[00:00] http://pastebin.ubuntu.com/446338/ <- somehow the echo status line gets eaten [00:02] ccheney: which of the echo lines? [00:03] the echo "status = $?" doesn't appear [00:03] i think maybe exec causes it to not be printed === AdamDV is now known as Guest1506 [00:03] i am configuring a new ubuntu server hiwch is working as a dhcp server (i think) but i can't ping it, ssh it, or get samba shares off it. is there anything obvious i'm missing? is it firewalled by default or something like that? [00:03] another option is to write the logs to individual files [00:03] hey guys, has anyone here installed Open-Xchange on teh new Ubuntu 10 ? [00:07] nm this idea isn't well thought out [00:07] i think it won't work period [00:08] i think the only easy way to make it work properly will be to log into a directory with full timestamp in the filename [00:11] ccheney: yes, I agree [00:13] my sshd config file says i'm listening on port 22, the service says it's running OK, but netstat doesn't report i'm listening on 22 [00:15] hggdh, ok i think i have a script that will log everything properly now :) [00:15] i got one status 5 so far [00:15] /sbin/vgcreate vg-1n2UFQ.. /dev/loop60 [00:15] /var/lock/lvm/P_orphans: flock failed: Resource temporarily unavailable [00:15] Can't get lock for orphan PVs [00:15] status = 5 [00:17] another thing....service smbd status says it' [00:17] cool i think this might actually be useful [00:17] says it's running, but sevice --statusall shows a ? [00:17] ccheney: maybe we are using all lock memory [00:17] ? [00:18] hggdh, hmm maybe so how does that work? [00:18] ccheney: good Q, I do not remember anymore... [00:18] * hggdh goes back to RTFM [00:21] i've upgrading this server from karmic to lucid and the only thing that's running is dhcpd...my old config files are intact...i'm lost [00:25] hggdh, this is weird i'm seeing what appears to be euca trying to force remove active pv's [00:26] ccheney: you mean some other pv (and not the one that would be removed? [00:26] oh i see why now [00:29] hggdh, my grep was a bit faulty and matched too widely [00:29] oh, OK [00:29] hggdh, so now i see how it is working [00:30] ccheney: is there contention for the lock? [00:30] it calls eg, lvremove that works, then calls vgremove and it fails, then calls pvremove without caring it failed [00:30] without caring vgremove failed [00:30] vgremove failed due to [00:31] /sbin/vgremove vg-g7thcw.. [00:31] /var/lock/lvm/P_orphans: flock failed: Resource temporarily unavailable [00:31] Can't get lock for orphan PVs [00:31] status = 5 [00:31] so maybe contention for lvm lock [00:31] I had a feeling euca was not worried about failures when I was cleaning up the mess I was left with [00:31] but since euca doesn't check its return codes and continues on like nothing is wrong it dies horribly [00:32] ok, well got to run and pick up food for my wife, bbia 30-45m [00:32] k [00:32] i wouldn't be surprised at all if the losing ip issue is the same type of problem, not checking error codes [00:33] i'll follow up to the bug report when i get back [00:34] this is sooo wrong... === unreal_ is now known as unreal [00:48] mathiaz: acked on the format change [01:05] mathiaz (or zul): the samba apport hook should be initially packaged for Maverick or Lucid? [01:13] there is already one for lucid [01:13] hggdh: so patches accepted [01:13] zul: yes, I am adding the test for /etc/samba/smb.conf [01:13] hggdh: cool === matthew is now known as Guest22728 [01:25] zul: I submitted a merge request for maverick (and added you as reviewer to my branch). I *think* this is what I should have done, but I am not sure [01:28] hggdh: actually no...the right thing to do in this case is to open up a bunch in launchpad with the patch and then nag me to death ;) [01:29] zul: a bug, you mean? [01:29] darn! [01:29] hggdh: yep [01:30] zul: doing it right now. And please consider yourself being nagged to death ;-) [01:30] hggdh: ok thanks [01:30] hggdh: yay! [01:30] :-) [01:33] oooo [01:33] elinks has pretty colored build scripts [01:43] zul: bug 591043 opened, can you please review? [01:43] Launchpad bug 591043 in samba "merge request: changing source_samba.py" [Undecided,New] https://launchpad.net/bugs/591043 [01:44] hggdh: yep first thing tomorrow [01:44] zul: thank you. Please correct me as needed === oubiwann_ is now known as oubiwann [01:45] I am running Ubuntu Server on VirtualBox. I am not using X Windows, only the command line. It looks like it is redrawing the screen for every line that causes the screen to shift down. In other words if there is empty space on the screen below the current line, it works normally, but if the current line is at the bottom of the screen, when the system moves to the next line it refreshes the whole screen line by line. This causes [01:45] Why does Ubuntu Server redraw the screen for each line? Is there a way to make this behaviour go away? [01:46] is it related to the framebuffer? [01:50] * ccheney back [01:51] New bug: #591043 in samba (main) "merge request: changing source_samba.py" [Undecided,New] https://launchpad.net/bugs/591043 [01:53] wtf [01:53] returnValue = removeVolumeGroup(vgName); [01:53] returnValue = removePhysicalVolume(loDevName); [01:53] yea... store the returnValue and then just ignore it [01:58] and lvm does not appear to document the return codes it uses === sentabi_ is now known as sentabi [02:03] heh [02:04] ccheney: good chances of getting it better there. I really cannot understand you errors are not orked on [02:05] i posted an unspecified rant about this mess, heh [02:05] ccheney: care to give the link? [02:05] Why does Ubuntu Server redraw the screen for each line? Is there a way to make this behaviour go away? [02:06] http://chrischeney.wordpress.com/2010/06/07/error-handling/ [02:06] heh :) [02:06] dammit, how can I have so many errors in one single line? [02:06] hggdh, can't parse your comment as it had too many errors :) [02:07] ccheney: there you go... bloody dixlexia, plus