[00:01] I am tempted to chmod -R 750 * for a bunch of places, including /var, /etc, /home - is this a bad idea? [00:01] generically, yes, it's bad [00:01] for example, ~/.ssh [00:02] jMyles: Yeah, don't do that. [00:02] Well I created a new user and I'm surprised how much access they have [00:02] I want them to only have access to ~ [00:03] for example, they can read everything in /etc/apache2/sites-enabled - I do not want this [00:07] How about instead of 750, I go o=X ? [00:12] how would I look up the cpu info ? [00:12] orudie_: cat /proc/cpuinfo [00:13] thanx === dendrobates is now known as dendro-afk [00:18] anyone know if I add /etc/event.d/ttyS0 so I can support a serial adapter, is there any way to apply that device without rebooting? [00:41] Hypnoz, sudo start ttyS0 [00:42] qman__: OH SNAP you da man! [00:43] qman__: was trying telinit q and telinit u, but I'm wondering if those only apply to the old init.d system, not event.d [00:43] yeah, all that stuff is gone away in ubuntu now [00:43] qman__: "start" and "stop" are for event.d I'd guess? [00:43] yes [00:44] and 'service' is now in ubuntu by default, for the rest [00:44] like the RH command, i can do service network restart? [00:44] yes [00:45] maybe some motivation to move up from 8.04 at some point [00:45] but I'm kinda hesistant to dist-upgrade all my servers ... if it ain't broke don't fix it [00:48] oi... just booted up an AIX machine. this thing is like a haunted house. [00:48] ksh default, and the up arrow moves the cursor upwards on the screen [00:48] hah... boy do I love gnu. [00:49] eek quick reimage [00:49] if you can believe it, some of our customers actually still use AIX. [00:55] And I bet service(8) ignores policy-rc.d(8) >:-/ [00:56] I'm considering trying out the Ubuntu-Server 10.4 beta1 for it's UEC. How reliable is that if anyone else by chance uses it yet? [00:57] Psi-Jack: i'm running it in a vm to host mediawiki and do miscellaneous tasks and it's working fine for me. [00:57] Oh yeah? Cool. I know 10.4 is due to release by the end of next month, so I hoped it was reasonably stable at the moment with all the freezes. [00:58] And 10.4 is the new LTS? [00:58] Psi-Jack: yeah, LTS. [00:58] Will be anyway. ;) [00:58] Psi-Jack: it seems like a lot of the issues, as always, are tied to GUI apps, so if you go headless you're much less prone to problems. [00:58] Psi-Jack: that being said, i just upgraded a gui install with zero problems. [00:58] Right, and my server is headless. ;) [00:59] I only plan to have GUI by UEC standpoints. [00:59] Which is totally web based as I understand. [00:59] Web/CLI that is. [00:59] Psi-Jack: well, join us in #ubuntu+1 if you're taking the plunge [01:00] Psi-Jack: beta2 comes out april 8th, so you may want to hold out another 2 weeks... [01:00] er... week and a half [01:00] I can always upgrade or dist-upgrade, no? ;) [01:01] These aren't live production servers anyway, just gathering ideas and trying out various options. [01:01] OpenSUSE with it's Xen Hypervisor approach, which seems really nice.. Citrix XenServer, which is effing awesome, but limited. And now, UEC, which I don't like some of it's approaches, but until I actually try it, I can't knock it. [01:17] for those in on the question re XDMCP, has anyone tried a simple ssh -X login with a gnome-session, and / or seen scripting to make that work well and automated for a client ? [01:17] fluvvell: ssh -X/Y work fine, as does Xnest [01:18] lullabud, my understanding was that xnest uses tsclient ? [01:18] fluvvell: i think that's vice versa... [01:18] heh. [01:18] In the UEC, what is Walrus? [01:23] lullabud, I think i mean that the client machine logs in with tsclient. I'd need to install xnest on the server. Prob I have is that we have virtualbox running with port 3389 in use and I haven't checked if I can do both at once [01:24] fluvvell: no, that's not true. tsclient is a client for logging in with various other methods, like MS RDP and such. xnest