[00:10] hey guys im trying to build a home server just for file sharing and whatnot, can anyone give me a hand im having a lot of trouble finding good guides online [00:11] New bug: #588523 in autofs (main) "Autofs5 fails on cascading/hierarchical mounts" [Undecided,New] https://launchpad.net/bugs/588523 === Guest27712 is now known as bogeyd6- [00:23] GuyCanada: how familiar are you with Linux and command line? [00:24] if you just want to share files http://www.howtogeek.com/howto/ubuntu/install-samba-server-on-ubuntu/ is a good guide [00:34] vladimCK im fairly fammiliar, no expert tho [00:35] what do you want to use the server for? [00:36] GuyCanada: ^ === WALoeIII is now known as Roundabound === Roundabound is now known as Roundabound[yout === Roundabound[yout is now known as Roundabound [01:22] I have changed the physical sata port that my primary drive is plugged into, and now I am being dropped into an initramfs prompt [01:23] how do I a. determine the uuid now that the drive is moved and b. get the system to boot from it? === JanC_ is now known as JanC [01:48] tyler_d8, #ls -al /dev/disk/by-uuid/ [01:48] or #blkid [01:49] hey everyone quick question, where do i go if I want to submit something to the software center? or just request an update? [01:52] is there a way to limit the amount of bandwidth an ethernet interface can use? [01:55] how doesn't the server version have a desktop environment? like gnome? [01:56] because its a server version, its done to remove the weight and reduce the potential security threats [01:56] if u need a gui try a webgui like ebox [02:01] wow, so the server admin has to run everything from the cli? [02:02] it's easy to install a GUI on top if you so wish. The Ubuntu server is built out of the same repository as the Ubuntu desktop edition. [02:02] just a question of defaults. [02:04] I've installed ubuntu server as a virtual machine on virtual box but now every time I sudo shutdown now, I'm reffered to a menu rather than system shutdown. [02:08] anyone ever setup a cert for their imap mail? === Nigel_ is now known as G [02:27] anyone have experience with the "disk utility" -- I am getting an error on connect -- Unexpected stderr output - expected `udisks-tcp-bridge: Waiting for secret' but got `stdin: is not a tty' [02:45] k so we are one step further, just now need to get the usage details for the udisks package, assumedly the command udisks-tcp-bridge is supposed to be somewhere within the daemon [02:45] ?? [03:13] Hi, I have a website that sends out automated messages to registered users. I'm getting blocked by just about every other email provider out there. I've removed myself from the spamhaus blacklist, and cannot find my IP's on any other list (I checked dnsbl.info as well). [03:13] I also set up an SPF record for my domain. What else can I do? [03:19] looking for the optimal mysql db raid-5 stripe size? [03:20] savid: you may need to e-mail spamhaus to ensure your ip address or range is listed within "commercial" not "private" [03:21] tyler_d8, I put my IP in the spamhaus tool, and it says it's not blocked [03:21] thats not what I said [03:22] ok, just figured that would accomplish the same thing [03:23] Does the reverse-dns IP for the domain have to match the originating server? If so, not sure if that's even possible for my setup [03:28] tyler_d8, ok, so I need to email spamhaus -- I can't find any method for contacting them by email on their website. Is this something you've done before? [03:29] savid: had to get my ip changed in their system as some isp's block based on where your ip is linked too, commercial or residential [03:29] cannot remember the exact details so I apologize [03:30] tyler_d8, so that's not the same as getting removed from the SBL/PBL/etc.. ? [03:30] because that's what I did -- I also checked on dnsbl.info and I'm not appearing on any blacklists. [03:34] whats a good benchmark utility for my raid array ? [03:35] hi there, I need to install in a server a mail client to send email. Somebody know how to do that, a good manual? [03:40] Roxyhart0 try zimbra [03:40] it's easy :D [03:40] I will check thanks, it works with command lines? [03:41] Anyone know of a good, inexpensive SMTP relay? [03:41] just extract and answers some questions then do the rest with web [03:44] i only know to use zimbra as mail server but i don't know smtp relay [03:48] whenever I run drive tests on my raid 5 I get kernel panic errors, and ultimately a frozen machine [03:48] ?? [03:48] the drives are all healthy, and the OS is on a separate drive(not on raid) [04:07] Hi, I was wondering if xubuntu with gdm killed would be lighter on ram than Ubuntu-server? [04:08] ubuntu-server, the thing with no gui ? [04:08] ? [04:12] basically, is there any differences between different variants of ubuntu besides desktop environment. [04:14] no, they are all the same packags. [04:15] ah, good, thanks [04:15] Of course, some packages are EOLed earlier than others. [04:16] my box with only 256 MB ram running xubuntu with GUI killed uses only ~100 MB RAM, however my -server edition is using something like 450MB RAM on another box.. of course it has more tasks.. [04:17] I'll go with -server and see what happens [04:18] thanks [04:19] Datz: it's trivial to find out what is using up memory, and either remove, replace or reconfigure it. [04:21] yea [04:21] I hope it will be trivial if I run into that problem [04:22] how come I can't find server edtion torrent here: http://torrent.ubuntu.com/ [04:22] 10.04 that is [04:25] also, I'm not able to donwload the 32bit version from the main page [04:25] Presumably nobody bothered to create a torrent for it. [04:25] I always use my ISP's mirror, since that's not counted against my monthly cap. [04:25] torrents are one of the least efficient ways to download things [04:26] hugely efficient for the torrent publisher [04:26] lifeless: I contend that is not necessarily true. [04:26] I was just worried about getting throttled by the download server [04:26] but since I can't find a download yet, that's not my biggest problem [04:26] :) [04:27] this is where the 32bit download takes me: http://quigonjinn03.appspot.com/www.ubuntu.com/start-download [04:27] I just get a blank screen [04:27] If you have gigabit within campus and a shared 10MB link to the US, and four other people are seeding from within campus, then it would be faster to torrent from your local peers than to use a bulk-bucketed HTTP GET over the 10MB line. [04:28] Datz: what is your country code? [04:29] In US [04:29] +1? [04:29] I meant the two-letter code, which would be .us [04:29] oh [04:30] You want Ubuntu Server 10.04 for x86-64, or i386? [04:30] i386 [04:30] and yes 10.04 [04:31] Argh, it seems that archive.ubuntu.com's CD/DVD images aren't hosted in the same place on us.archive.ubuntu.com as they are on my ISP's mirror [04:31] humm :-\ [04:31] I can always get it some other time I guess [04:32] You could always use http://us.archive.ubuntu.com/ubuntu/dists/lucid/main/installer-i386/current/images/netboot/mini.iso [04:32] That's a generic installer that downloads packages on demand, rather than up front. [04:32] ah [04:32] interesting [04:32] netboot.. humm [04:33] would I be able to remotely install? [04:33] The machine is in one place, and you're in the other? [04:34] it is tucked away between the back of my desk and the wall :P [04:34] During the install, you can choose to install an SSH server in memory, and then perform the rest of the install via SSH [04:34] (All install media, except the "desktop" CDs, support that.) [04:34] ah, so I need a monitor to do the inital setup? [04:35] Datz: or a serial line [04:35] Or, to provide some answers in advance ("preseeding"), but that's probably too much effort for you [04:36] ah [04:36] perhaps [04:36] maybe I will just move some stuff around [04:36] Debian's installer is a good deal cleverer than Ubuntu's silly wrapper on it [04:36] humm [04:37] well, I've saved link, not going to do install tonight, so I'll check if download is working in the morning :) If not I'll try