a glass of wine :-( [02:07] dammit! again [02:07] hggdh, once i dig around in lvm and determine what kind of error code 5 is i will try to fix up the euca code [02:07] ccheney: roj. Glad to have you working with us :-) [02:08] hggdh, :) [02:08] hggdh, i worked for several years at a place that was pedantic about checking every error code even if it was near impossible to actually ever see the error [02:09] iirc some of the errors we checked for would have likely never gotten to the code path due to major system failure if it actually happened [02:09] ccheney: that helps, indeed. I worked on a *lot* of places where errors were only important if management saw them [02:09] heh [02:10] that's it. I am out for the night [02:12] my first guess os is create, how can i access it? [02:20] hggdh, oh yea and better, we had to document our functions [02:20] hggdh, so people actually knew what they were supposed to do and return, etc :-\ [02:20] * ccheney looking at lvm and has no clue what it is doing with its return values, weird crap [02:22] ah i think i found where the error codes are, but they aren't actually useful [02:22] lib/commands/errors.h [02:32] anyone using libvirt? [02:32] how to access my new guest os? [02:35] unewbie: look at virt-viewer [02:41] mathiaz: do i need a desktop for that? [02:42] can i just ssh? === jballard_ is now known as jballard === unreal_ is now known as unreal [03:08] ccheney: I will look at this tomorrow, done for today/night [03:08] ok [03:08] have a good night === ajmitch_ is now known as ajmitch [03:09] ccheney: you too. Please call it a day also, otherwise I will feel guity ;-) [03:09] hggdh, ok :) [03:09] * ccheney just noticed its 9pm here already, heh [03:20] it's 9 am here :D === bc_ is now known as bc [04:06] hi [04:06] all [04:07] what is the command to update the system... everytime i turn in on it says there are 27 updates availabale.. how do i run them/ === geneticx is now known as acron [04:21] how can i connect my guest os with kvm [04:30] apt-get update [04:30] apt-get upgrade [04:30] Yosi: may be that's what you need === jldugger is now known as pwnguin [05:22] Do SATA optical drives work in 8.04? [05:23] I remember that 8.04 d-i doesn't handle them, but I don't remember if the installed system will. [06:24] ahhh [06:24] ! [06:24] anyone here? [06:24] many people here [06:24] ! [06:24] just ask :D [06:25] alright i tried to ask this in the ubuntu channel but it got washed away [06:25] so [06:25] i'm trying to follow this guide here: [06:25] err... accidentally copied another link just a moment [06:25] http://www.debuntu.org/how-to-set-up-a-serial-console-on-ubuntu [06:25] ERR... no not that one either [06:25] http://derrick-caluag.blogspot.com/2006/10/how-to-setup-dial-in-server-on-linux.html [06:26] that one. i need to be able to connect to my headless server box at any time [06:26] and the only connection it has is wireless [06:26] and sometimes it fails to start [06:26] so since it's headless and if wifi goes down... well i can't do anything really. i can try to do it blind but sometimes that doesn't work :/. So I want to be able to dial into it. both pc's have modems [06:27] and its running 32bit server lucid [06:27] but /etc/inittab doesn't exist in these versions, nor does /etc/event.d apparently [06:27] so im not even sure how to do this... [06:28] so if i could have some help with that guide that would be excellent [06:29] Mgamerz: /etc/event.d because /etc/init for no good reason [06:29] ? [06:30] Sorry, predictive text [06:30] Mgamerz: /etc/event.d became /etc/init for no good reason [06:30] hmm [06:30] inittab is gone and its supposed to bind to an interface [06:30] im not sure what the dial up/telnet terminal would be labeled [06:30] although since it is a 'LAN' connection (of sorts), I assume telnet would work fine. [06:30] just having to bind a listener to the ubuntu side [06:31] I have a router running 10.04 that has a serial interface. I can dig out the /etc/init/ttyS0.conf entry if you want. [06:31] that'd be great :) [06:31] But I'm a bit confused because you're talking about serial and wifi, which is unrelated. [06:31] although is a modem (with phone line) the same as serial? [06:31] Well [06:31] the thing here is: [06:31] No, a modem is different. [06:31] ... -_- [06:31] I have a server [06:31] and i don't have access to the router from where i can put it [06:32] so i have to use wifi [06:32] but occasionally the wifi... eh just dies [06:32] and since the server is headless, i can't just do ifdown ifup [06:32] well i can but if i mistype something then well... i dont wanna have to move this monitor again and again [06:32] Mgamerz, you probably wont have telnet, but if it is server install you should have ssh [06:32] http://pastebin.com/qxBYPwHK [06:32] so if i could just dial into the server [06:32] and use a terminal that'd be excellent [06:33] Mgamerz, so are you on the network or not? [06:33] yes. [06:33] I am [06:33] telnet is also unrelated to serial lines, modems, and wifi. [06:33] Mgamerz, does the machine have an ip address and do you know it ? [06:33] i can access the server right now [06:33] im vnc'ed to it (have tightvncserver) [06:33] mgamerz .. take a look at this link ... the issue will be in finding all the pieces to get wvdial to work properly ... [06:33] http://ubuntu-virginia.ubuntuforums.org/showthread.php?t=1374108 [06:34] Mgamerz: to clarify: are you talking about a dialup (56k) modem? i.e. not ADSL, ATM, etc. modems? [06:34] yea [06:34] a pci modem card [06:34] one of the ones from the 90's hehe [06:34] too cheap to buy a null cable [06:35] A null cable is likely to work [06:35] How much is your time worth? [06:35] well i have a ton of time [06:35] Mgamerz, just to verify, the server has access to the internet - but you have no access to its router? Are you trying to access it from a remote spot with no internet ? [06:35] so not much [06:35] the server can access the internet [06:35] it can access everything i