question ? [14:46] Daviey: did you get a similar Q when installing the CLC/Walrus/etc? I did not [14:47] Linux command line: [14:47] ttx, On your NC install... did you get grub asking you to check the "Linux command line" it apparently extracted from /etc/default/grub? [14:47] smoser, ^^ [14:47] i think you can probably leave it blank. [14:47] i left it balnk [14:47] Daviey: I somehow regret that we introduced that version with new depends now [14:47] Daviey: and it failed, right [14:47] ttx, That might not be the issue.. ofc [14:47] my functional laptop has: GRUB_CMDLINE_LINUX="" [14:47] ttx, yes.. grub didn't handle the question well asking me which device to put it on [14:48] ttx: on the single install I let proceed, CMDLINE="" also [14:48] and it booted oK [14:48] whatever is it being asked for now [14:48] Daviey: ok, so that's another issue, trying with empty [14:48] hm.. that is strange. [14:48] Daviey: could you file a bug about that, we need to find a way to not have that question asked [14:48] i get annoyed by that in cloud images also [14:49] ttx: why did I *not* get this Q on the CLC install? [14:49] ttx, yup [14:49] hggdh, I didn't either [14:49] hggdh: it's a NC dep [14:49] ttx: what dep? [14:49] so, seriously, grub-pc is not installed on your CLC ? [14:50] * hggdh doubts [14:50] that question is coming from grub-pc [14:50] Hmm grub != grub-pc [14:50] smoser: it *is* installed, 1.98+20100722-1ubuntu1 [14:50] smoser: grub-pc is installed later in the process, by making the NC depend on it you force earlier install [14:51] ah. thats what i was thinking. [14:51] :-( [14:52] i'm sorry for the pain. [14:52] smoser, don't be [14:52] i'm going to run, thoguh, if i am needed, please call me. [14:52] smoser: makes life interesting ;-) [14:52] and if the solution is "drop out smoser's changes", i wont cry too much. [14:52] smoser, Before you go.. [14:52] so this is on preseeded install that we're hitting it ? [14:52] smoser, What would happen if grub was removed from a depends? [14:53] just assume it's already installed? [14:53] smoser, and ISO [14:53] well, grub-mkrescue will fail. [14:53] now, I did not get Q-ed on the linuxc cmdline, but on whether I wanted or not to install grub [14:53] is there any addtl firewall on top of apache in an ubuntu server? [14:53] i dont know. [14:53] i mean if everything is truely there, then it should have no problem.s [14:53] smoser, ok, cheers. [14:53] Daviey: I can't make it install [14:53] ttx, same here :( [14:53] silly me thought that putting it in depends was right, since it depended on it :) [14:53] I'm stuck at the grub-pc "upgrade" [14:54] i just started my NC install again to see [14:54] you can probably put anything in that string [14:54] ttx, Keeps telling you that you have chosen not to install grub? [14:54] its just going to get onto the linux command line [14:54] which, who cares. [14:54] smoser, That is isn't the killer issue [14:54] Daviey: yes [14:54] grub-pc/install_devices is your issue? [14:55] Daviey: that's actually a problem. The NC is completely uninstallable [14:55] smoser, Near the end.. grub asks you waht device you want it on [14:55] well, you can tell it. [14:55] or maybe not [14:55] you select a device [14:55] and press continue [14:55] oh. [14:55] :-( [14:55] grub then returns "Are you sure you don't want to install grub?" [14:55] well, i'm sorry for this. i have to run. again, if "back out those changes" is the solutuion, then thats fine. [14:56] I thought i was being a plum, so started the install again to reproduce.. [14:56] seems ttx has just hit the same behaviour [14:56] Daviey: apparently if you let it continue it works [14:56] i admit to not ever tested an full install with those packages. [14:56] ttx, Daviey: the NC *is* installable, I just installed two of them [14:56] smoser, Well it's not terribly easy to test packages at install time.. [14:56] hggdh: I was fearing it was not installable from ISO [14:57] hggdh: apparently it's just a misleading message [14:57] i made a seperate local repo for doing this.. and it's not ideal [14:57] Daviey: could you file a bug ? [14:57] ttx, yes [14:57] yes. I tried to say I did *not* want grub, and it kept on hapilly [14:57] Daviey: so that the message is in english :) [14:57] rigth, we need to document that [14:57] hggdh, crikey [14:57] well it's not as bad as we were thinking then [14:57] Daviey: in