basically is network redirected X sessions, just like ssh -Y but with a windowed desktop. [01:25] fluvvell: you can get something similar if you ssh -Y and then run startx, but it's not in a window. xnest nests the remote X session in a window so you don't end up with two docks, two of everything [01:25] all overlapping each other... [01:26] fluvvell: first google link validates most of that... - http://en.wikipedia.org/wiki/Tsclient [01:27] ..that is, first google result for "tsclient" [01:30] fluvvell: oh, so back to server vs client, you don't need to install xnest on the server, you have to have xnest on the client. [01:31] lullabud, yes - though the nail I'm trying to hit is to replace XDMCP. My user was working well in Jaunty, when upgraded to karmic the XDMCP support failed. So she only needs to show one desktop, that of the session on the server. I've gotten around the lack of a server-locating utility in gdm by starting her client machine with kdm. But for some reason, (I suspect the version of xorg at this stage) there are no XDMCP servers showin [01:31] g. [01:32] oh, i see, i see... [01:32] then no, xnest is not what you need. [01:32] Sry, that was a long post [01:33] fluvvell: there are some conf files you can tweak, i forget which, that enable/disable xdmcp. i think it's xorg.conf... might have some notes around somewhere.... [01:33] no, but If I could think through the ssh /startx thing, I could probably work it out. [01:33] lullabud, yes. I've just been down to the site and I think the fact that he has the ATI commercial x server stuff installed, thats where the issues begin - that and possibly compiz [01:34] hmm. compiz, i don't know... never done that with xnest. [01:34] which i should disable, and start with a bog standard setup. [01:35] and the fact that its a cafe and its the lunch /2h period, I've skipped away to lick my wounds and think it out. :-) [01:37] well, they are surely moving away from xdmcp. i hadn't used it in a long time, but looking now it's clear that they're doing away with it for some reason. [01:37] they say it's insecure, which i can't validate... [01:37] perhaps they're moving to vnc or something... [01:38] vnc has never been reliable for me in linux though, so it's disappointing to see xdmcp being dropped. [01:39] the xdmcp support on the 6.06 livecd saved my ass one time when i had to configure a solaris machine that had no monitor port. [01:39] What's this about dropping XDMCP? [01:40] twb: eh, gnome is moving away from xdmcp [01:40] twb: it's no longer a login option to connect to a remote machine rather than the local... [01:40] and it's missing from other places too [01:40] FSVO gnome = gdm? [01:40] fsvo? [01:40] For Some Value Of [01:41] uh... yeah, gdm-2.20 is now considered "legacy" [01:41] and it's the last version (afaik) that supports xdmcp [01:41] the newer versions do not support it, and are default in 9.10 and 10.04 [01:41] Annoying, but as long as Xorg (and xdm) continues to support it, I don't care too much. [01:42] I wonder what the LTSP people had to say about it [01:42] well, that's part of my disappointment... i just did "grep -ri xdmcp /etc/" and didn't find it in anything xorg related on 10.04 [01:43] there didn't seem to be much development going on in the ltsp crowd either [01:44] but the XDMCP indifference is definitely at Gnome level rather than Ubuntu [01:45] RFB is retarded compared to X11. [01:45] and it *is* insecure, but what with ssh or ssl surely there is a way to seal any cracks - [01:45] It's only when you get insane amounts of layered gradients, like in GTK2 or OpenOffice or xulrunner, that RFB actually starts being faster. [01:46] Oh, and bloody client-side font rending [01:46] RFB ? [01:46] fluvvell: RFB is the protocol that VNC is a thin wrapper on top of. [01:47] fluvvell: oh, well, it looks like it's dead simple to enable xdmcp server-side stuff in 10.04. edit /etc/gdm/gdm.schemas, just change "disable" to "enable" and it