want it to [06:35] BUT!!! [06:36] sometimes the wifi that connects it to the network dies [06:36] and since it is headless, i have NO way to make it come back up without forcing a power off (not shutdown) [06:36] fluvvell: I think he has a headless box with a wifi NIC, a serial port on the motherboard, and a PCI 56k modem card. [06:36] what is the distance between where your remoted computer is and where the server is? [06:36] Mgamerz, and there is no way to wire it to the router; I think I understand [06:36] about 5 feet [06:36] my telephone wire is barely long enough to reach that far [06:37] and I'm using my ethernet port on my remoted machine [06:37] He wants to connect to it over a RJ11 cable or something [06:37] I have a mini wifi router here that also acts as a passive wifi adapter [06:37] Mgamerz, What about using a reverse network patch lead? [06:37] all you need then is a 10' (foot) cable to connect to the router then .. you don't need to worry about hooking up to wireless .. [06:37] noooo [06:37] my box is wireless too [06:37] the router is like 40 feet away [06:37] azteech, my thoughts too. [06:38] Yeah, BY FAR the easiest solution is to put in a $10 100baseT ethernet card and run a cat5 cable. [06:38] ... [06:38] then all you really need is a hub ... to connect the server and the remote unit ... and you can run cat5 cable up to 150 feet ... with no problem ... [06:38] i don't have access to the router itself [06:38] If you can't do that, the next easiest is to buy a nullmodem cable and talk serial between two serial ports. [06:38] Hi All, I hav two drives in my system.. a single sata drive on the internal controller and a RAID on a controller card.... when i installed ubuntu on the raid, i think it put the bootloader/grub on my single sata drive, is that a problem? everything seems to work fine.. [06:38] and if i want to not spend money i could get this to work [06:38] Yosi: yes, it's a problem. [06:38] Yosi: if either drive dies, you won't be able to boot [06:39] true... [06:39] better having the bootloader on the raid... [06:39] i do not feel the need to spend $10 on something that i could do for free with a bit of work :l [06:39] twb: any easy way to move it over, or can i have the bootloader on both? [06:39] Yosi: first, confirm your diagnosis by trying to boot without sda [06:39] i.e. with it unplugged [06:40] Yosi: then, you will want to install grub on all drives. Google for the phrase "setup (hd1)". There should be adequate documentation online somewhere. [06:40] i disabled the internal controller that its on, any nothing worked.. but maybe it better if i just unplug the drive [06:40] why don't you have access to the router? distance, or is it someone's elses hardware? [06:40] (You could also try "grub-install /dev/sdb", but I don't trust that.) [06:40] it is ours [06:40] but the router is a long ways away [06:40] and we have no power outlets that are open next to it [06:40] it's in our living room, and this server is... well... 'ugly' [06:40] as my parents say [06:41] so it lives in my room [06:41] twb: is there an easy way to check where the bootloader is located, i can't unplug the drive easily, i forgot the case is now locked... [06:41] i have looked on the internet and everywhere i go people always just say do this do that instead of just helping with the person's questions [06:41] so i guess [06:41] where can i find ubuntu's version of /etc/inittab? [06:41] Yosi: it's difficult to check, but it should be easy to blow it away with something you think is correct [06:41] Mgamerz: you can't. [06:41] would be the answer i am looking for [06:42] ... [06:42] okay ... then what you need to do work on, is how to get wv to work on lucid ... like I said above, and what has been said in the link I pointed you to, is to get all the pieces together ... and set up wvdial to make it work for you .... [06:42] Mgamerz: in all current versions of Ubuntu, inittab support is negligible. [06:42] i know it is [06:42] thanks, brb [06:42] i just need to know the thing it turned into [06:42] Mgamerz: it turned into /etc/init. [06:42] like how can i initialize this on boot [06:42] hmmm [06:43] There is no direct analogue of the old inittab [06:43] the problem is you are using a how-to that was done in 2006 .. there has been a lot of changes made since then ... recommend looking at the man page for wvdial . and google for wvdial to see what the latest is for lucid ... [06:44] k [06:44] believe twb is correct .. all that is now /etc/init .. [06:44] i'll see about that [06:44] yea that is the only guide i could find [06:44] i figured it had probably changed but it's the only one i could find about a dialin server [06:45] I *know* I'm correct, because it has pissed me off several times before [06:46] google for wvdial lucid .. or ubuntu 9.xx and see what you get, and look at the man pages for it ... you might get a lot closer to the answer your looking for, than a 4 year old how-to ... [06:46] lol twb ... [06:47] Mgamerz, sorry to bludgeon on with an idea you've discarded, but if money is an issue what about scavenging an old patch lead and wiring in an extension to it? You say you have access to network ports on both the server and the router could it not be wired somehow ? [06:47] or maybe you should figure out why the wireless is dropping out? [06:48] if you have so much time on your hands. You will find dial up networking speeds unbearable. [06:48] Or just call up your local luv and ask them if they have a spare 100baseT card [06:48] no [06:48] i am not using dial up to access the internet [06:49] !!! [06:49] Er s/LUV/LUG/ [06:49] i am using it to access the terminal [06:49] terminal does not require copious amounts of bandwidth [06:49] which 802.x wireless dongle are you using .. a b, g or n type? [06:49] g [06:49] it has always done it to ubuntu [06:49] when it was on this box before the HD died it killed it as well [06:49] whereas you may find many people with the knowledge to help solve your