fact the last install step catches back and installs it [14:57] i just killed the install and started again [14:58] * ttx runs an instance [14:58] ttx, good luck. [14:58] oh noes, robbiew is here. [14:58] * hggdh crosses fingers and toes [14:58] everyone, make as if eucalyptus was working ! [14:59] * hggdh starts behaving nicely [15:00] hggdh: I'm using your registration bug to as a metabug [15:00] ttx: roj [15:01] instance running [15:02] hmm [15:02] i'm still getting the traceback [15:02] when registering an image [15:02] works for me [15:02] what iumage are you using ? [15:03] Daviey: have a bug number for the grub-pc issue ? [15:03] I'll mention it on my test report [15:04] ttx, I'll generate one now [15:04] ttx, Are both issues the same bug IYO? [15:05] both ? [15:05] Linux Command Line one and the "Sure you don't want grub?"? [15:05] you just shouldn't get prompted [15:05] so its the same issue [15:05] they do sound related, but I did not get the cmdline [15:05] file just one, we'll fix it by removing that depends [15:06] * ttx really needs a pause now [15:08] Daviey: let me know the bug number when its in [15:08] * ttx pauses [15:08] Any ideas how I can make ssh log into a user so it has all the users aliases and functions defined for it? [15:09] so then I can do ssh server.com bash_alias [15:09] MTecknology: search for 'environment' in sshd_config man page [15:10] MTecknology: if you put your public key inside their authorized_keys you can go right into that account. [15:11] ttx, Sorry for the delay. [15:11] bug #613463 [15:11] Launchpad bug 613463 in eucalyptus "[10.10 - Alpha 3 (candidate)] Prompts misleading grub dialogs during UEC Node installation." [Undecided,New] https://launchpad.net/bugs/613463 [15:12] Daviey: no errors registering (lucid|maverick) (amd64|i386) [15:13] I mean, images [15:13] pah [15:13] I experienced this yesterday, and it fixed itself [15:13] pmatulis: I must be missing it. I tried AcceptEnv * [15:14] ttx, Ok.. i've been infront of the computer for 9 hours now.. i need a break :/ [15:14] I also tried out PermitUserEnvironment [15:17] Daviey, ttx: I will go now for a (CLC+Walrus), (CC+SC), (NC)+ install. Who knows, the "Walrus first" may have been fixed... [15:17] hggdh: or you might get lucky, if you do the right dance [15:18] heh. [15:18] * ttx will lay down a little longer to let the meds calm down the fever [15:18] * hggdh picks up the "Great Book of Magical Dances" [15:18] I plan to run a UEC image on the cloud I luckily got running [15:18] oh... it's trying to execute the alias on the local system instead of the remote system.... [15:18] and a tomcat6/amd64 test [15:19] mathiaz_: I'll let you coordinate the rest of regular ISO testing [15:20] mathiaz_: with zul and spamaps [15:30] ttx: okdiko [15:30] ttx: good luck with your dinosaurs battle [15:38] * ttx is back, fully drugged. [15:39] ttx, Daviey s were ok [15:39] smoser: ? [15:39] ? with a wierd workaround? [15:39] so we're ok? on euca-nc install [15:40] yeah! [15:40] so no kittens have been killed :) [15:40] hs any one done euca\ [15:40] err uec-provisioning? [15:40] * Daviey gives smoser a new keyboard [15:41] smoser: yesterday, not today [15:41] smoser: But that isn't a concern for A3 [15:41] if you just took the 2 month old from my arm i might not need anothr keyboard :) [15:42] alright. me is off again. later === dendrobates is now known as dendro-afk [15:42] Daviey: I'm going to scrap my working UEC setup, unless you need it for some tests [15:43] ttx: sounds fine with me. [15:43] thanks for jumping in ttx ! [15:43] I'll run a few regular ISO tests, then be back with i386 topology1... and maybe more details on that strange registration issue [15:43] I want to be able to clearly point to euca_conf behaving strangely, keep the autoreg stuff out of the equation [15:44] so that we can push it upstream === dendro-afk is now known as dendrobates [15:49] Daviey: tomorrow morning we'll have to come up with release notes for those funny bugs [15:49] funny release notes for the bugs.. got it! [15:54] im going nuts trying to setup a NFS server that osx 10.6 can connect to. is there some sort of magicial param that i'm missing? [15:59] tesseracter: nfs3? [15:59] give it the insecure option in exports [16:00] the insecure option in this entry also allows clients with NFS implementations that don't use a reserved port for