works.... [01:47] i jsut tested it between OSX and 10.04 [01:47] http://en.wikipedia.org/wiki/RFB_protocol [01:47] twb, yeah I'm using vnc in a number of cases, and while it solves the immediate problem its not the answer to everything [01:47] lullabud: my last OS X host was a glorified X thin client ;-) [01:47] fluvvell: RFB is a "lowest common denominator" protocol. [01:48] twb: yeah, but it was probably the thinnest client ever. [01:48] :P [01:48] following after steve job's as the thinnest ceo ever.... [01:48] s/job's/jobs [01:48] lullabud: it was a 330MHz bondi blue running 10.2, acting as a remote head for a 3GHz workstation [01:49] hey, g3's are still useful! [01:49] now i'm going to have to run out to the curb and get that one that's next to the trash can... === dendro-afk is now known as dendrobates [03:31] New bug: #552238 in postfix (main) "package postfix (not installed) failed to install/upgrade: subprocess new pre-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/552238 === rgreening_ is now known as rgreening === manjo is now known as manjo_ === manjo_ is now known as manjo [04:17] I just set up parted (from the expert install of the alternate CD) with a RAID1 with /boot on it and all three drives it mirrors have a bootable flag. And a RAID5 with a lvm on it that contains root filesystem and swap space. Will my kernel be able to find and mount the root filesystem at boot? [04:17] Try it and see. [04:17] I will. I am hoping for the best twb. [04:18] Finishing install of base system now... [04:18] As you were told yesterday, this is the wrong channel for support for desktop ("desktop" and "alternate" CDs) installs. [04:18] Nobody knows what a LVM on that channel is... [04:18] that doesn't surprise me [04:19] Well, we *did* suggest you use the server install CD. [04:20] Right, I will probably have to do that if this fails. Even if I do drop to command line to set up fstab and mdadm, how can I be sure what I am doing is right. I first started with gentoo and followed the docs and got help on IRC and that had a kernel panic where it couldn't find the root partition. [04:23] Maletor: I can help you if you go to the debian-installer partition manager (d-i partmgr) step on the server install CD. [04:23] And assuming a clean install. [04:24] Doing it any other way is more effort than I can be bothered going to on your behalf. [04:24] I may take you up on that offer twb. I'll let you know how this goes. [04:34] Maletor: it should be able to, though there's been some trouble in lucid booting off of lvm+raid. [04:35] sbeattie: thanks, yes, i've heard lucid is very unstable from pretty much everybody [04:37] Until it's released, it is /by definition/ unstable. [04:37] * sbeattie eyes his two lucid systems carefully. :-) [04:37] I installed karmic server on a raid 1 with LVM just yesterday [04:37] it was smart enough to figure it out [04:37] That's not the same as "buggy" [04:37] which surprised me greatly [04:37] since it's the first time that's ever happened for me [04:38] I wouldn't have said it was pretty unstable. but I don't have any of my raid machines running it. === Maletor_ is now known as Maletor [04:46] hqman__: did you install from server? [04:46] qman__ [04:50] yes [04:51] Maletor, you could always install ubuntu server, then install the ubuntu-desktop package on top [04:51] of course the GUI is not supported here [04:52] but the install process would be [04:52] Houston, we have lift off. [04:52] Let me just verify the RAID and LVM and I will be ecstatic. [04:53] cat /proc/mdstat [04:54] And what is the command 'dh -i' for where filesystems are mounted? [04:54] I usually use 'mount' to do that [04:55] and df -h to check filesystem usage [04:55] Yup, everything is working magically. This is fantastic. [04:58] lvextend works too. This is sweet. Thank you everybody. [05:00] OK, I have a host here with three rtl8169's in it. mii-tool can't tell me which one is connected, because apparently the driver