wireless troubles. [no not heaps of bandwidth, but you also said you were using vnc] [06:49] fluvvell: he wants to run a serial line between a server and a serial console, except he wants to somehow do it using two modems instead of a nullmodem serial cable. [06:49] i am using vnc right now over the wireless. [06:50] i use vnc over the wireless, when it is up [06:50] twb, yes - not a wise choice in my book as I've worked with pci modems and frankly nowadays they are a huge pain. [06:50] and when it goes down (which is occasional), i cannot access the box at all [06:51] and [06:51] fluvvell [06:51] Mgamerz, what phone system is between these computers? [06:51] a single cable [06:51] i'll take a picture it'll just be easier to show [06:51] two wires? [06:51] or four ? [06:51] 4 wire phone line [06:51] cat3 [06:52] err [06:52] gotta find my phone... [06:52] brb [06:52] with 'g' you should have a issue maintaining a connection .. unless there is a lot of walls/electronics/ or something else running in the house that is in the same freq range as that dongle ... have you tried to switch channels on it .. and see if you get a clearer signal? [06:52] yep [06:52] this is the best setup i can get on chan6 [06:53] my phone is somewhere near this pc as my monitor has fuzzed up once cause of the interference. [06:53] but where [06:53] and you have no cat5 patch leads that you could wire onto them? A reverse network patch lead only uses four wires. 40' would probably still work ok. [06:54] monitor ? Interference ? from what ? the wifi ??? [06:55] no [06:55] my phone [06:55] i have a CRT [06:55] and when it gets a call (and sometimes not) the monitor will F up [06:56] Mgamerz, my understanding of phone frequencies is that they overlap at the lower end of the wifi spectrum, therefore using channel 11 or higher if possible is desirable. Sounds like the phone is the cause of most of your problems. [06:56] fluvvell: hei is getting freq interference ... from electronics around him ... that limit him to getting best connection on channel 6 .. but he said that somewhere close to where he has his unit, something else is also causing interference with his monitor from time to time as well ... almost sounds like a grounding issue ... [06:56] no [06:56] ffs [06:56] my phone has nothing to do with this [06:56] ie wifi greater than channel 10 [06:57] alright [06:57] so to imageshack i go [06:58] Mgamerz, sry we're just trying to help. Sometimes people can be fixated on a solution which has many repercussions [06:58] but troubleshooting that is not gonna solve his request .. he needs/wants to set up wvdial on lucid ....we shouldn't be guessing him, or trying to talk him out of it ... [06:58] azteech, you're quite right of course. [06:59] yea i know [06:59] Mgamerz, what type of pci modem? [06:59] umm [06:59] lemme pull it up [06:59] i have scanmodem on the vnc window right now [06:59] scanmodem ! [06:59] Conexant Systems HSF [07:00] do you physically know the chipset ? [07:00] HSF... [07:00] just tell us it isn't a wintel modem ... cause if it is .. will be more issues to try to correct .. [07:00] ie have you eyeballed the card - believe me it helps [07:00] of course it is. [07:00] I'm just gonna go ahead and plonk this modem thread, mmmmkay? You're not gonna get anywhere. [07:00] didn't see anything about wintel in the info [07:00] yea its Conexant [07:01] azteech, that was the angle I was trying to come from. [07:01] alright [07:01] so... http://yfrog.com/3dimage1813j [07:02] on the right is my computer that i am talking to you from, this is my main pc [07:02] the monitor is hooked up to it [07:02] on the left bottom is the server [07:02] and you probably can't see it, but there is a telephone line going between them directly [07:02] via pc modems [07:02] pci* [07:03] sort of figured it was ... :) [07:03] i want to have my server act as a dial up server '(like an ISP), NOT DO CONNECTION SHARING, and just use the terminal to do stuff [07:03] that way i have an 'always on terminal' if stuff goes south [07:03] any chance of a high res photo of the actual pci cards ? [07:03] eh... they are already installed [07:03] but [07:03] the one in the server (ubuntu) is a conextant [07:04] when i looked at them earlier today that is about all i could gather from the sticker [07:04] the one in the windows is... well the one that was in the server originally [07:04] didn't think i'd ever use a modem again [07:04] also didn't realize how serial and modems are painfully intertwined [07:06] I have 2 spare ethernet pci cards [07:06] BUT [07:07] i don't have any pci room in my home PC. [07:07] I have my Graphics card which is ginormous (and sucks), a fan for it (cause my pc overheats... its like spilling its guts out) [07:07] and well... that takes up 4 slots O_o [07:07] Mgamerz, the motherboards don't have built in network ports? [07:07] they do. [07:07] But. [07:07] I have this [07:07] uh [07:07] i'll give yo uthe page it'll explain better than me [07:09] i pulled out the cable by accident reading the name [07:09] http://www.trendnet.com/products/proddetail.asp?prod=175_TEW-654TR&cat=77 [07:10] its that thingamagic [07:10] i can never spell ma-gij [07:10] ... [07:10] anyays [07:10] anyways* [07:10] it also acts as a passive client for wifi. [07:10] It's way better than dealing with the setup of windows, plus i can plug it into anything without any driver hassle. [07:11] I would use it on my server... but im greedy and use it for my pc instead. [07:11] WAIT [07:11] hmmmm [07:11] Connexant HSF 56 K Fax/modem .. is that the modem you have? [07:11] yep [07:11] gotta go, bfn. [07:11] i jsut had an idea [07:11] bfn? [07:11] I have TWO extra ethernet pci cards [07:11] fyi .. it is a windows soft modem ... meaning to make it work you will need to find a windows driver that will be compatible for linux ... [07:11] i just realized [07:12] haha i am stupid [07:12] if i can install the pci modem then i can swap it out and instead use the pci ethernet