NFS. [16:00] thesheff17: You available for a little help? [16:00] /home/steven foo.bar.0.0/16(rw,sync,no_subtree_check,insecure) is the line [16:01] papertigers, ^ [16:01] MTecknology: yup [16:01] jk_cp -f /jail /usr/bin/{vim,vim.basic,vimdiff,vimtutor} [16:01] thesheff17: ^ I run those commands and then vim is available - but broken [16:03] tesseracter: try this on mac sudo mount -o -P server:/exported/path /local/path [16:03] I used to have to do that [16:05] MTecknology: haha I think because vi isn't an alias to vim :) [16:05] thesheff17: hm? [16:05] MTecknology: so vi is really vim [16:05] thesheff17: I never typed vi without the m right after [16:05] MTecknology: right I wonder where they make that alias [16:06] thesheff17: I'm considering making a really big chroot... then using that [16:06] 'which vi' [16:06] /usr/bin/vi [16:06] vi != vim [16:07] yay! bacula 5.0.3 next week...WOHOO! [16:07] right.. which is why I always type vim [16:07] MTecknology: I would just create an alias from vi to vim [16:07] hggdh: you mind if I rewrite the desc. for bug 609112 to talk about component resgitration rather than node registration ? [16:07] Launchpad bug 609112 in eucalyptus "[maverick] Registration unreliable" [High,Confirmed] https://launchpad.net/bugs/609112 [16:08] thesheff17: /etc/alternatives/vi and that is vi -> /usr/bin/vim.basic [16:08] hggdh: I think the current desc is misleading [16:08] thesheff17: did that - no difference [16:08] someone knows what could be wrong? [16:08] modprobe kvm_intel [16:08] FATAL: Error inserting kvm_intel (/lib/modules/2.6.32-24-server/kernel/arch/x86/kvm/kvm-intel.ko): Operation not supported [16:08] binBASH: sudo? [16:08] I'm root ;) [16:09] binBASH: see if $ kvm-ok , gives an insight [16:09] Ahh [16:09] thx Daviey [16:09] MTecknology: /usr/bin/vi -> /etc/alternatives/vi [16:10] So that provider disabled virtualization in BIOS, bwarg :D [16:10] heh [16:10] and then /etc/alternatives/vi points to /usr/bin/vim.basic [16:10] thesheff17: the problem is a lot of undefined functions inside vim [16:12] MTecknology: what function isn't working [16:12] thesheff17: I just gave you that vi link, it goes to vim.basic [16:12] thesheff17: E388: Couldn't find definition <- one - but I get a lot of them [16:12] thesheff17: and the display is pretty mucked up [16:12] thesheff17: perhaps vim-nox has more [16:13] papertigers: ^^ that was for you [16:13] MTecknology: maybe try vim-tiny [16:13] pmatulis: spiffy idea.. [16:13] MTecknology: /usr/bin/vi -> /etc/alternatives/vi -> /usr/bin/vim.basic [16:14] papertigers, just hangs, are there logs i can tail to see if the mac is hitting the server? [16:15] pmatulis: still messed up - but much better [16:15] tesseracter: nfs? [16:15] pmatulis: I keep thinking I'm probably missing some files that it wants [16:16] MTecknology: yes, probably some libraries [16:16] tesseracter: from the mac when you mount give it the -v [16:17] for verbose see what its waiting on === mathiaz_ is now known as mathiaz [16:19] MTecknology: /usr/share/vim/ you might want to copy all this into the chroot env [16:19] MTecknology: I wonder though how much of that is symlinked around :( [16:21] thesheff17: easy enoguh... [16:21] Why not take the vim tar ball and compile it in the chroot :) haha [16:21] MTecknology: use ldd to determine what libraries you need [16:21] thesheff17: find /usr/share/vim/ -exec jk_cp -f /jail {} +; :D [16:21] papertigers: that is what I was thinking...maybe even add sudo and apt-get into chroot and could you apt-get vim :) [16:22] thesheff17: if you had the tools in there to do that :P [16:22] papertigers, 1. -P wasn't considered a valid option, 2. v, vv ,vvv aren't giving any text at all, 3. it returns mount_nfs:cannot access /home/steven: Permission denied when it did return after the hang. === dendrobates is now known as dendro-afk [16:22] papertigers, although without -v, -P seemed to work, thats when i got permission denied. [16:22] I just made a really hefty chroot - we'll see how that works [16:24] tesseracter: hmm you are on server, do you ahve a gui or just cli [16:24] papertigers, just GUI on server. [16:24] tesseracter: okay go to finder [16:25] papertigers, heh, i meant to say just cli [16:25] and then go to file connect to server and do nfs://myip:/path/to/share [16:25] /facepalm [16:25] tesseracter: ahh okay, I was gonna say I know the gui way works on