the netboot/mini.iso [04:37] brb [04:37] ok [04:58] excellent, so my motherboard is fried in my one machine(server-64) and my backup I have found out now need to be upgraded(in the process now) but as well has gnome-packages installed, what is the most effective way outside of re-installing to switch to ubuntu-server? [05:03] tyler_d8: that was confusing [05:04] lol, sorry [05:04] You have a host with GNOME installed, and you want it to be a server? [05:04] I believe its the entire ubuntu-desktop package installed [05:04] based on what I am seeing within the upgrade logs [05:05] Well, either don't run, or remove, those packages. [05:05] The difference between -desktop and -server is largely the set of packages that are installed by default [05:05] so I would be safe in doing sudo apt-get remove gnome-desktop [05:05] If you're not using it, sure [05:06] I would probably do: [05:06] aptitude --schedule-only markauto ~i; aptitude --schedule-only keep ~aremove; aptitude markauto ubuntu-desktop [05:06] And carefully review the packages that it offers to remove, in case you DO need any of them. [05:09] very good, ty [05:18] twb, I like the look of that, not being a big aptitude user - what does the ~aremove do ? [05:18] ~a means "what's gonna happen" [05:18] So ~aremove means "packages scheduled for removal" [05:19] so, it's not just I having the problem of not being able to download ubuntu-server? [05:19] aptitude-doc-en has a chapter on such patterns. I highly recommend learning them, since the alternative is usually a tedious chain of grep-dctrl (or worse, dpkg -l/apt-cache and sed/awk/perl). [05:19] Datz: I've no idea [05:20] ok.. can someone else check? :P [05:22] it would be most excellent [05:22] twb, thanks - I've generally just used apt-get and apt-cache, dpkg -l with grep etc. my awk got lost about 10 yrs ago (have used it thoughand i've only ever used what others have sed for me ;-) [05:23] oops, dropped a ) [05:23] RoyK: FYI, the RAID5 sync finished and is [UUU] this morning [05:24] hi there i want to configure in my cups server a printer pdf, somebody have any good link to do that [05:31] twb: Found the torrent I was looking for. I don't know how I could get to it from the linkI posted before FYYI http://torrent.ubuntu.com:6969/ === twister23_ is now known as twister23 === twister23_ is now known as twister23 [07:08] anyone setup dovecot+ssl? [07:09] I have been working at it all day with no luck [07:15] Elad: did you read the ubuntu server guide? [07:15] I read the dovecot guide [07:17] twb, looks like they have some stuff in the ubuntu guide - reading now [07:19] hi [07:19] I have deleted the boot file on Ubuntu 8.04 server, is there a way to regenerate it again. I did rebooted and it did not created it, Also I touched the file and there were no entries in that file after reboot. I have followed also http://ubuntuforums.org/showthread.php?t=49925, it did not worked either. Please suggest/guide [07:20] kaushal: which boot file? [07:21] twb: /var/log/boot [07:21] kaushal: that file is created at boot by /etc/init.d/bootlogd, iff /etc/default/bootlogd contains BOOTLOGD_ENABLE=Yes [07:22] It doesn't matter if the file exists beforehand [07:22] However, it might be that Ubuntu has broken that functonality in recent versions, since it's largely a sysvinitism and Ubuntu moved most stuff to upstart since 8.04... [07:23] Oh, you're on 8.04. Carry on. [07:25] is there a command to get your external ip? or do i have to setup a .sh script? [07:26] corpse: WHICH external IP? [07:26] And, no, not portably. Conventional wisdom is unfortunately to postprocess the output of "ip address show dev eth0" or some variation thereof. [07:26] twb: my servers webhost ip [07:27] corpse: www.whatismyip.com [07:27] corpse: so when you have zero NICs, or fifteen NICs, or one NIC with three IPv4 addresses and two IPv6 addresses, which address do you want? [07:27] wise_crypt: oh, right, maybe he meant that. [07:28] yeah thats kinnda what im talking about, but i need to run that on my server [07:28] i have set up some .sh scripts based on what i have found on google. but none see to work as they are all a few years old [07:28] w3m or lynx [07:29] that was it =D thanks alot! [07:29] didnt know about the w3m command [07:30] corpse: dyndns can do that for you www.dydns.com [07:31] corpse: dyndns can do that for you * www.dyndns.com [07:40] so I just got done reading the ubuntu guide, and it didn't answer my question [07:41] although it is possible I didn't ask it the right question [07:43] I can connect fine to dovecot+ssl, but I get a warning of an untrusted cert. I change the cert location in dovecot.conf to my qualified certs, but my mail client still sees the old certs; makes me think I did something wrong [07:43] Did you bounce dovecot after reconfiguring it? [07:43] yes /etc/init.d/dovecot restart [08:24] ttx: question about cassandra's huge pile of dependencies.. [08:24] SpamapS: yes [08:24] ttx: It seems I'll probably have to replace w/ symlinks. [08:25] ttx: but one in particular is tricky... antlr3 [08:26] ttx: we package 3.0.3 in main, I don't see a merge in merge-o-matic .. but cassandra wants 3.1.3 [08:26] SpamapS: welcome to hell [08:26] SpamapS: does cassandra want 3.1.3, or need 3.1.3 ? [08:27] also I'd recommend patching the build.xml rather than using symlinks, if you can [08:27] it's cleaner than reconstructing symlinks on package unpack [08:28] (source package unpack) [08:29] * ttx grumbles [08:36] SpamapS: there are several options... [08:36] SpamapS: (1) Cassandra builds with 3.0.1 alright -> DONE [08:36] (2) Cassandra can be tweaked to build with 3.0.1 -> PATCH -> DONE [08:37] (3) Update antlr3 to 3.1 -> prep package, test all reverse build deps to see if they accept 3.1 instead of 3.0.1, upload as the new antlr3, depend on that -> DONE [08:38] ttx: one of the reverse deps is eucalyptus-java-common .. ;) [08:38] (4) If some require 3.0.x and the others require 3.1.x... [08:39] antlr is a build dep iirc, so it's quite easy to test [08:39] ttx: ok, my preference was to just build w/ 3.0 .. but the cassandra guys are somewhat militant that we should build with the version as close as possible to the one in lib [08:40] ttx: yes its a build dep, but it is a code generator, so it may generate different code ... [08:40] SpamapS: if it builds with 3.0, the plan would be to just build with it... and ask debian how open they would be to an update [08:40] it's a question of how many revdeps [08:41] SpamapS: that's a calssic case where Java collides with distributions [08:41] SpamapS: distributions try to have as little different versions of the same libs as possible [08:41] SpamapS: Java actually encourage devs to pick their specific version [08:41] and maven made it even worse. [08:42] ttx: thats annoying. :-P [08:43] SpamapS: it's funny the first time it hits you. Over the long run, you just wish java didn't exist. [08:57] SpamapS: do you know how to do ISO testing ? [08:57] http://iso.qa.ubuntu.com/qatracker/build/ubuntuserver/all [08:57] That's one of the sevarl things we also have to do :) [08:57] several, even [08:57] Upstream are jackasses, film at 11 [08:59] twb: I wouldn't say that. But in some cases, the way tehy are used to work doesn't play nice with the needs of distributions. [08:59] When they respond to that difference in the style of schily or tuomov or moco, they are definitely 100% jackasses. [09:00] right :) [09:11] ttx: Does it count if I use the iso to install in VMware? [09:12] SpamapS: if you mark so in the comments, that's ok [09:12] SpamapS: I use KVM for most of my testing. [09:12] SpamapS: I run a few basic installs on real HW just to check. [09:14] ttx: I've no real hardware to test on really.. just my laptop :-P [09:15] I moved everything else into the cloud ;) [09:16] Though I have been thinking about putting in a server of some kind to host my local mirror of the archives [09:18] ttx: so 20100601.1 ? [09:19] SpamapS: whatever the ISo testing site says. Looks like it is 20100601.1 for ISOs, yes [09:19] ymmv for UEC images