adapters and just do... [07:12] ah damnit! I don't have a crossover cable [07:12] the only one that I know of that is available .. is from linuxant .. and they charge for it .. [07:12] man linux was not made for dial up eh? [07:12] yep ... that would be the way to go ... [07:13] i need a crossover cable... [07:13] not really ... [07:13] i wish there was software based MDIX [07:13] was back in the day when all there was, was dial-up .. but even then it was a pain to set up .. [07:13] i remember when the internet used to be free [07:14] had yahoo internet [07:14] back in 96 [07:14] was the slowest pos ever [07:14] azteech: I still use dial-back modems for remote access to a secure facility. [07:14] I use dial up occasionally to telnet/console into a cisco router. [07:14] Mgamerz: and yet you still live with your parents [07:14] well [07:14] twb .. didn't say they aren't being used ... just said for him .. his situation is almost impossible unless he can find a windows driver that can be installed ... [07:14] im 17 [07:15] will be 18 in a month. [07:15] azteech: oh, I thought we were getting into a retro tech pissing match :-P [07:15] providing one has the right hardware, dial-up can be, and is still used today ... [07:16] I could make a crossover cable [07:16] but i don't have crimpers [07:16] nope, being a old mil comm tech .. I know better ... lol [07:16] or a cable... whateveryoucallit ends [07:17] would recommend cutting and splicing the cable .. but that is a pain .. and unless you have a solder gun, probably not the right solution ... [07:18] i have a coupler so i do in fact have a cable long enough to reach the router itself [07:18] but having a cable run through the house is not really an option [07:18] iwas thinking of using bluetooth serial [07:19] then i was like nahh [07:19] bluetooth on ubuntu is a pain in the rear anyways [07:20] your better off ...going down to the local second hand shop, getting a older 10baseT hub for $5 buck .. and run cable .. at this point ... [07:20] heh [07:20] well [07:20] where i live it probably won't have it [07:20] i don't live in that big of a city [07:20] Mgamerz: you don't need a crossover if at least one NIC supports autonegotiation of same, which is a requirement of the 1000baseT spec. [07:20] yea... i don't have a gigabit. [07:20] only 10/100 [07:20] Many 100baseT card provide it, too. [07:20] i'll look up if it has MDIX [07:21] twb .. true ... [07:21] Admiteddly, not any of the shitty ones that *I* get... :-/ [07:21] ehhh wtf... where'd the card go [07:21] oh [07:21] man my orom is a disaster since i dismantled some old machines [07:21] 2001 D: [07:21] sounds like it is time for a yard sale ... lol ... [07:22] lol [07:22] i just took out a AMD K6 processor [07:22] 233 [07:22] Or just give it to your local hardware recycling group. [07:22] mainly so i could remember my first computer [07:22] which was a pentium that ran on the same socket... [07:22] well [07:22] don't worry .. have the same issue here .. trying to marry two households together .. that are both geek based .... [07:22] They spruce it up, install xubuntu, then sell the machines for like $10 to poor people. [07:22] until i accidentally ran over the processor with this chair here <_< [07:22] was busy screwing with it and i bent some pins [07:23] wow this card i can't find a serial fro [07:23] for* [07:23] hmm [07:23] perhaps the cisco one [07:25] LNE100TX [07:25] v 4.1 [07:25] doesn't seem to have MDIX [07:26] but perhaps if i can find a firmware for it [07:27] good luck in your endeavors ... [07:27] hopefully [07:31] mgamerz .. your out of boise and they don't have second hand computer shops? [07:32] kind of [07:32] i don't know what you mean by second hand relaly [07:32] really* [07:32] I know of some computer shops, but I doubt they'd sell a cheap hub (eww hubs) or a switch for less than $30 [07:33] omg [07:33] i cannot find out who the maker of this card is [07:33] basically one that takes in computers, fixes them, upgrades them, etc ... and the parts they don't use, or older stock they sell at heavily, used parts pricing .. [07:33] yea they have some here [07:33] like... 2 [07:33] i can't find the maker of this card.. the whole thing seems unbranded [07:34] Pulse H1012 google search shows someone with the same query but it's not the card maker [07:34] those would be the ones that would most likely have the older hubs dirt-cheap .. if they still have any left ... [07:34] well i'll try this one first [07:34] also, sometime thrift shops have them as well .. just have to look around ... [07:34] yep [07:34] see if the cared has a pn on it .. and google for the part number ... [07:34] card .. [07:35] theres a [07:35] circuit? or chip idk soder(ed?) (??) onto it [07:35] and, sometimes you can find info about the card by doing a fcc id search ... [07:35] but it comes up with nothing useful [07:35] and theres on other one [07:35] it looks like [07:35] atmel? [07:35] no, the pn would be printed on the circuit board ... [07:36] theres a lot printed on this card [07:36] sometimes though, manufs use stickers/labels .. and if that is not present .. you may have issue id'ing the card .. [07:36] the cisco one is much better [07:36] amtel is the chip maker ... [07:36] there is a sticker but it's just a serial number... not like a id [07:37] nothing that may look like a p/n: or pn: or pn# .. [07:37] i'll punch some into google [07:37] theres a huge amount of numbers on this one [07:38] eh wtf [07:38] it comes up as a intel motherboard [07:38] probably a old intel nic then .... that was coupled with a intel mobo ... [07:38] yea [07:39] its relatively... unbranded [07:39] all that is there is the text that you find on the circuitboard [07:39] all the tiny R1's and stuff [07:39] nas, what you need to find is a pn or mdl # ... [07:40] it is indeed an intel card [07:40] can't type this evening [07:41] see anything that says fccid on it? [07:41] yea [07:41] im tracing it [07:41] comes up as a intel server card [07:41] type that in as fccid and the number [07:41] pila8460b [07:41] well, that is getting you closer ... [07:41] is the id... i think [07:42] [07:42] Intel PRO/100plus Management (PILA8460B) Network Adapter [07:42] probably a 8460b based nic card then ... [07:42] eh... good/bad? [07:42] there you go ... now you have info .. and can search ferther ... [07:43] probably better than the cisco one you have, but also probably older ... [07:43] it is older [07:44] the cisco one is from like [07:44] 2004 [07:44] may or may not have the mdix ... have to research it and see ... [07:44] or i could just plug it in [07:44] seeing as i'm not finiding much info about this [07:44] you could ... would suggest that intel site might have the best info on the card .. [07:45] Intel 82559 Fast Ethernet controller [07:45] its not on their site [07:45] all it shows is a discontinuation [07:46] http://www.latestdriversupdate.com/driver/34635.html [07:46] http://www.intel.com/support/network/sb/cs-006103.htm [07:47] dis a search in google for what you listed the card as ... and came up with lots of reviews .. sales locations .. the the above two links ... [07:47] and am sure there is more .. [07:48] hmmm [07:48] trying to see if nayone has hacked the firmware [07:48] doubtful they could add mdix [07:49] PILA8460B (PBA# 721383-xxx). [07:49] that is my card [07:49] my card is 007 :P [07:49] i've noticed on the end of the ethernet card there is 3 pins [07:49] the same as the cisco one [07:49] there you go ... see, if you look, sometimes ye shall find ... [07:50] look at the intel link ^ ... and am sure there is additional info you can find out about the card, besides the one ^ ... [07:50] now, will leave you to your research ... and escapades/endeavors ... have fun .... [07:51] imma just install it [07:51] lol [07:51] hopefully tonight was a learning experience for you ... [07:51] Hi all, question: everythime, I "shutdown -h now" the next bootup i get the following error "fsck from util-linux-ng 2.12.2" [07:51] i'll put the linksys in this machine and the intel in the unix [07:51] and the screen locks up [07:51] i (L) webmin [07:53] anyone know why i;m getting this fsck error [07:53] ? [07:53] yosi .. sounds like you might have a hd that may be starting to flake out ... or, is not getting enough time to complete write cycle before it is powered off ... and it is corrupting the data somehow ... [07:53] smart test :) [07:53] they are brand new SSD's in a RAID 5 setup.. [07:53] Mgamerz: good luck bud ... [07:53] i just installed them all today [07:54] doesn't mean thay can't be bad out of the box .... [07:54] true... [07:54] is there a way of slowing down the shutdown process.. [07:54] cause when i trype shutdown -h now, its like instant [07:54] azteech [07:54] why do you always have ...'s after your lines if i may ask [07:55] just the way I type is all ... [07:55] and a way to seperate parts sometimes ... [07:55] is there any other ways to shutdown [07:55] ? [07:56] press the power button...? [07:56] on mine it shut down [07:56] well [07:56] after i set it to do that [07:58] whew man my hard drives are arm [07:58] warm [07:58] Yosi: you can always try doing a halt instead of a power down ... option and then power it down manually ... [07:59] how would i do that? [07:59] I just googled my issue, and lots of ppl with SSD's and ext4 have this issue [07:59] sudo shutdown --help and man shutdown should point you in the right direction ... [07:59] some ppl recomend movinback to ext32 filesystem [07:59] typo [07:59] ext2 [08:00] lot of folks are having isses with ext4 as a whole .. not just with the new ssd's ... [08:00] fair enough [08:00] ext4 IMHO .. is still too new to trust my data to it ... so I stick with ext3 .. [08:02] Mgamerz: try running systems with temps up in the 105+ range .. and see how well they do heat wise .. when they aren't in a heavy a/c'd data center ... [08:03] oo fun [08:03] what i don't get [08:03] is how do pc manufactuers ever think you can clean the heat sink in a pc without taking the heatsink out [08:03] I have 120 fans in my cases ... plus a/c going, and a fan blowing on the case .. and they still run 40-50C ... [08:04] lol i have a fan on my case too!! [08:04] well [08:04] i did [08:04] like 3 days ago [08:04] until i bought a fan cable splitter [08:04] now i have like [08:04] 5 fans in my pc case [08:04] so [08:04] im going to log into here with my tablet [08:04] Mgamerz, canned air [08:04] they expect you to take it out, clean it once in a while ... and re-apply temp cream between the plate of the HS and the cpu ... [08:04] bad for the environment [08:05] I even do it powered on sometimes [08:05] hi to all :-) [08:05] but I guess I just live dangerously ;) [08:05] yea... i didn't use any thermal compound [08:05] so it probably overheats cause of that lol.. [08:05] wait a minute [08:05] OH DANGIT [08:05] i installed the wrong card [08:05] wait [08:05] no i didn [08:05] t [08:06] this boot rom chip area of the linksys card has this huge spaced contacts that makes it appear super old [08:06] alright [08:06] unless you buy coolers with the silver compound on it already in place, you should ALWAYS use thermal compound ... [08:06] meh [08:06] quote 'i live dangerously' [08:06] no not really i awnt a new pc and if it died i'd have reason to get a better one than this hunk of junk [08:06] that was fairly sarcastic [08:07] I've worked on computers for years, never once used ESD straps [08:07] and I work on systems live somewhat frequently [08:07] ello! [08:07] lol gman, we were talking aobut static issues ... [08:07] die mgamerz [08:07] As long as you're holding the case and aren't a frail old woman, it's TOTALLY SAFE [08:07] s/were/weren't ... [08:07] never fried anything [08:08] just get some canned air, aim the straw at the heatsink, and blow [08:08] cleans it right up, never even have to shut down if you don't want to [08:08] the issue was to use thermal paste or not ... between the cooler and the cpu ... [08:08] it's compressed CO2, doesn't conduct electricity [08:08] you should always use a thermal compound [08:08] and how to properly clean it ... and your right canned air can be used ... but some folks don't like using it ... [08:08] i have 2 power supplies... [08:08] not using one voids warranties [08:09] and fries chips [08:09] does that prove my pc's terribleness [08:09] my graphics card is sooooooo power hungry [08:09] a 2004 server graphics card at a gis workstation [08:09] I would never remove a heatsink from a chip to clean it [08:09] that's just silly [08:10] if anything, I'd shut down and use an air compressor [08:10] well to blow out the heatsink i had to pull out the thing. [08:10] or be careful with a vacuum [08:10] if you could access the level and keep both toether that'd be great [08:10] actually, removing the heatsink from a cpu isn't silly ... did it several times ... and today was the worst ... dust caked so thick under the base and around the cpu ... that had to be scraped loose ... [08:11] yes, it is [08:11] Or just replace your server every five years [08:11] the only time you should ever remove it is if you want to use a new heatsink or compound, or replace the chip [08:12] it's somewhat risky and each time you do it you have to clean the chip with alcohol and use more compound [08:12] removed the sink to do through cleaning .. and re-apply new compound ... hadn't been done in 5 years ... [08:13] unless you used really, really cheap compound, you don't have to redo it at all [08:13] qman ... it was factory ... [08:13] factory is designed to last for decades [08:14] I would agree, but in this case, the dust was so bad of a build up .. that getting to it .. meant also getting the old off ... and put on new .. [08:15] chip was getting to the point of total burn-up .. and unless something was done it was gonna be gone ... [08:15] nothing a blast of air won't clear [08:16] the place where the dust was really bad was at the base up under it .. and around the chip itself ... had to remove the sink to get it out ... therefore breaking the seal on the paste ... === lifeless_ is now known as lifeless [08:17] and it was onle of the older style P4 fan systems that are both hooked and pressure ... [08:17] so couldn't just blow it out .. [08:17] don't have to get that out, as long as the fins are clear and getting airflow, it'll work [08:18] been done several times .. and temps kept rising ... so, today ... took it totally apart and cleaned it good ... now temps are back to decent range ... [08:21] okay ... let me get off here ... will see you all later ..night ... [08:24] hhh [08:24] this is pissing me off [08:24] where'd the screw go === cemc1 is now known as cemc [09:01] New bug: #591128 in lm-sensors (universe) "sensord cannot RRD log with multiple chips" [Undecided,New] https://launchpad.net/bugs/591128 === wieshka_ is now known as wieshka === funkyHat_ is now known as funkyHat === unreal_ is now known as unreal [09:10] Whuh? No ttx? [09:31] twb: hi [09:31] Thanks and appreciate it [09:31] I could resolve the ipp2p match module on iptables [09:32] kaushal: unless you want something specific, please don't greet me by name. [09:39] twb: ok === viezerd- is now known as vizerd === vizerd is now known as viezerd [09:51] * RoyK greets twb by his name [09:52] RoyK: plonk [09:53] * RoyK has just decided to finally give up the attempt to use opensolaris for a host for virtual machines [09:53] * RoyK gets out a CD to burn ubuntu lucid server [10:02] Are there good books to learn about ubuntu-server ? [10:04] The Ubuntu Server Guide [10:05] is it https://help.ubuntu.com/10.04/serverguide/C/index.html ? [10:05] yes [10:06] at hospital, wife having baby [10:06] with a laptop, chatting_ [10:06] ? [10:07] ccheney: Awesome! [10:08] ccheney: If it's suitable, are you avaliable to upload your euca lucid sru.. or would you rather someone else sponsors it? [10:08] hmm.. it must be silly o'clock in the morning for ccheney? [10:12] yea no sleep for me irc from iphone [10:12] this is one app the extra resoluyion would be helpful [10:13] ccheney: debuild from the iphone.. there MUST be an app for that :P [10:13] ccheney: Okay, is it ok to get your current branch sponsored? [10:14] ccheney: I'm guessing you are going to be afky for a while from today. [10:15] yea that is fine the prob i was having yesterday appears to e bad error hNdling in euca [10:16] :/ [10:16] i filed a bug with most the findings [10:16] I read! Good work [10:18] thanks i found a spot where to fix it i think but i dont know java well, so i [10:19] might have misread the code [10:19] ttyl === register is now known as Guest70139 [10:37] hi there: does anybody knows why eth0.10 cointinues going UP at networking reboot [http://paste.ubuntu.com/446554/ < /etc/network/interfaces]? [10:41] e-DIO-t: you probably have NetworkManager installed. [10:42] It's designed to royally screw up any network you might have. [10:42] guess nope: its a server install [10:42] btw, on server reboot no more "cached" vlan interface... [10:42] e-DIO-t: is this a VM? [10:42] yep [10:42] Vbox [10:43] so guess, it's a known-one is it? [10:43] Does vbox a persistent MAC across reboots? [10:43] udev will assign a different interface name if the MAC changes. [10:44] ^_^ tnx [10:44] e-DIO-t: Error: "_^" is not a valid command. [11:50] Oh, ew. [11:50] So for the first time, I happened to boot 10.04 on a machine that actually had a GPU supported by this new fedora splash crap [11:50] It's all... magenta [11:51] Come on, man, I only just got used to 8.04 and you change the splash and GTK theme and colour pallette again? [11:54] 8.04 was years ago. [11:55] My users are prisoners, so I don't bother to give them shiny/broken non-LTS releases. [11:56] I meant years to get used to. [11:57] You know, Knuth only upgrades *his* system once a decade [11:57] So compared to that a two-year window is pretty small. [11:59] That and most of my gear doesn't run X, so it's not like it's