my mac [16:26] thats frustrating... [16:28] there we go..... [16:28] papertigers, now youve got me confused. the server is on ubuntu, the client is a mac. the finder connect to server thing hasn't worked....although it worked a while ago with a different guys laptop. it just hasnt been consistent between macs. [16:28] thesheff17: pmatulis: Fixed... debootstrap lucid dir; chroot dir; aptitude install vim [16:28] :D [16:28] MTecknology: sweet :) [16:29] less secure chroot becuase you can do more in it.. but... [16:29] it's working [16:30] what's the stable version of debian? [16:30] lenny.. there we go [16:30] If I go with a complete chroot, I may as well go with one that doesn't change much [16:31] is there anything in the works for making ubuntu virtualisation work like vmware or hyper-v where you can setup 2-3 boxes and have them work as a HA-cluster without anything in front? [16:37] Any ideas how to stop these messages? http://dpaste.com/224964/ [16:38] heh... simple enough [16:40] pmatulis: still doesn't work exactly right - vim acts more like vi [16:41] pmatulis: actually.. it acts more like some of the keys are messed up.. I press I, Delete and the position turns into a ~ and then the letter I was on comes back and the case of the previous letter changes [16:43] MTecknology: why not set up a machine for this express user-login purpose? i still don't know much details of what you're doing [16:44] pmatulis: hm? [16:44] pmatulis: just setting up user jails [16:44] MTecknology: seems like it not working out [16:44] pmatulis: my question about ssh is completely different [16:45] It seems like it's something with escape sequences of special chars [16:47] I like Delete but if I can't use it I guess that's fine - it seems like that's the only issue with it [16:49] mathiaz: is kirkland with you today ? [16:49] mathiaz: see why I want a presence tool :) === dendro-afk is now known as dendrobates [16:51] MTecknology: I wonder if vim is just confused on what keyboard you are using. I have seen del and backspace both break on vim before [16:53] ttx: yes - he is busy === jj-afk is now known as jjohansen [16:53] mathiaz: right, he synced with me [16:53] ttx: good :) [16:54] thesheff17: any ideas hwo to fix it? [16:56] Daviey: got my i386 setup working the first time I tried [16:57] ttx: you should enter the Lotto [16:57] sounds like your numbers came in! [16:59] SpamapS: please coordinate with mathiaz to get some ISO testing coverage today [16:59] * ttx completes one last test and jumps directly to bed to heal that headache [17:00] SpamapS: I'd suggest you to start with the i386 images [17:00] SpamapS: as I'm testing amd64 right now [17:00] mathiaz: we'll need coverage on the upgrade and netboot side as well [17:01] ttx: upgrade? [17:01] ttx: we're already testing upgrade in alpha3? [17:01] New bug: #613309 in cloud-init (main) "upstart scripts do not wait for 'cloud-config' status" [Undecided,New] https://launchpad.net/bugs/613309 [17:01] mathiaz: dunno, it shows up in http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [17:01] ttx: I thought upgrade started to be tested at beta time [17:01] mathiaz: see with ara if it's superfluous ? [17:02] ttx: ok - I'll sync up with ara [17:02] ttx: have a good night [17:02] not before I can start an instance on that UEC :)à [17:02] upgrading uec from lucid was known to be broken [17:03] The recent upload has hopefully addressed this, but i haven't tested it [17:05] MTecknology :fixdel does that help at all? [17:06] arh [17:06] instance automatically shutting down [17:06] ... [17:07] thesheff17: *blink* :D [17:07] let's say its a dupe of bug 610479 [17:07] Launchpad bug 610479 in eucalyptus "Instance fails to start" [High,Confirmed] https://launchpad.net/bugs/610479 [17:09] MTecknology: so del is the only key not working? Sorry a developer was bugging me about something :) [17:09] thesheff17: I think so.. that's kinda interesting [17:10] MTecknology: http://vim.wikia.com/wiki/Backspace_and_delete_problems [17:11] thesheff17: thanks :) [17:12] ttx: nope, you cannot manually register the walrus on a separate install -- euca_conf requires the CLC to be running [17:12] ttx: so... this really breaks eucalyptus on distributed installs [17:13] yeah, I've run into those issues with backspace and delete on solaris, though I've never run into them on linux [17:13] good document there [17:16] hggdh, Daviey: ran into bug 613514 trying to start instances on my i386 UEC setup [17:16] Launchpad bug 613514 in eucalyptus "Instance does not start" [Undecided,New] https://launchpad.net/bugs/613514 [17:16] and now eod [17:18] ttx: I am going back to an all-in-one, and will check it [17:23] Registering images would be nice :) [17:27] Hello, I'm trying to install php5-mcrypt in ubuntu server 10.04 - but I'm getting this message: The following packages have unmet dependencies: php5-mcrypt: Depends: phpapi-20060613+lfs -- However http://packages.ubuntu.com/lucid/php5-mcrypt says the phpapi-20090626+lfs is a virtual package provided by packages I already have... So whats going on here? [17:32] Can anyone help me install php5-mycrpt? [17:33] Does anyone know how I can query a DHCP server without having any of the settings take effect on my system but where I can have it display everything it says? [17:39] I forgot a hard link can't cross partitions [17:39] fooye :P [17:45] mathiaz: ok, will start i386 stuff. :) [17:46] SpamapS: hm - well - I'm about start i386 tests [17:46] SpamapS: I'll be finished in 30 minutes [17:46] SpamapS: http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [17:47] Daviey: still awake? [17:47] SpamapS: netboot seems to be the main part missing test coverage [17:47] hggdh: sadly [17:47] SpamapS: I'd suggest to take the amd64 netboot tests [17:48] SpamapS: and check whether you can run them [17:48] mathiaz: I'm pretty sure ttx said netbooting isn't a priority for A3 tests [17:48] Daviey: have you ever heard of a programme called 'bttrack'? I see an error in the cloud-debug.log stating it cannot find it [17:48] Can anyone help me figure out why ubuntu server refuses to install php5-mcrypt? [17:49] nicetry: what is the exact error you get? pastebin it if necessary [17:50] hggdh: interesting.. that is a torrent client [17:50] mathiaz: I don't have any way to test net boot [17:50] BTW was there some email I missed saying the iso's were out? [17:50] hggdh: I suspect that is a new feature he hadn't yet heard of [17:51] Daviey: I know of 'bttrack.bittorrent', but not of just 'bttrack' [17:51] at least we are not packaging anything with this name right now for Maverick [17:51] hggdh: hmm.. it's provided with both bittornado and bitttorent packages [17:52] hggdh: http://pastebin.com/VZdkYmgb [17:52] hggdh: Thankfully, they are both main - so it's not a biggy if we need to add them.. [17:53] hggdh: I sort of expect it's a feature of distributing the images via torrent to the nc's.. but this is nothing more than a hunch [17:53] SpamapS: I'm subscribed to test cases in the iso tracker [17:54] SpamapS: that's how I get notified whenever new images are waiting to be tested [17:54] SpamapS: The ISO's aren't out.. we are testing the candidates :) [17:54] SpamapS: http://testcases.qa.ubuntu.com/Install/NetBoot [17:55] SpamapS: ^^ this outlines net boot [17:55] SpamapS: it turns out that netboot is actually about testing the mini.iso === lag is now known as Guest17866 [17:56] hggdh: Any ideas? [17:56] I guess mini.iso is pretty close to a PXE enviroment, if you can't change your dhcpd settings [17:56] Although.... a valid test could be using gpxe [17:57] mathiaz: ah [17:58] I have a laptop [17:58] nothing else [17:58] not really interested in risking accidentally wiping it with a pxe install. ;) [17:58] SpamapS: virtulisation baby! [17:59] kvm and virtualbox can both PXE boot :) [18:00] SpamapS: time to get some more hardware [18:01] zul: I spent years kicking all the hardware out of my house [18:01] Daviey: right - we should add some pxe testing as well [18:01] its nice and quiet [18:01] and cool [18:01] SpamapS: I hear you ;) [18:01] SpamapS: heh....well different career path requires different hardware :) [18:01] I do have an old G5 mac but I doubt people care about powerpc testing. :) [18:01] SpamapS: I was able to setup a netboot installation environement using vm only and libvirt [18:01] mathiaz: i can't, over the noise of these servers [18:02] SpamapS: I have a virtual network within libvirt where i have an pxe server running [18:02] SpamapS: that how I can test network installation within my laptop [18:02] zul: true, I have the whole cloud at my fingertips though.. who runs real servers anymore anyway? ;) [18:02] nicetry: what version of Ubuntu? [18:02] SpamapS: umm...when you are needed to test isos