and ec2 amis [09:36] ttx: ok, see you after some sleep. :) === _chris__ is now known as chris-- === chris-- is now known as chris- [10:20] anyone in here know mysql server? [11:06] a little off-topic, but still https://secure.avaaz.org/en/gaza_flotilla_7/?rc=fb [11:06] tyler_d8: I know it quite well [11:07] ever migrated a db in 5.1 --- ubuntu [11:08] what are obvious things to check if I'm getting a 503 on a basic apache2 setup on 10.04 ? [11:08] localhost works fine but for some reason request to the actual host name are getting a 503 [11:10] actually scratch that it's working with the host name - it just won't work from request that aren't on the host box [11:11] tyler_d8: usually it just works - export the db from the old version with mysqldump [11:11] no no, just the physical location on the same disk [11:11] ie. modify the instance to have the data stored on a raid lets say [11:50] is there a list of people providing ubuntu vps services around the world? [11:51] eeeeerm ... seems there may be a kernel memory leak in lucid [11:56] http://pastebin.com/XK8aDVkc [11:59] it seems there are 16 processes running, each using 2,5GB or so, I have 64GB memory and 16 cores in this baby. According to ps, it's not using a lot of memory, but it's swapping like crazy [11:59] RoyK: uptime? [11:59] incorrect: only google AFAIK [12:00] Daviey: 18 days [12:01] Daviey, i found one that the xen guys do, [12:03] RoyK: can you pastebin the headers of top? [12:04] http://pastebin.com/msXYduBX [12:06] RoyK: erm.. "top"? [12:06] oh [12:06] top [12:06] :) [12:07] ie, http://pastebin.daviey.com/wytX/ [12:09] http://pastebin.com/cwQYihHe [12:11] RoyK: to me, it looks like Flexpart is leaking [12:12] RoyK: Are you able to test Flexpart against valgrind? [12:15] Daviey: should work, but then, why on earth doesn't ps report this? also, these are jobs that are started and run for, say, 2-3 days. then new jobs take over. new processes, just scripted [12:15] I love my personal server now that it runs Ubuntu :D [12:15] I was running it on FreeBSD but it was starting to get kinda hard-to-manage that way [12:16] (keeping ports updated and all that) [12:16] how do I thank Canonical for their great work? [12:18] RoyK: you could be right, but it would be interesting to see the results of valgrind to rule it out. [12:19] RoyK: a resource leak you are experiencing in quite a short time, i would have expected others to have reported that by now [12:19] Not saying that ISN'T the case, but be good to rule out things first. === twister23_ is now known as twister23 [12:27] Daviey: this is old fortran code, so I don't know - might be worth a try [13:07] hi guys! [13:07] anybody wiling to help me a bit with a weird behaviour of SAMBA? [13:07] !ask [13:07] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [13:13] Guys, I have the following SMB settings - http://pastebin.com/pfahh509 . When a user from creativeshop group creates a directory and uses Windows OS, it correctly creates a directory with 0770 privileges, but a user under MacOSX creates a dricetory with 751 and I really dont know why [13:13] any idea what is wrong ? [13:20] how to count the number of files in a dir ? [13:20] ls /path/to/dir | wc -l [13:21] Pici, and recursively? [13:22] find ./ | grep wc -l [13:23] Pici, thanks! [13:23] sure! [13:23] kirkland/Daviey: could you cover http://iso.qa.ubuntu.com/qatracker/result/4188/494 in Alpha1 ISO testing ? [13:24] ttx: may I ask you another question regarding UEC ? [13:24] spineau: sure [13:25] ttx: When you've just downloaded/registered an image [13:25] ttx: Does every node cache it or only the one that will be running the instance ? [13:26] spineau: the node caches it the first time it starts it [13:26] ttx: If there are extra nodes on the CC, they are not caching for future needs ? [13:27] morning [13:27] spineau: I don't think so. [13:27] ttx: ok, thanks [13:31] smoser: ping [13:34] ttx: I could indeed! [13:35] Daviey: cool ! You can cover the UEC images test with the same environment: [13:35] http://iso.qa.ubuntu.com/qatracker/test/4216 [13:35] http://iso.qa.ubuntu.com/qatracker/test/4217 [13:35] Daviey: ask me if you have any question, the docs should be straightforward though [13:35] hmm.. ok [13:35] * ttx updates them to use the lucid image by default [13:35] ttx: When is the deadline? [13:36] Daviey: end of today ? [13:36] ttx: It's going to be tight! But think it can be done. [13:36] Jeniczek|vPraci: You're setting a mask and you're not forcing the mode, so whatever the mac user put in his profile, will fit your mask. [13:40] I'm setting up a Samba server agains 2008r2 DC's, but the realm= in smb.conf seems a bit unclear to me. Is it the domain that goes here, or is it the DC itself, if it's the DC itself, can I list more than one DC? [13:40] alvin hmm [13:41] zul: for alpha1 testing, could you cover one of the EC2 regions ? Not sure smoser will show up. [13:41] sure [13:41] zul: that would give us confidence, or time for a respin if needed [13:41] Jeniczek|vPraci: You can change that on the macs. (in Globalpreferences.plist or something. I don't use samba for macintosh. Resource forks are still a cause for trouble) [13:41] alvin so I have to use ,,force create mode (S) ,, and force directory mode (S) ? [13:41] ttx: gotcha [13:42] ttx, i'm here. [13:42] Jeniczek|vPraci: Yes, I'd experiment with that. It's easier than changing all those macs [13:42] alvin the SAMBA runs on ubuntu server, not on MacOSX [13:42] and i've run the tests last night. [13:42] smoser: ha [13:42] alvin okey than [13:42] i just need to open a couple bugs and record results [13:42] thanks alot [13:42] smoser: care to start your magic EC2 automated tests ? [13:42] zul: then you can pick random ISO tests [13:42] Jeniczek|vPraci: I know, but the macs will try to put their resource forks on your server. How that happens is dependant of the version of OS X [13:43] * ttx is on random ISO tests as well [13:43] zul: mark them started when you start them [13:43] ttx:yep yep [13:43] ttx, i already did on 20100601 [13:43] was re-spin required ? [13:43] smoser: no [13:44] smoser: the tests don't show up as completed (or started), so I figured you didn't run them :P [13:44] right. i just kicked off the test before i went to bed [13:44] i have to open some bugs and record results. [13:44] alvin thank you very much, am gonna do it === bogeyd6 is now known as Guest47942 [13:45] smoser: I had ara promote the UEC image to the tracker, btw [13:45] have you sniffed it ? [13:45] smoser: no. [13:45] :) [13:45] i can sniff quickly here on lucid UEC [13:45] smoser: that would be great [13:46] I didn't plan on reinstalling UEC today, was trying to push Daviey/kirkland into covering that part [13:46] Jeniczek|vPraci: good luck. (btw, people told me netatalk has come a long way. Apple will still work better with afp than samba.) [13:47] alvin well for me it sounds, that I have to force both create modes ( dir and file) to 770 on all shares that MacOS users have access to... [13:47] ttx: I really wanted to get this merge polished today, and the rig isn't yet setup.. That shouldn't take tooo long, but need to dig out some cables. [13:48] Daviey: maybe kirkland can cover the alpha1 UEC testing (at least on amd64) === Guest47942 is now known as bogeyd6- === bogeyd6- is now known as bogeyd6 [13:50] ttx: I think i can cover it, but depends on any hiccups getting the rig deployed. I did want to borrow kirkland to review a diff.. it's quite large @ 648364 lines, and a second pair of eyes would be handy. [13:52] Jeniczek|vPraci: I'm using force create mode = 0660 and force directory mode 0770, but I don't let apple users use samba. (using a commercial solution for that). Wen using those masks, I set the umask to 002 on the Linux clients (NFS) in /etc/profile. It results in directories drwxrwxr-x and files -rw-rw-r--. [13:53] alvin why you dont let use apple users samba? [13:53] zul: please concentrate on i386, I've trouble downloading the i386.iso [13:54] ttx: ack [13:54] Jeniczek|vPraci: That'd make a mess of the shares. That's