doesn't support that ioctl. [05:00] How can I non-destructively determine which NIC has a cable plugged in (so I can label the NICs with eth0 through eth2)? [05:01] Hmm, I suppose worst case is I can connect it to another host (instead of just a switch), and stick pings over a transiently-static configuration. [05:01] twb, mii-tool works for me on rtl8169s [05:02] those are what I mostly use, since they're a great way to go gigabit for about 15 bucks [05:07] is there a way to check which computers are connected to samba ? [05:08] maxagaz, I'm sure it could be done with netstat, though there may be a higher level tool for it [05:08] maxagaz: smbstatus maybe? === aetaric_ is now known as aetaric [06:55] hey guys im having issues with vi over ssh [06:55] eagles0513875: expand. [06:56] twb: basically work in vi then all of a sudden it feels like it freezes [06:56] eagles0513875: you're hitting ^S. Type ^Q to fix it. [06:56] im not hitting annything [06:56] this is while typeing lets say some programming in c++ [06:56] OK, what kind of line speed are we talking? [06:57] lan [06:57] 100mbps [06:57] server nic being gigabit [06:57] and desktop im on gigabit as well [06:57] what even stranger it only happens with vi [06:57] OK, check the load average and the iowait on the remote host. [06:58] atm i have no load on it [06:58] im the only one who has access to it [06:58] if i use nano this doesnt happen [06:58] Talk to #vim, then. [06:59] twb: could it be an issue with ssh [06:59] cuz it seems from what i can tell by the ssh program im using it ends up one way the connection [06:59] the server is getting input just not sending back output [07:00] How did you determine this? [07:00] im using xshell ssh program on windows and it has arrows to show input to the remote machine and whats getting sent back out to said client [07:02] Just because your client receives nothing doesn't mean the server isn't sending anything. [07:02] If you kicked out the cable, you'd see the same behaviour. [07:02] i havent been anywhere near them [07:03] Ask dmesg if the link has been bouncing lately [07:04] I get that kind of behavior when I have a poor wireless signal [07:04] moving my laptop to get a better signal fixes it [07:04] but, on a LAN [07:04] it would have to be bad cables or faulty hardware [07:04] qman__: or something like NM playing silly-buggers [07:04] or some bizarre software issue [07:08] Ah, xshell apparently proprietary crap. [07:08] twb: and qman__what doesnt make sense though is why does it only do this on vi and not anythign else like nano [07:08] twb: i like it so i can consolidate all ssh connections into one program with multiple tabs [07:11] I presume there are extraneous factors that prevent you simply running Ubuntu on your workstation [07:11] twb: said desktop is windows said machine in question is nix [07:12] bah got no help in vim [07:12] random thought [07:12] since im on 64bit ubuntu server should i install ia32libs to be save [07:12] I think you're misdiagnosing the symptoms, and I'm still waiting for the results of the diagnostics I asked you for. [07:15] twb: not showing anythign out of the ordinary in dmesg and syslog [07:15] Can you reproduce the symptoms on demand? [07:15] twb: i can reproducd them yes just have to work with vi for a few [07:16] Can the server address your Windows desktop directly (i.e. no NAT)? [07:17] im using static ips on both and yes it can [07:17] On the server, run "mtr ", and hit d once. [07:18] You should see horizontal line(s) of "." crossing the screen. [07:18] When there is no network, it'll print a ? instead of a . [07:18] didnt do anything like that [07:18] it just brought up an empty traceroute [07:19] You need to hit "d" [07:19] i did [07:19] i have some latency values [07:20] eagles0513875: you should have at least one line in the table, starting with "1. " [07:20] going to take a few need to hook up my monitor directly to the server [07:20] Whatever [07:21] twb, I get