perhaps? ;) [18:03] Its an interesting idea to create a virtual netboot test env. [18:03] seems like cheating. ;) [18:03] hggdh: 10.04 [18:03] hggdh: but this also happend before I upgraded in 9.10 [18:05] mathiaz: so I shouldn't bother downloading the i386 iso then? [18:05] * SpamapS only has the amd64 isos [18:05] SpamapS: nope [18:06] SpamapS: if you have amd64, you could also run through the raid1 install: http://iso.qa.ubuntu.com/qatracker/result/4389/286 [18:06] SpamapS: and report it to http://iso.qa.ubuntu.com/qatracker/test/4389 === nicetry is now known as thecaptainmurphy [18:07] nicetry: phpapi-20060613+lfs does not exist on Lucid === thecaptainmurphy is now known as nicetry [18:08] nicetry: but there is phpapi-20090626+lfs [18:08] nicetry: are you sure you are using the Lucid repositories? [18:09] Read maverick-server-amd64.iso. Target 41.7% complete. [18:09] Wow we changed a lot. ;) [18:11] hggdh: no! the universe/multiverse were still karmic for some reason [18:12] hggdh: thank you - that did it [18:14] nicetry: welcome, glad to be able to help === luis__lopez is now known as luis_lopez [18:17] would anyone suggest a reverse proxy? I know there is squid, pound, apache, nginx and basically just looking for high performance active checks on the backend and sticky persistence. [18:21] nginx doesn't do sticky persistance, IIRC [18:21] Jeeves_: thx [18:22] Oh, wait. [18:22] It does [18:22] It was varnish that didn't [18:22] ah ok...I just actually started looking at that. [18:30] varnish is really high performance [18:30] and I believe while it doesn't do session persistence, it does utilize keepalives on the backend connections [18:30] haproxy is also really nice [18:31] IMO, needing sticky persistence means limiting your application quite a bit. [18:31] thesheff17: Do you really need sticky persistence, or just think it will improve performance? [18:35] SpamapS: well I did in the past but the developers might have fixed their crappy cache system on the code side :) [18:35] I get this http://dpaste.com/224992/ from doing this --> sudo -s /bin/dash -n -H -u demo.kalliki.com -i -- yes y | /usr/local/sbin/drush/drush -r /jail/home/demo.kalliki.com/drupal -l demo.kalliki.com up [18:36] GRR [18:36] thesheff17: www.memcached.org :-D [18:36] thesheff17: I'm trying to make something happen outside of the chroot as that user [18:37] SpamapS: thx I sent it to the developer...I'm sure he will love that link :) [18:38] MTecknology: what are you trying to do? [18:39] thesheff17: honestly, if they are in a web environment, and they haven't heard of it yet.. they're fired. ;) [18:39] thesheff17: I want to run a php process as that user - but I want to not be confined to their chroot [18:40] thesheff17: if I could change their shell to /bin/bash and their home dir to /jail/home/user instead of /jail/./home/user it might work.. [18:41] MTecknology: -f you drop -H, you won't inherit the user's whole login environment, you'll just get their uid/gid [18:41] s/-f/if/ [18:42] MTecknology: you could write some script to check for something within the chroot to let them run something on the main server. [18:42] SpamapS: I tried without the -H too [18:44] root@incipio:~# sudo -s /bin/bash -n -u demo.kalliki.com -i -- "/usr/local/sbin/drush/drush -r /jail/home/demo.kalliki.com/pressflow -l http://demo.kalliki.com up" [18:44] /bin/bash: demo.kalliki.com: No such file or directory [18:58] I'll have to finish fighting this when I get back. === mrjazzcat-afk is now known as mrjazzcat [19:10] is there a channel to discuss type 1 virtualization with ubuntu ? [19:11] axisys: type 1? [19:12] pmatulis: like xen .. hypervisor.. not virtualbox (type 2) [19:12] axisys: try #ubuntu-virt for any "types" [19:12] pmatulis: thanks [19:20] Daviey: rejoyce! I ran 81 instances on an all-in-one + 5 NCs, and 78 succeeded! [19:20] hggdh: \o/ [19:20] yeah... who would have said... [19:20] hggdh: what was your sleep interval between starting instances? [19:21] Daviey: the usual ramdon between 5 and 11 seconds [19:21] ok, thanks [19:21] hggdh: If you fancy testing it when you go to bed, fancy leaving it at ~60 seconds? [19:21] on my rig, this guarantees I will be usually getting all possible instances to run [19:21] (or more) [19:21] Daviey: yes, no prob. Any special reason? [19:22] hggdh: I am interested to see if it's just a bad timing issue, or if spawning