not too bad, but their applications would have trouble finding out what filetypes are on the share. [13:54] Jeniczek|vPraci: They could use extensions for that too, but well... [13:55] hmm interestin [13:55] Jeniczek|vPraci: Our mac users are used to OS9. They don't use extensions and you can't even begin to imagine what symbols they try to put in filenames [13:55] the force create modes are for specific share or as global option? [13:56] i have just se tit for every share and the newly created dir is again with bad mask [13:56] Oh, you can choose that. I use them globally [13:56] maybe samba needs to be restarted [13:56] or refreshed. [13:56] how to do that? [13:56] I dont want to break current open files [13:56] Oh. On Linux it's $ sudo service samba reload [13:56] reload, ok [13:56] that won't break open files [13:57] Reloading /etc/samba/smb.conf smbd only [ OK ] [13:57] smbd should be enough, right? [13:57] yes. nmbd is only for netbios [13:58] Here: http://en.wikipedia.org/wiki/Resource_fork There's a paragraph about SMB shares in that article [13:59] ok, lets try it again [14:00] but different versions of OS X will save their resource forks in different ways and you can get compatibility problems [14:08] alvin hmm, the MacOSX still ignores the force mask [14:08] I must have somethin wrong then [14:08] It sill creates just 751DIrs [14:10] What is the umask you'd like to see as a result? 770 dirs? [14:10] smoser: have your thought of creating the rootfs for the maverick ec2 images to be ext4? [14:10] alvin yeah, for most of the dirs, 770 [14:10] zul, thats the plan, yeah. [14:11] move to ext4 so i can move to btrfs next : [14:11] smoser: k [14:11] heh [14:12] ttx, images are booting, but i can't get to them. [14:12] (uec) [14:14] no ssh into them ? No network ? [14:14] never mind [14:14] user error [14:14] i hate --key [14:16] lol [14:17] ttx, when the user uses the system correctly it seems that the instances boot and are reachable via ssh [14:18] smoser: good news, mark the tests ok while it works [14:19] alvin any luck? [14:20] ttx, its valid to call it passed having tested on lucid host ? [14:20] Jeniczek|vPraci: I denied access to samba from the macs, so I can't test there, but I tried this from Windows. I want directories as 2775 and that works. Maybe try create mask = 02770 and directory mask = 02770 [14:20] smoser: it's actually how it should be tested. [14:21] smoser: moving guests on stable host, moving host with stable guest [14:21] though when everything works the shortcut is tempting. [14:21] Jeniczek|vPraci: You can also try 'force directory mode' [14:21] yeah, thats what I was trying [14:22] The samba howto is pretty complete about that. (difference between mask/mode etc) [14:22] <\sh> ttx, regarding the ipv6 problem with authbind, wouldn't it be a good idea to add ipv6 support to authbind? [14:22] \sh: that sounds like the right solution. [14:23] \sh: haven't looked into what that would mean, though [14:23] * ttx watches his upload rate go from "almost unbearable" to "completely useless" [14:24] Jeniczek|vPraci: Aha, look here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=532856 I don't know if it applies. The version of samba will be more recent in Ubuntu. [14:24] Debian bug 532856 in samba "umask settings overridden by Mac OS X 10.5 (Leopard) clients" [Normal,Open] [14:24] can't even send an email with attachments wth [14:26] <\sh> ttx, when I find the time, I'm trying to google some examples of how to implement such logic for ipv6...anyhow I need to find a way to test it though, and this will be a complicated task for me ;) [14:27] ttx: sure [14:27] Daviey: morning! [14:27] kirkland: see with Daviey to not duplicate effort -- also I don't think we need to tset UEC/i386 or the alternate topologies for this milestone [14:28] ttx: great [14:28] ttx: i'm supposed to review Daviey's merge of eucalyptus-devel today [14:28] ttx: i think it would be best if he installed a Maverick UEC for ISO testing [14:28] ttx: and then we upgrade that to his merge ;-) [14:28] ttx: two birds [14:29] kirkland: I like that [14:29] hmm, alvin thats pretty crazy [14:29] Jeniczek|vPraci: I concur [14:29] am tryin to set unix extensions = no if it make any effect [14:30] it seems that it helped to those guys [14:33] Jeniczek|vPraci: Yes, I'm reading that here. Now I finally know what those unix extensions do. I'll leave them in here. [14:33] Jeniczek|vPraci: If you have the time, I'd try netatalk. [14:34] I have never heard about netatalk [14:36] I use a commercial version of it. Keep help those Apple machines in check. Apple has always used appletalk and netatalk will bring that to your linux. [14:37] (In my world, NFS is for Unix/Linux, SMB for Windows and AFP for Mac.) [14:38] hmm [14:38] ill have a look [14:38] lets try those extensions to be off [14:38] I dont have other Linux clients or server in this LAN, so turning them off should do no harm [14:40] If I understand it correctly, it shouldn't harm you anyway. only the umasks, set on Linux/Unix will not be set on the shares. [14:46] how can Iinstall perl ? [14:46] oru_work: Perl should be installed by default. [14:46] Pici, ok [14:47] hm [14:47] alvin its funny, now it creates 775s ;) [14:48] Jeniczek|vPraci: If that's not what you want, look in man afpd. There's an option there to set umask [14:57] Hi everyone. Was looking for some advice with a stange issue I am having with an Ubuntu 10.04LTS server I set up last week. cron.daily scripts do not seem to be being run. Also /var/log/messages is completely empty. [15:01] Daviey: ping me when you're around [15:01] kirkland: PING [15:02] Daviey: heellllloooooooo dave [15:02] kirkland: Hellloooooooooo ... Dustin [15:03] Daviey: alrighty, so today, we need to do 2 things ... a) run the alpha1 uec iso test, and b) review your eucalyptus-devel merge, right? [15:04] kirkland: correct-a-mundo. [15:04] kirkland: The diff is huge, so a second pair of eyes would be awesome [15:05] hand to re-introduce the makefile diff, as it was failing to compile [15:05] Daviey: ah, hmm, we'll need to look at that one closer then [15:05] *had to reintroduce [15:09] Daviey: let me reboot, and then let's get a schedule together for those two items, to make sure we sync up on them today [15:09] kirkland: Sounds good. [15:09] * hallyn waits for wiki access... [15:10] i'm reviewing general todos, pls lemme know if i'm supposed to be doing anything else in particular [15:11] hallyn: welcome [15:11] hey hallyn [15:11] thx :) [15:11] hey [15:28] hi guys [15:28] I'm configuring kvm and I'm getting some troubles. [15:29] I created a VM and I've configured bridge network interface [15:30] although I can ping the host from a VM and ping a VM from the host, I can't ping the internet. [15:30] do I have to configure tun/tap interface to get it working? [15:31] what could be missing? [15:31] any ideias? [15:32] Hello. I have configured my Ubuntu 10.04 with LDAP. At logon, appear a list of users who have previously logged. How can I remove these "cached" users? I only want to appear the box to write the username and password. [15:33] LinuxAdmin: did you start from cmdline or from virt-manager? [15:34] if cmdline, what cmd? [15:38] sudo mysql -v [15:38] ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: NO) [15:38] why would it do that [15:38] i'm running ubuntu 10.04 [15:39] it doesn't prompt me for the password :( [15:39] oru_work: you're not telling it to prompt [15:39] how would I do that ? [15:39] oru_work: mysql -p [15:40] sudo mysql -p [15:40] Enter password: [15:40] ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: YES) [15:40] I have to be root right ? [15:40] I don't think i have root enabled [15:41] You should enable the password first :P [15:43] can I install bugzilla from one of the packages or should I do it manually ? [15:46] you dont have to be root [15:46] on linux [15:46] if thats what you meant oru [15:56] oru_work: mysql -u root is the command you need, if you do not set up the password [15:57] hallyn: I started from the command line [15:58] but