nothing of the sort doing that [07:21] twb: if im doing it via ssh its not givine me an ip but latency values [07:22] It should look like this: http://paste.ubuntu.com/406887/ [07:22] yeah, it doesn't look like that [07:22] im not getting anythign of that sort either [07:22] ...except with fewer entries, because there aren't any hops on your LAN [07:22] also, that's what page it opens to [07:22] but I don't get anything at all in the list [07:22] me neither [07:23] Try mtr 8.8.8.8 [07:23] that works as expected [07:23] Or even "mtr 127.0.0.1" should have at least one entry [07:24] ah, working now [07:25] I don't know why it isn't/wasn't working for you. [07:25] 8.8.8.8 and lo device both have something for me [07:25] I don't either [07:25] You could try mtr -n , if it's a DNS issue [07:25] maybe something wrong with the client IP I was using [07:25] when you say workstation ip do you mean teh host or destination [07:25] taht you are on [07:25] eagles0513875: AFAIK you have two hosts, a workstation (running xshell) and a server (running sshd and vim). [07:26] On the server, run "mtr -n ' [07:26] ok [07:26] nothing still [07:27] Then your network's totally buggered, and you need to make that work. [07:28] :-/ [07:28] eagles0513875: are you the network admin, there? [07:28] its a small home network [07:28] lol only one in this family who understands in computers [07:29] wait a min [07:29] could me having bind installed but not configured be the issue here [07:29] at least to networking [07:29] not really [07:29] cuz i tried to ping my current machine im on from server and it wasnt resolving [07:30] it would have an effect but by IP things should still work [07:30] super high latency more like it [07:30] well, there you go [07:30] found the answer [07:30] find out what's causing the latency and fix/replace it [07:30] restarted networking [07:30] and its fine [07:30] so its a bug in networking somewhere in ubuntu server [07:31] qman__: it wouldn't surprise me if he has an appliance router or switch, and it got confused after a power outage. [07:31] no power outages here though twb [07:31] im on a linksys [07:31] yeah, I've had that happen before, as well as them burn up [07:31] Stupid cheapo switches... [07:31] qman__: think im goign to buy a new one cuz this one is starting to piss me off [07:32] I use a PC running ubuntu server with 4 NICs in it as my router [07:32] Real Soon Now manglement will turf them and actually fix the wall points, so that we can just use the UPS-backed switch in the rack. [07:32] but I have a somewhat complex network [07:32] Oh yeah, and replace the battery in the UPS :-/ [07:32] heh, yeah [07:32] UPS makes all the difference, even without power actually going out [07:33] everything just works happier [07:33] * twb hits manglement with a trout [07:34] * eagles0513875 tries vi again [07:35] not happening again [07:36] seems like its a bug in ubuntu networking [07:36] cmeiklejohn, yes smbstatus! [07:36] cmeiklejohn, thanks [07:37] eagles0513875, a hardware problem is far more likely [07:37] qman__: this server is brand new [07:37] qman__: as it only does this with vi and no other editor [07:37] cables, your linksys [07:37] that stuff does get flakey sometimes [07:37] qman__: true [07:37] qman__: after resetting networkign on ubuntu its fine and happy [07:38] could also be a problem with the NIC driver [07:38] or if you've got networkmanager running [07:38] which you shouldn't, networkmanager just causes problems [07:38] trust me i know [07:38] about nm [07:39] im nothing but commandline on my server [07:39] tbh kinda prefer it [07:39] the nic's are both broadcom nics [07:39] one is a pcie nic the other is a normal nic both gigabit [07:39] what do you mean by "normal" [07:40] NM is a headless daemon. [07:41] qman__: normal pci nic im guessing and the other uses the pcie bus [07:41] it did it again [07:41] froze or lost connection [07:46] New bug: #552360 in squid (main) "package squid 