too fast causes it [19:22] i suspect it's the luck of the draw issue we've been experiencing with a few parts of the stack [19:22] Daviey: oh. OK. I will change the test not to do a -n on euca-run-instances [19:23] super! [19:23] Daviey: how many? Fancy 1000? [19:23] (we will be sleeping, after all) [19:25] hggdh: while true: [19:25] :) [19:26] 99999999999999999999 should cover it.. the kill it when you start tomorrow [19:26] is that ok? [19:27] yes, no prob (but I could never run more than 10k before having som many errors that I would kill it [19:27] MAX_INT64 it is [19:28] Oh, sorry - pick a suitable value then :) [19:37] running the beast now, Daviey [19:38] (we actually will not get much done, one every 60sec == 1440 in 24 hours) [19:40] yeah :/ [19:40] thanks hggdh ! [19:43] Daviey: my pleasure. I believe you ar now due for bed, correct ;-)? [19:43] hggdh: yeah right! [19:44] sleep is a commodity that Ubuntu doesn't like :) [19:44] heh === mrjazzcat is now known as mrjazzcat-afk === _Techie_|_Away_ is now known as _Techie_ === _Techie_ is now known as _Techie_|_Away_ [20:41] how can you resize a XFS partition? === mrjazzcat-afk is now known as mrjazzcat [20:44] you can only extend it [20:44] and, notice that XFS is file system, not a partition [20:53] can i use ftp client on a ubuntu server? [20:53] hy there. anyone know how to configure tomcat on 8.04 to listen in ipv4? it only listens on ipv6 :( [20:53] Sander_: there is the ftp - CLI client [20:53] so why not? [20:53] Sander_: lftp is a great CLI client [20:54] apt-get lftp ? [20:54] iowahc: it should listen on ipv4 by default [20:54] iowahc: if you have ipv4 address [20:54] Sander_: you already have ftp client installed [20:54] Sander_: just type ftp and hit enter [20:54] ivoks: well. should :D it doesn't [20:55] done [20:55] but i don understand that ftp client :$ [20:55] you cannot shrink the xfs filesystem? [20:55] Kaffien: nope [20:55] that is retarded [20:55] Sander_: open ftp.server.com [20:56] thanks <3 ;p [20:56] ftp --help would help [20:56] er... man ftp [20:57] iowahc: i have 0 experience with tomcat configs :/ [20:57] np. thx anyway :) [20:58] Where can I find the Sun JRE package for Ubuntu Server? I thought it used to be in the repos as sun-java-jre-whatever but I don't see any sign of it in karmic [20:59] erm [20:59] AlexMax: sun-java5-jre [20:59] make that lucid [20:59] apt-cache search sun jre [20:59] or sun-java6-jre [20:59] !java [20:59] To install a Java runtime on Ubuntu, see https://help.ubuntu.com/community/Java. For the Sun Java products and browser plugin, search for the sun-java6- packages in the !partner repository on Lucid (which must be enabled), or !multiverse repository on older releases. [20:59] AH! [20:59] partner repos [21:00] AlexMax: you need the partner repo enabled [21:00] i don't have it enabled [21:00] but i have sun jre [21:00] hm? [21:00] oh, i have :D [21:04] thanks [21:04] Daviey: I just noticed you applied for bug-control last year (!!) [21:05] hggdh: Yeah.. probably did [21:05] heh reason i said karmic was that this box has been upgraded from karmic to lucid and parts of the sources (commented out) still said karmic :P [21:05] sources.list [21:05] AlexMax: this hould not have happened on a kosher upgrade [21:06] hggdh: Decided to just be a member of ~ubuntu-dev instead :) [21:06] and got your -control as bonus :-) [21:07] Daviey: OK. I am considering cleaning up the cruft on the -control proposed list... I will start by declining you (since you have it via -devs) [21:07] hggdh: feel free :) [21:07] * hggdh goes on. One down, 300 to go === _Techie_|_Away_ is now known as _Techie_ [21:11] Daviey: interestingly, with interval betwee submissions ~60 sec, I already got 11 failures in 81 tries [21:13] pah [21:13] hggdh meet LP-API [21:13] i'm pretty sure could could auto decline if left idle for 30 days :) [21:14] yes... but we did not have it, and there are requests pending from 2008 or earlier... [21:14] I am considering writing a hack to clean it up [21:14] hggdh: Oh aye.. i'm just saying rather than spend the time declining them manually.. the same time on a little python.. :) [21:15] :-) [21:17] hello. i'm running 10.04 server i'm wondering if there is any issue with running ACL's on my filesystems. i'm using the server as a web-hosting server. in particular LAMP services [21:21] I have a setup with users