I've also tryed from virt-manager [15:58] it's the same behavior [15:58] what cmd? [15:58] with virsh [16:00] what I'm burning my head is why can I ping the host from the outside world (it means that bridge is well configured) [16:00] why can't the VM ping the internet [16:00] ? [16:01] these are the last few lines of installation of /usr/bin/perl install-module.pl DateTime http://pastebin.linode.com/3895 [16:02] can someone help me with this issue? [16:02] LinuxAdmin: when you started from the command line, what command did you use [16:04] sudo apt-get build-essential [16:05] is this misspelled ? [16:05] LinuxAdmin: i generally start vms with something like [16:05] kvm -drive file=server1.img,if=virtio,index=0,boot=on -m 1000M -smp 2 -redir tcp:2222::22 [16:15] I defined auto start [16:15] when I started kvm the VMs start automatically [16:15] sorry hte delay [16:17] I used "virsh -c qemu:///system start vm1" [16:24] i got this error after apt-get isntall yaml-mode [16:24] hmmm http://pastebin.linode.com/3896 [16:24] any ideas ? [16:27] hi Deji [16:28] hgghd : ping [16:28] brianchidester : pong [16:28] hggdh: ping [16:28] LinuxAdmin: can you pastebin the result of 'virsh dumpxml vm1' ? [16:28] brianchidester : thanks for the help [16:28] Deji: yarp [16:33] Deji: pong [16:36] can someone help me with this error http://pastebin.linode.com/3898 [16:47] oru_work: There's not much context there. What printed this error? [16:48] SpamapS, i'm already passed that , thanks though. It was after I ran the installer for one of the perl modules [16:48] SpamapS, was missing one dev library [16:55] sorry again hallyn, It's been a busy day. here is pastebin link: http://paste.ubuntu.com/443428/ [16:56] hallyn: do you see anything strange with the config file? [16:57] br1 it's my bridge interface [16:57] I can ping it from the outside world [16:58] so, I assume is ok with the bridge configuration on the host [16:58] what do you say? [16:59] LinuxAdmin: really, you can ping it from the outside world but can't get to the outside world from it? [16:59] I can get to the outside world from the host [17:00] what I can't is ping the outside world from any VM [17:00] and you mean you can ping the host, but not vm1, from the outside word? [17:00] yes [17:01] what exactly do you get if you do wget google.com from vm1? [17:01] (i'm wondering whether your /etc/resolv.conf is just not right) [17:01] can you pastebin the output of 'brctl show'? [17:02] as well as 'ifconfig -a', 'netstat -nr' from the host, and 'netstat -nr' from vm1 [17:02] (at your convenience :) [17:03] ok [17:03] just a minute [17:11] hallyn: the link is http://paste.ubuntu.com/443436/ [17:13] Does anyone have experience having nagiosgrapher graph things from statd? Specificlly, I'm trying to get a graph of the data reported from check_load_statd. [17:13] hallyn: do you see anything wrong? [17:17] LinuxAdmin: well you have br1 containting only eth1, which appears to be unconfigured [17:18] where is eth1 supposed to lead? [17:18] i suspect if you hook vm1 up to virbr0 instead of br1, it'll "just work" [17:19] (and i suspect you want to just get rid of br1 altogether) [17:19] (leaving it there won't cause problems, just be confusing) [17:19] but I want to use bridge, and I should point bridge interface to physical interface br1 --> eth1 [17:20] * ccheney wondering where his ac repairman is at [17:20] but eth0 is the configured netif [17:20] or not? [17:20] oh, i see. [17:20] eth0 is an interface used to a private lan, just to synchronize share storage [17:20] so is 194.79.75.210 your real external ip? [17:20] yes [17:21] if I want outside world connect VMs without nat, I mean directly, I should configure bridge [17:21] yes [17:22] I installed a debian guest and did not use the --accelerate falg, can I enable it some where in the xml file? [17:22] otherwise I have to make nat roules [17:22] rules, I mean [17:22] and every manuals tells me to make bridge this way [17:23] LinuxAdmin: yes you need a bridge, what ubuntu release are u using [17:23] and it seams that is working because I can communicate with bridge interface ip from the outside world [17:23] 9.10 [17:24] (sorry, having troubles of my own, metacity hanging) [17:24] I've read a lot on the internet, and I saw an article that says ubuntu 9.10 is getting problems with this [17:24] LinuxAdmin: you can't ping outside world by ip address either? [17:24] no [17:25] it's not a dns issue [17:25] I can't even ping an ip of the same network [17:25] I mean, in the same switch [17:25] LinuxAdmin: i did use 9.10 with virtio over bridge, but i had to forward bridge to wireless card so wasn't exactly the same setup [17:26] LinuxAdmin: can you pastie your /etc/network/interfaces [17:26] yeh [17:26] ok [17:26] I use 9.10 host with bridges guests all the time [17:26] just a minute [17:27] do you want /etc/network/interfaces from the host or from the VM [17:27] ? [17:27] I'll post both [17:27] just wait [17:27] LinuxAdmin: the last thing looking suspicious to me is that your gw on vm is .209, not .210 [17:27] might be right, but i would have thought the br1 ipaddr woudl be the right value [17:28] hallyn: btw, welcome! :) [17:29] SpamapS: thx [17:33] here is the link: http://paste.ubuntu.com/443449/ [17:34] sorry for delay but I can't copy/past from VM [17:34] I'm using virt-manager and mouse does not work there :) [17:35] any ideia? [17:36] diago: do you use karmic guests? [17:39] diago: do I have to configure any firewall rule to forward traffic between interfaces? [17:39] I haven't try that yet [17:43] hey y'all! i want to change the path a symbolic link points to WITHOUT deleting it. i tried the following and somehow it failed http://paste.ubuntuusers.de/398435/ <- do you folks have an idea? [17:45] diago: are you still there? [17:47] could someone have a look at this http://pastebin.linode.com/3902 [17:47] i do have apache2 installed [17:48] oru_work: What are you trying to do? [17:48] Is Munin relatively light weight? [17:49] Pici, sudo /usr/bin/perl install-module.pl mod_perl2 [17:50] oru_work: mod_perl2 is in the repos, why are you trying to install it from scratch? [17:50] Pici, i'm following the isntallation guide [17:50] oru_work: What guide? [17:50] Pici, bugzilla [17:51] Pici, i tried apt-get install mod_perl2 , can't find package :) [17:51] oru_work: its libapache2-mod-perl2 [17:52] oru_work: Installing these packages from source means that you will not get updates on them when they become available and will render your install somewhat unsupportable. [17:53] Pici, i didn't know that [17:54] oru_work: If you really feel that you need to install bugzilla from source, I highly suggest that you install its dependencies from the packages available in the repositories. [17:54] Pici, well right now here is what i see, thanks to you i got the last bit piece of it and ready to go on with my isntallation http://pastebin.linode.com/3903 [17:54] Pici, well they are already installed, what do you suggest at this point ? [17:55] is it just me or do none of the "Store" images work on UEC 10.04? [17:56] Pici, oru_work: even better, install the bugzilla3 package :) [17:56] pmatulis, :) [17:57] installing [17:57] 80 mb of space :) [18:04] hello I have problem with ssh I get this error PEM_read_PrivateKey failed and I am sure private key is at right place on server. [18:04] I am trying to connect to Ubuntu 8.04 from my Ubuntu 10.04 and I did everything online manual said [18:05] Does anyone have experience having nagiosgrapher graph things from statd? Specifically, I'm trying to get a graph of the data reported by check_load_statd. [18:11] for pdfedit sudo apt-get install pdfedit [18:14] zul: wondering why bug 564920 was declined for lucid [18:14] Launchpad bug 564920 in php5 "PHP5 under Apache2 on 64 bit system is not completely 64 bit " [Low,New] https://launchpad.net/bugs/564920 [18:15] pmatulis: because its not fixed in maverick [18:16] zul: can you elaborate? no SRU possible? [18:17] Pici, is request tracker in ubuntu repos ? [18:17] pmatulis: because i havent