2.7.STABLE7-1ubuntu11 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/552360 [07:51] * eagles0513875 will have to try ssh via my macbook to soee if same issue with vi locking up occurs [08:08] wow... incredible how some people pay up to $3 700 for having RHEL pre-installed on their servers [08:08] I wonder if they come with something beside installation and upgrades [08:12] <_ruben> support probably [08:18] _ruben, right... forgot about that one, he he [09:11] hi [09:17] sebsebseb_, hi === sebsebseb_ is now known as xylos [09:25] hi [09:27] model name : Intel(R) Xeon(R) CPU E5504 @ 2.00GHz [09:35] do i need to install amd64 or is there a x86_64 hardy image ? [09:35] amd64 will do [09:38] Jeeves_: Thanks [09:39] so there is only i386 and amd64 image only ? [09:40] afaik, yes [09:44] <_ruben> amd64 == x86_64 [09:45] <_ruben> as in: the amd64 install media install a x86_64 kernel [11:39] How would you determine the ftp software on a server if you are sitting at the client. Is it even possible? [11:39] alvin, if you telnet to it, it should say the software and version [11:39] nimrod10: Thanks! [11:43] Hi, I want to restrict access to the nbdrootd service to local networks only. Is the following correct? [11:43] a) in hosts.deny, put: nbdrootd: ALL [11:43] b) in hosts.allow, put: nbdrootd: 127., 10., 192.168., 172.16.-172.31.: keepalive === swift__ is now known as swift === swift__ is now known as swift [12:42] how to dpkg use wildcard character [12:47] dpkg -l mysql* ? [12:50] obea: never tried but *if* it works then you should put quotes around the pattern to hide the * from the shell [13:06] yeah dpkg -l ¡®*mysql*' thx! [13:06] ;) === rgreening_ is now known as rgreening [13:45] can anyone comment on mdadm using "raw" disks (/dev/sdx), or partitions (/dev/sdx1) and which is better and why? [13:48] Hi, is it possible to setup an ubuntu entreprise cloud across multiple data center ? [13:49] proppy: it is possible to setup separate "clusters" (availability zones), but they will depend on the same "Cloud controller" front-end. [13:50] ttx: is LAN connection required between all the "clusters", and the "controller" ? [13:51] proppy: yes. Otherwise you should install two separate "clouds" [13:51] well, not LAN connection [13:51] can be WAn [13:52] the cluster controller and the cloud controller need to be able to talk to each other, but not necessarily on the same network. [13:53] one "cluster" = one "physical host" ? [13:54] one "cloud" = multiple "physical host" on the same network ? [13:54] no, a cluster is made of a cluster controller, a storage controller, and one or more node controller [13:54] oh ok, "node controller" = one "physical host" [13:55] node controller is one physical host running lots of VMs, yes. [13:55] so we can have a "cluster" in house [13:55] another in running on a dedicated server [13:55] and having them part of the same "cloud" [13:56] given that each cluster controller is able to talk to the cloud controller [13:56] is it possible to provision image on multiple clusters ? [14:02] thefish: I have no idea, but the Debian installer uses partitions if you set up md devices during install [14:03] alvin: cheers, ubu does the same [14:03] but ive set up raid6 here on raw devs [14:03] i cant think why that would be bad [14:03] thefish: I meant Ubuntu actually. Ubuntu uses the Debian installer. (unmodified?) [14:03] me neither. Go ahead [14:04] if mdadm scans /proc/partitions it sees both [14:04] If I set up LVM, I also use raw devices. (Same for ZFS) [14:04] ttx: thanks for all these informations [14:04] im going to stick to the debian method, im sure theres something they know that i dont [14:04] ye ill be using lvm on top of this md [14:05] proppy: yes, the image registered in one "cloud" should be available to all "clusters" [14:05] For ZFS, it is recommended to use raw devices, but I'm pretty sure the reason for that is not valid for mdadm. [14:10] ttx: is there some kind of synchronisation