authenticating via LDAP. How can I force a password change on first logon? [21:44] RyanP: Set the "lastChanged" field to 0. [21:45] ...or whatever it's called. [21:47] soren, Thanks. What would you consider to be the easiest way to do that? [21:49] RyanP: Using whatever tool you use to deal with other aspects of user management. [21:49] what user does apache run as in ubuntu 10.04 lts server by default/ [21:49] * ? [21:50] nm it's www-data [21:50] idlemind: www-data [21:54] SpamapS: any progress on the raid1 iso test? [21:55] mathiaz: My internet connection went down 2 hours ago, just came back up so I am just now finishing the download and starting on it. :( [21:55] SpamapS: ok - cool [21:55] SpamapS: are you doing amd64 or i386? [21:55] mathiaz: I have the amd64 iso downloaded... so I'll start there. i386 has 11m left on the download [21:56] SpamapS: 11 *months* - where do you live? [21:56] mathiaz: Alpha Centauri [21:56] mathiaz: once the bits start flowing, it will be over in 5 minutes.. but the latency is AWFUL [21:56] SpamapS: no wonder that your internet connection breaks down once in ahiwl [22:16] Anyone have experience with http://libslack.org/daemon/ ? [22:16] libslack? [22:17] heh [22:17] comron: guess not [22:17] EvilPhoenix: heh. It was worth a try :) [22:19] Hey all [22:21] I was wondering if anyone had some thoughts for me...I have a Dell Rackmount server (2850) and it's on the network with two IPs and receives pings, ssh, etc, but cannot connect out, no ping, looks like no dns [22:22] can't even hit the gateway [22:23] n00b3rz: can it ping itself? both IP adresses? [22:24] guntbert: Yes it can [22:26] n00b3rz: please pastebin the output of: ip ad and of: ip r [22:27] I'm using static IPs, under /etc/network/interfaces I have auto eth0 [22:27] iface eth0 inet static [22:27] should the first line have static eth0 ? [22:27] n00b3rz: no, those are actually commands you should enter [22:28] !pastebin | n00b3rz [22:28] n00b3rz: For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://tinyurl.com/imagebin | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [22:29] http://pastebin.com/mn32WyBr [22:30] thanks BTW [22:30] I've been googling around and haven't run across anything that seemed to describe my situation, my google-fu must be weak :-| [22:33] n00b3rz: 1) whats the idea of 2 nics in the same network? 2) you should always have only one default route [22:34] guntbert: I don't believe that #2 is true. its perfectly fine to have multiple routes to any destination [22:35] as long as the gateway knows how to get to the destination [22:35] guntbert: 1) No point right now, only one network to work with and I believe one is unplugged. 2) I can see about disabling the second default route if necessary [22:35] that could be it though, if one cable isn't connected and that default is used first...I did try swapping cables and it still had no connectivity [22:35] n00b3rz: could be :-) [22:35] if one is down, the routing layer won't use it. [22:36] n00b3rz: why is one disconnected? [22:36] yea, that fixed it, thank you guntbert and SpamapS [22:36] SpamapS: in my experience multiple default routes lead to trouble [22:37] !yay | n00b3rz [22:37] n00b3rz: Glad you made it! :-) [22:37] guntbert: yeah, it doesn't do what most people think it does, but it should work fine. [22:37] SpamapS: I only have one place to connect the server to [22:37] n00b3rz: you can put two IP's on one interface, if you need both IP's [22:37] only one network here at work, only need one IP too [22:38] SpamapS: not with default route - it doesn't [22:38] well then stop doing weird stuff with two IP's ;) [22:38] hehe, good to know! [22:38] I'm on my way to bed - glad it works now :-) [22:39] if I want to disable the second interface would the best place be under /etc/network/interfaces? [22:39] removing the entry for that connection? [22:39] n00b3rz: yes === ND-movie is now known as NightDragon === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates [23:38] http://penguindroppings.wordpress.com/2010/08/04/fsck-mountall-var-and-lucid/ [23:38] ugh [23:38] we gotta fix stuff like that :-P [23:40] I would like ubuntu without plymouth as a possibility [23:40] or a "server mode" for plymouth [23:40] especially cause I never see my server boot anyways :) [23:40] When you are watching your server boot.. [23:41] its usually because something hit the fan