had a chance to look at it properly. usually once its fixed in the development release i usually open an SRU [18:17] zul: gotcha [18:18] hrmm [18:19] where can i find info on how to setup heartbeat resources [18:19] like for postgresql [18:19] pmatulis: if you want to try to fix it...be my guest :) [18:19] oru_work: Yes: the package name in lucid is request-tracker3.8 [18:19] i'm looking around can't find what i need [18:20] oh wait nm got it [18:20] Path PostgreSQL data directory. (optional, string, default /var/lib/pgsql/data) << that's annoying [18:20] why wouldn't it default to PG_DATA [18:20] global var [18:21] zul: will do [18:22] zul: ;) [18:26] Pici, just finished installation now sure whats next, what would be he url in the browser ? [18:30] oru_work: I'm not sure, I've never installed bugzilla myself. [18:33] i have a probelm with ubuntu enterprise server, im trying to set it up with Eucalyptus but my Node does not see the Cluster. Ive checked to see if they are on the same subnet and they have there own class B ip [18:37] hggdh : ping [18:37] deji I am here [18:37] ac repairman here now, and working on the unit :-) [18:37] Pici, i'm really having trouble locating it :) [18:38] ccheney: COOL! Er, not yet, I mean great! [18:38] hggdh, yep soon :) [18:38] * look sobs and runs for a corner [18:38] hggdh : hey, i talked to kirkland and he said i have trouble with cc and the logs were not registering [18:38] hggdh : were u able to access the systems? [18:39] oru_work: I'd start by looking in /etc/apache2/conf.d/ [18:39] 28C and rising in the house, so it will be good to have this done before the hot part of the day [18:39] Deji: yes, I was. Also, it seems to be working now [18:39] hggdh: i'm on the phone at the moment, and hoping you might be able to help deji with his test rig [18:39] hggdh: this was a package-based uec install [18:39] hggdh: the components are registered, but the nodes are not [18:39] kirkland: they are now [18:39] hggdh: the problem, as i see it, is that the CC is not quite running all the way [18:40] hggdh: hmm... [18:40] Pici, nope nothing there [18:40] AVAILABILITYZONE |- m1.small 0000 / 0000 1 192 2 [18:40] kirkland: that was my feeling also. [18:40] ohhh [18:40] bad [18:41] hggdh: can you help deji debug this? [18:41] Pici, i have /usr/share/request-tracker3.8 [18:42] kirkland: will do. Deji see also https://pastebin.canonical.com/32887/ [18:43] oru_work: Oh, I thought you were talking about bugzilla.. one moment. [18:44] Pici, no RT3.8 now :) [18:44] Pici, bugzilla i found easy it was mydomain.com/bugzilla3/ [18:44] Pici, now i'm having trouble locating login url for request-tracker8 [18:44] hggdh : its working now, thanks a lot. Its registered [18:45] oru_work: domain.com/rt perhaps? [18:45] Pici, tried it nope [18:45] Deji: but still euca-describe-availability-zones verbose lists 0 [18:46] Deji: of course, the NCs are not registered yet [18:47] oru_work: before I look further, can you check /rt3 ? [18:47] Deji: ceylon states a system restart is required. Were package updates installed there? [18:48] hggdh : i think so would do a reboot on that system now [18:48] Deji: OK. Are you driving the reboot, or you want me to do it? [18:48] hggdh : i'll do it, thanks [18:49] Deji: roj [18:49] Pici, i checked rt3 , not found [18:50] * ccheney will be in his attic for a bit with the repairman, will check for messages when i get back [18:51] oru_work: Okay, looks like you may need to symlink one of the conf files from /etc/request-tracker3.8/ into /etc/apache2/conf.d/ then [18:51] And then it will be located at /rt [18:52] Pici, symlink? i don't understand [18:53] oru_work: like: sudo ln -s /etc/request-tracker3.8/apache2-modperl2.conf /etc/apache2/conf.d/ [18:54] Pici, :) [18:54] Pici, should i run this command ? [18:54] oru_work: yes. [18:55] Pici, okay I did [18:55] Pici, i tried url.com/rt still not found [18:55] oru_work: Okay, restart apache first. [18:55] oru_work: sudo service apache2 restart [18:56] * Restarting web server apache2 [18:56] apache2: Syntax error on line 233 of /etc/apache2/apache2.conf: Could not open configuration file /etc/apache2/conf.d/bugzilla3.conf: No such file or directory [18:56] ...fail! [18:56] errr [18:56] bugzilla3 [18:56] ... [18:56] I uninstalled it [18:57] with eucalyptus it says when i enter the "euca-describe-availability-zones verbose" it gives me this error 'EC2_ACCESS_KEY veriable must be set' [18:58] could i get some help with that? [18:58] New bug: #588605 in groovy (main) "Groovy package in Maverick should be 1.7.2" [Undecided,New] https://launchpad.net/bugs/588605 [19:01] how would i specify line number in nano ? [19:03] i give up no one wants to help [19:05] sommer: the server guide does not mention nfsv4, any reason why? [19:07] pmatulis: no specific reason... the nfs section could be updated [19:07] pmatulis: contributions welcome :) [19:07] sommer: alright. i thought there might be another reason [19:11] something seems to be rather bad in 10.04 - just had to reboot this server after the _kernel_ ate half its memory. I stopped all memory-hungry processes and ps axfv showed very low usage. Still, there was ~32GB used [19:12] and, no, this was not buffers/cache [19:16] RoyK: what is the server doing? [19:18] hello has anyone here used a sunfire v100 [19:18] hggdh : ping [19:19] Deji: go ahead [19:20] hggdh : hey please am trying to get the credentials for the clc using the steps on https://help.ubuntu.com/community/UEC/PackageInstall [19:21] hggdh : i can't get on tamarillo using the web [19:21] pmatulis: scientific models [19:21] Is there a good writeup somewhere that describes advantages of ubuntu server over debian? [19:21] pmatulis: but it shouldn't really matter, it's something in the kernel, not userspace [19:21] hggdh : since it is protected by a proxy batuan.canonical.com [19:21] jimcooncat: it's mostly the same [19:22] over 45C+ in the attic with the plenum venting into it, would be much worse if it was closed up i'm sure [19:22] Deji: you either have to use FoxyProxy, or set up a tunnel under ~/.ssh/config [19:22] RoyK: I gathered that, I'm installing some new ones and want to figure out whether to use the new LTS or lenny with selected backports [19:23] how would i specify line number in nano ? [19:23] jimcooncat: we're mostly using 8.04 so far [19:23] has anyone here increased the number of available file descriptors before? [19:23] Deji: then you can access the UEC web admin via https://localhost:8443 [19:24] jimcooncat: 10.04 is still fresh and fresh distros usually contain fresh bugs as well [19:24] RoyK: I get that -- I usually wait a couple months after release [19:24] Deji: for example on ~/.ssh/config, see https://pastebin.canonical.com/32874/ [19:24] Deji: you will have to adjust as needed, of course [19:26] RoyK: I guess I'll have to look at the individual apps -- there were a few things on squeeze I would love to have, but I haven't the patience to play with anything non-stable [19:29] jimcooncat: 10.04 is stable, though I've had a terribly strange problem with this 16 core box. suddenly the kernel was eating 32 gigs of RAM and the box was swapping wildly [19:29] jimcooncat: I think it might be related to compcache, but don't know yet [19:29] hggdh : so when i create a ~/.ssh/config file, what next? [19:30] Pici, do you mind having a look at this http://pastebin.org/301485 [19:30] Deji: man ssh_config? [19:30] RoyK: yikes! I've got three 4GiB boxes to set up in a very small office [19:31] oru_work: try enabling mod_rewrite: sudo a2enmod rewrite [19:31] jimcooncat: this one has 64 gigs of ram [19:32] and since compcache was set to 50%, I guess that's the bad one [19:32] the models are quite memory-hungry as well, but no more than about 3GB per job, so it should run well, and it did for 16 days or so, until a couple of days ago [19:33] RoyK: htop is my friend :-) [19:33] what does htop do better than others? [19:33] it's top with training wheels [19:34] heh - I've been using linux since 1994 or so, I don't think I need training wheels :) [19:35] I'm