available for the multiple storage controller, across the cloud [14:11] for EBS-like storage [14:11] proppy: I think multiple SCs will sync, but I'm not exactly sure, sounds like a question for #eucalyptus [14:11] ttx: yes, I will ask there, thanks a lot for your help [14:11] proppy: np [14:19] ¡° /usr/sbin/dpkg-reconfigure: uptrack is broken or not fully installed === dendrobates is now known as dendro-afk === RoyK is now known as Guest19685 [15:15] Hi, does uec support non-VT hardware throught Xen or Qemu/kqemu ? [15:18] proppy: Kirkland should be able to give a conclusive answer but I believe the reponse will be no as UEC uses KVM. [15:19] the fish: re: MD on top of raw partitions. According to Steven Ellis at Linux.conf.au 2010 : [15:19] http://mirror.internode.on.net/pub/linux.conf.au/2010/tuesday/index.html [15:20] They are not so good - makes it harder to move disks to another system (backup/recovery). A single partition takes a very small amount of space and is more compatitible with the tools. [15:20] Even : thefish :-) [15:20] sherr: ok ta [15:20] sherr: same for lvm? [15:21] thefish: I use LVM over MD over partitions. Works well. [15:22] sherr: so just pv on eg /dev/md0 ? === klaas- is now known as klaas [15:27] hi. solaris has a command, pstack, to see the stack from a coredump - any idea if linux has something similar? the pstack available in ubuntu doesn't show this [15:27] thefish: yes, that's right. [15:28] cool, thanks sherr - thats what ive usually done [15:46] proppy: Omahn: well .... i suppose it kinda does support non-accel QEMU, as UEC just runs "kvm", which by default, if no VT hardware is present, it just disables kvm-acceleration and runs in godawful-slow mode [15:47] proppy: Omahn: as to the other points, upstream Eucalyptus supports Xen, I believe, but Xen in Ubuntu is universe-supported, and very much untested with UEC [15:47] proppy: Omahn: and kqemu doesn't exist anymore, sorry. === dendro-afk is now known as dendrobates [16:13] ttx: a bit of PM? [16:15] Why is this command not working? [16:15] $ sudo qemu-img -c -O qcow2 disk-raw.img disk.img [16:15] Oh,... I know [16:16] (Sorry to bother. The 'convert' part was missing) [16:21] hggdh: just a sec [16:26] hello. i have a server that now is a NX server for several clients on the office [16:26] now i want just to export /home with nfs and use some kind of Network Authentication for account [16:27] the nfs export goes fine... but when it comes to the central network authentication thing i don't know really which tool plays what role. OpenLDAP, Kerberos... [16:28] * zul dislikes dentists [16:31] ihernandez: I'm also interested in that. Let us know if you find documentation that works with Ubuntu [16:33] alvin, i have been told by a user on #ubuntu that with ldap + nfs all users can login from all machines and have they data [16:33] the nfs docs on ubuntu site are great, and they work pretty well [16:33] Yes, they should. [16:33] https://help.ubuntu.com/9.10/serverguide/C/network-file-system.html [16:33] VMBuilder.exception.VMBuilderUserError: No such hypervisor. Available hypervisors: vmserver esxi xen kvm vbox vmw6 [16:33] where did qemu go in 10.04? :D [16:33] but ldap docs are a nightmare [16:33] zul: ^^ ? [16:34] ihernandez: which ldap document is wrong? [16:34] ihernandez: which version of ubuntu are you using? [16:34] I'm about to upload a debdiff that (hopefully) resolves an old issue in exim4 in hardy. Once I upload the debdiff should I subscribe ~ubuntu-sponsors to get someone to build the -proposed package? [16:35] There is nothing to find about NFS4+kerberos, and there are bugs filed against the ldap documentation. ldap has split the packaging. For someone without experience, the whole ldap story is very confusing. [16:35] http://doc.ubuntu.com/ubuntu/serverguide/C/openldap-server.html [16:35] ldap ^^ [16:36] looks better than last time, thanks [16:36] I'll try them out in a few weeks [16:36] that's for karmic and newer [16:38]