a bookkeeper and only admin on the side -- so sometimes I forget stuff I use only once in a while [19:36] I'm a unix sysadmin :) [19:44] and drbd experts around [19:47] RoyK: It looks like ubuntu server is making itself a player in "cloud" computing, which I don't (think) I have need for. So I guess I'll go with lenny, unless there's someone else that can help point out an advantage [19:48] jimcooncat: it doesn't do that unless you install it as a cloud node [19:49] you get that choice in the initial install menu (grub)¨ [19:49] RoyK: I haven't a clue what it's usefulness is. I use a linode for stuff I need on the net. [19:50] it doesn't use any cloud stuff unless you specifically ask it to [19:50] calm down :) [19:51] jimcooncat: If you look at the security hardening specs for Ubuntu Lucid and how much of that is implemented in Debian already, I think it shows a clear win for Ubuntu. [19:51] hggdh : ping [19:52] No, it's just that I've left a lot of comments on blogs, and forums asking what the cloud thing is all about, and I've not had much practical feedback. I'd see it's usefulness if it did async drbd or some other automatic failover thing [19:52] jimcooncat: It's completely possible to entirely ignore the cloud stuff if you don't care. It's what I do. [19:53] how can i tell with drbd which cluster is set to primary currently [19:54] ScottK, jimcooncat: for details on the security features, see https://wiki.ubuntu.com/Security/Features [19:54] Thanks. [19:55] jimcooncat: Compare that with what is in Lenny. [19:55] the big things that Debian is missing is NX emulation, stack protector, PIE, fortify source, relro, bind now, and (i think) CONFIG_CC_STACKPROTECTOR [19:55] with some individual packages build with PIE, stack protector, fortify, relro, and bind now. [19:56] (under 20, last I checked) [19:57] thanks much, kees, ScottK. NX emulation != freenx stuff, correct? [19:58] It's explained on the wiki page. [19:58] oh, found it. [19:58] jimcooncat: that acronym is heavily overloaded, unfortunately. I mean "non-executable memory", which goes by both XD and NX. [19:59] Pici, e2enmod rewrite helped, apache now restarts without error [19:59] Pici, going back to the URL for rt3.8, not sure what it is i tried mysite.com/rt and its not it [20:08] Hey guys [20:09] i think the ac repairman just dumped 15lb of freon into our house [20:09] ccheney: dude i lived in an apartment and they would just go around refilling them instead of fixing it [20:10] ccheney: nice for the indoor climate [20:10] there is so much in the house its cloudy [20:10] does this look right [20:10] http://pastie.textmate.org/private/7edo79qjnce191w8wsnea [20:10] * ccheney wonders if it is safe for me to be in here, i sent my wife out immediately [20:11] i'm having problems with getting the secondary going it seems [20:11] when i start the secondary and it is selected to be the current dc [20:11] i get this [20:11] http://pastie.textmate.org/private/f1nhragk7mch2smgnn18lq [20:12] this kinda sucks [20:12] i don't know what to do [20:13] ccheney: it's probably not CFC unless the fridge or whatever it is, is really old [20:14] webPragmatist: your pastebin isn't working for me [20:14] jimcooncat: hrm [20:14] one sec [20:14] let me try something and i'll repaste after that [20:14] royk: r22 whatever that is [20:14] royk: the ac unit is about 6 years old, before the changeover to 401 [20:15] jimcooncat: heartbeat is supposed to mount it for you using Filesystem right? [20:15] http://en.wikipedia.org/wiki/Freon#Phase_out_of_CFCs [20:15] webPragmatist: I'm just lurking, I've only researched drbd, not implemented it yet [20:15] hmm i think i meant 410 is used now apparently :) [20:16] CFCs are BAD [20:16] good thing i sent my wife out then, heh [20:16] hrm [20:17] drbd is working find it's just Heartbeat that is hosed [20:17] ccheney: lol - they're only bad in the stratosphere... [20:17] "Overexposure may cause dizziness, loss of concentration, Central Nervous System depression and/or cardiac arrhythmia." not sure if 15lb in a house is enough for that though [20:17] or just if you try huffing the stuff [20:19] perhaps airing out would be a good idea.... [20:20] yea i think i will once he is done, my wife is 9mo pregnant so i didn't want to risk any trouble for her [20:21] always good. Keep in mind that high concentrations will replace oxygen, so (apart from all else) oxygen starvation is a risk [20:21] lol this thing keeps trying to promote and demote the drbd [20:23] webPragmatist: using heartbeat/pacemaker? [20:23] yes [20:23] it's like neither are Master [20:23] so it's just like wtf promote promote promote [20:24] it will say master for like a second but i think the promote fail [20:24] fails* [20:25] RoyK: have you done htis? [20:25] no, but I'm working on a test setup [20:25] probably be finished in a few days if I get the time [20:25] http://pastie.textmate.org/private/ynmooiesn2noryowbzl7ow [20:25] webPragmatist: what do you intend to use this stuff for actually? [20:26] an ecommerce site [20:26] webPragmatist: is it something required now? i'm asking b/c there is a new HA stack in the works [20:26] are we talking days? [20:27] i mean heartbeat has been around forever…. [20:27] I somehow doubt we're talking days [20:27] yea [20:27] more like 10.10 [20:27] webPragmatist: yeah, and as you've seen. it doesn't work very well [20:27] 2012 [20:27] https://blueprints.launchpad.net/ubuntu/+spec/server-lucid-cluster-stack [20:27] i'm just not doing something right [20:28] but yea we have already purchased the server [20:28] so i need to make it do stuff [20:29] heh, is that directly from your boss' mouth? [20:30] no i am the boss [20:30] my boss wouldn't understand [20:30] ok [20:31] but i don't care to get involved in a server crashing… [20:31] because it's my but [20:31] butt* [20:31] okay question [20:31] http://pastie.textmate.org/private/rjbtogxzgqbjtdvwc2bq [20:31] i remember trying to do the heartbeat/drdb thing. stuff not starting, unpredicatble behaviour, etc [20:32] Not an ubuntu-server question as much as an ubuntu-client question, but I figured you folks may have experience with it. I'm trying to get ubuntu-clients to connect to an ldap server, and they dont' seem to be reading their /etc/ldap.conf [20:32] location loc-1 fs_drbd -inf: cluster2 [20:32] location loc-2 drbd_disk -inf: cluster2 [20:32] which host is supposed to be that? [20:32] the "opposite" that runs corosync? [20:33] If I run ldapsearch from one of the clients, I need to specify the server and base that are arleady specified in /etc/ldap.conf. I'm guessing if I fix that, the rest of this should fall into place [20:33] hrm [20:33] i think all those meta tags shold be 1 [20:35] Anyone here with ubuntu+ldap experience who could lend a hand? [20:37] grrr, should postfix really be a dependency of mdadm? [20:38] pmatulis: Probably not, but that's package based distros for you :) [20:38] pmatulis: if you already have sendmail or exim will it use those instead? exim is tiny. [20:38] pmatulis: probably "mta" is a dependency [20:38] pmatulis: and postfix the first choice [20:40] http://www.drbd.org/users-guide/s-pacemaker-floating-peers.html [20:40] pmatulis: ^^ [20:40] hrm [20:40] i'm trying to do that but the HA testing doesn't do that [20:41] hi all [20:42] noob queston, how do I remotely log into ubuntu from cold start? [20:43] Freud_: I'm not sure that's possible, due to nothing listening by default. You'd have to set up something like openssh-server. [20:46] Does anyone here know how to get ldap-users to appear in the login screen on 10.4? [20:47] pmatulis: it seems they get out of sync or something [20:50] damn spoke too soon or something [20:57] what is the difference between using kvm and eucalyptus? [21:00] one is an herbal remedy [21:00] anyone atleast just use drbd? [21:01] yep [21:01] i use it all the time [21:02] webPragmatist: give me a few days, and I'll know [21:02] yipee, ac is now working, and starting to cool the place down a bit [21:03] feels extra cool to me as i was up in a 50C attic for part of the time [21:04] incorrect: what if my nodes are stuf on WFConnection? [21:04] stuck* [21:04] might have even been higher than that, i just noticed my thermometer doesn't measure over that [21:04] webPragmatist, not had that problem, last time i trashed it i did a full rebuild [21:06] incorrect: well… that's not very confident [21:06] if your drbd goes down… do you just find another job lol [21:06] i take it you haven't read the manual [21:07] and what full rebuild means [21:08] guess not [21:08] but i think i just did whatever a full rebuild is lol [21:08] basically i invalidated the secondary… and reconnected [21:08] it's sinking now [21:09] synching rather [21:09] you should rtfm [21:09] you should diaf [21:09] has anyone here successfully setup ldap authentication with Ubuntu 10.04? [21:09] Kyle__, sure i have it running over many nodes [21:10] incorrect: Did you need to do anything special to get the users to appear in the login list? [21:10] incorrect: I can login via console, but not via gui. [21:10] oh that is a PitA [21:10] you can either install nscd [21:11] or you can edit /etc/networks/interfaces and make sure the interface comes up fast [21:11] also you will need to restart gdm first [21:11] I think that got installed as part of my futzing aroudn to get ldap authentication working. [21:11] nscd that is. [21:15] kirkland : ping [21:16] wish i could figure out what is causing win7 under kvm to crash on shutdown [21:16] Deji: hey [21:16] Deji: okay, made some progress on your cloud [21:16] Deji: got the cc registered and working [21:16] Deji: and your NC registered and working [21:16] Deji: however, your NC has VT disabled in BIOS [21:17] Deji: so you need to go into BIOS on that system and enable virtualization [21:17] Deji: on your NC, try running "kvm-ok" [21:18] kirkland : kvm is diabled on the nc [21:18] Deji: yeah; you need to attach to the console, reboot, go into bios, enable it, and hard poweroff/poweron again [21:19] Deji: then your cloud should be operation [21:19] what does UEC use ? [21:19] Deji: we'll need to document this [21:31] if anyone is interested, if you want windows 32bit to run nicely give it a arch='x86_64' not i686 [21:31] damn that was frustrating [21:46] I can't remember a very important mysql command that is used to change the default settings to make it more secure I think _secure is even part of the command does anyone know what I'm talking about or have I just lost it? [21:52] if making it "secure" was one simple command, it would make no sense to not have that be the default [21:53] hrm [21:53] true [21:53] anyone know why heartbeat would lose quorum but not try to start the resources [21:54] um it removes the default settings like remote access [21:54] also its one command but there are a few steps to it [21:54] remote access is not enabled by default [21:54] mysql listens on a local unix socket out of the box [21:54] you have to configure it otherwise to enable remote access [21:55] does mysql come with like an example database? [21:55] I know I'm not wording that right [21:56] it has a built-in database which handles users and permissions [21:56] but there is not an included 'starter' database or anything like that [21:59] thanks for trying to help me qman__ [22:00] I'm going to have see if there's anyway I can get into my old notes that were on the schools server most likely not :( === elb0w is now known as elb0w`done4day [22:34] what is this refering to it cant remove LAMP ????? tasksel: aptitude failed (100) [22:35] hallyn: https://wiki.ubuntu.com/Mirrors/Scripts [22:36] i would be careful with tasksel remove [22:36] i used it earlier today and it has a bug [22:36] what is this refering to it cant remove LAMP ????? tasksel: aptitude failed (100) [22:36] deleted many packages which it shouldnt delete [22:36] hallyn: my mirror is a lenovo q100 "nettop" box, with an internal 500GB hard drive [22:36] hmmm [22:37] hallyn: tiny little dual 1.6GHz atom processor thing, with 2GB of memory, and 500GB disk, but does actually have gigabit ethernet [22:37] the hell... I am trying to get rid of Apache,php, etc from the aptitude and use the package from ApacheFriends [22:37] dominicdinada: are u running lucid? [22:38] Yes. [22:38] hang on then [22:38] ok [22:38] kirkland: and how much of that 500G is taken up by the mirror? [22:38] * kirkland checks .... [22:39] hallyn: i also mirror about a dozen or so of the daily ISOs too [22:39] hm, did they have that thing at frys? [22:39] yeah i'd want those too i recon [22:39] hallyn: /dev/sda6 449G 367G 59G 87% /mirror [22:39] heh, cuttin' it close :) [22:39] hallyn: they do have them at fry's; on sale, they're $179 [22:39] www.pricewatch.com you cant beat their pricewatch. Its like the stockmarket for computers [22:40] hallyn: yeah, that includes the full mirror, plus the ISOs i copy [22:40] hallyn: 7.0G /mirror/iso/ [22:40] i was figuring i'd just do a usb drive though [22:40] hallyn: let me warn you about a couple of things wrt to the lenovo nettop ... [22:40] dominicdinada: Ok, I installed LAMP earlier today with tasksel, these packages were installed, you could remove them with apt-get remove [22:40] http://paste.ubuntu.com/443581/ [22:40] hallyn: a) the graphics is crap, no accelerated graphics upstream yet SiS crap [22:41] (s'ok my monitors are in another state) [22:41] ok i am looking [22:41] hallyn: b) it comes with 1GB of DDR2 desktop memory, which I upped to a 2GB stick, since I had a spare [22:41] hallyn: c) it comes with 160GB hard disk, which I also upped to 500GB, since I had one of those too [22:41] you dont' use it as a pc at all though, right? [22:41] hallyn: you might do better to just spend $75 and get an external drive and attach to another system [22:42] yeah, that's what i'm thinking [22:42] hallyn: if you go the external drive route, I recommend getting a laptop size one, so that you can travel with it if you want [22:42] hallyn: having a full mirror on a plane is fun :-) [22:42] :) [22:42] my laptop won't last long on a plane though [22:42] hallyn: your sony? [22:42] 8 cpus, but only a few hours of batt life [22:42] yeah [22:42] not that i've tested it [22:43] haven't even had time to set up the nvidia to work with compositing (so i'm using dwm, bc metacity locks up) [22:44] Gotta config php.ini... [22:44] Boring... [22:45] before I do this Why would 100% working sites fail it is like URI parser is not working for instance with a codeigniter application i am getting all of these 404s [22:45] kirkland: so you just do a release mirror? [22:46] no matter what link I click on it does not work! [22:46] hallyn: no, not quite ... [22:47] My undestanding is that sun-jdk is provided by deb http://us.archive.ubuntu.com/ubuntu/ lucid multiverse - is that not the case? [22:47] hallyn: i think ubumirror should do most of what you need [22:47] hallyn: i do an archive mirror [22:47] hallyn: plus a handful of rsync's on the key ISOs i want from cdimage.ubuntu.com [22:48] ah, looks like add-apt-repository "deb http://archive.canonical.com/ lucid partner" [22:58] kirkland: can I have the QA document for Dell? I would like to see what will be tested there === unreal_ is now known as unreal [23:02] how do i stop a crm resource or even delete one… crm_resouce is frickin confusing [23:02] before I do this Why would 100% working sites fail it is like URI parser is not working for instance with a codeigniter application i am getting all of these 404s [23:02] no matter what link I click on it does not work! [23:02] oh nm i had to delete the entire group [23:25] where does logrotate ... log to? [23:28] i set up a logrotate job to run mysqldump [23:28] but i haven't seen any action yet [23:44] how do i delete an ms type resource with crm_resource [23:44] i tried crm_resource -D -r ms_drbd -t ms [23:45] Hi everyone [23:47] on a running ubuntu 10.04 LTS server, is there an easy way of adding/configuring a software RAID5 array using some sort of menu-driven software, like in the initial installation of ubuntu? [23:50] nm [23:55] anyone here? [23:56] wyleyrabbit: nope