in my face every day [12:04] I have this laptop with wireless connection and I've installed LAMP on it. Now I would like for web server to be able to get online through the wireless without me having to login. Is this possible if I select "Available for all users" in the settings for the wireless network in the network manager? [12:04] Or is there another way? [12:05] X-Sleepy-X: sorry, I don't support hosts running Network Manager [12:06] :) [12:06] If it wasn't there, the normal way would be to configure /etc/network/interfaces with a plain roaming wpa-supplicant configuration, but that'll break your normal NM usage [12:06] isn't this ubuntu-server ? [12:06] twb: ok [12:07] Erm, I'm assuming your laptop is a wifi client, not an AP [12:07] yea [12:07] it's just annoying having to login in order for apache to go online [12:11] The good news is you can probably get support for NM in another channel. [12:11] The bad news is that channel is probably #ubuntu [12:13] twb: ok thx [12:13] :) [12:21] heh [12:29] twb: I've tested around a bit and the solution were to select "Available to all users". [12:29] twb: So I though I would share that information... ;) [12:29] X-Sleepy-X: and did that work? [12:30] twb: Yes. [12:30] Oh, I assumed it didn't, since you mentioned that early on [12:30] twb: I just didn't know a way of testing. [12:31] twb: But I then realized that I could try to connect with SSH by my cell phone. [12:31] :) [12:31] Anyhow, I'm off for yet another restart.... [12:46] morning === oubiwann is now known as oubiwann_ === oubiwann_ is now known as ouniwann [13:00] ubuntu-desktop's synaptic gui has a function to 'generate package download script'; is there something similar available in a terminal? [13:02] while trying to install lucid server on a xSeries 225 i get kernel panic... any suggest? [13:03] did you verify the iso after you downloaded and burned it? [13:04] panfist: jigdo? [13:04] panfist: apt-walkabout? [13:05] apt-walkabout sounds about right [13:08] Personally I found it a massive pain in the arse [13:08] It was easier to just post weekly rsyncdiff files of the mirror. [13:09] what's your use case like...i'm trying to configure a server that's on the private side of an airgap [13:09] I have secure facilities that aren't allowed to be networked [13:10] i think that means the same thing [13:10] I need to apply security updates to the servers and hundreds of workstations behind them [13:11] If there's only one host, it might simply be easier to do [13:11] apt-get install foo [13:11] ...then collect the failures into a list of URLs, wget them, put them in /var/cache/apt/archives, then re-issue the command. [13:12] but isn't that only going to show you one level of dependencies? for example in my case, sudo apt-get install mediawiki depends on php5 and php5-cli...but what do those depend on? [13:13] oh i guess it does show all dependencies... [13:13] Assuming you've got an accurate package listing on the offline side, it'll attempt to fetch all deps [13:13] It boils down to: is this a one-off, or a regular thing? [13:16] i don't think it will happen too often, and if it does i'll probably just do a private repository (i think you suggested that before) [13:21] maybe i'm being picky and i should just do it, but i'm trying to find a command that outputs the full url of a package, instead of the something like this "foo (1.1.1-ubuntu4.2 Ubuntu:10.04/lucid-updates)" [13:22] something like http://us.archive.ubuntu.com/.../foo_1.1.1-ubuntu4.2.deb would be nicer [13:22] got it [13:23] nope nevermind [13:34] panfist/twb: apt-get has an option --print-uris that will give you a list of download-uris instead of downloading & installing packages [13:34] I was just trying to find that one :-) === sanderj_ is now known as Snadder [13:35] it includes file sizes & hashes too, so should be easy to write a nice script around it (or one probably exists already) [13:36] i was trying that and it wasn't working...it didn't work like this: '--print-uris install foo'; but it worked like this: '--print-uris --yes install foo' [13:36] i guess it needed a little extra encouragement? [13:36] you can run it as a normal user [13:37] morning [13:40] panfist, try: apt-get --print-uris -qq install foo [13:40] then you only get the lines with the URLs [13:51] hggdh, headed off to bed, was up all night due to delivery, it appears that eucalyptus does not catch errors on creating volumes, at least if i understand the code correctly [13:52] hggdh, or rather it might catch it but does not do so correctly and does not appear to clean up afterwards [13:52] hggdh, getting to actually retry on transient locking issue seems rather hard due to the fact that lvm sucks [13:53] hggdh, it appears lvm always returns error code 5 for any kind of error, so you can't actually check the return code to do something sane [13:54] hggdh, i'm on paternity leave until june 18, but will probably be around and will also be available via email if needed for anything [13:54] Daviey, see above comments to hggdh also [13:55] if anyone has a bored moment and experience with half-failed (?) builds on launchpad, I'm trying to figure out the deal with http://launchpadlibrarian.net/49925137/buildlog_ubuntu-lucid-amd64.atlas_3.6.0-24ubuntu1%2Bregressiontest20100608_FAILEDTOBUILD.txt.gz [13:55] it claims debian/control failed, but I see nofailure in the log, and it did get uploaded to the ppa [13:55] (and a local build succeeded) [13:59] ...after all the trouble, us.archive.ubuntu.com is down. [14:08] I'm trying to setup shared storage for KVM and am having issues finding good documentation on how to do so. Is the correct approach LVM used for guest hard drives over NFS or iSCSI? [14:14] diago, i guess you would probably want a SAN [14:14] rather than NFS [14:14] iSCSI would do that [14:14] so LVM should work [14:14] or ATAOE [14:14] with rather than or [14:14] Yeah, I can't see how NFS would work in this situation [14:14] or was in relation to iscsiu [14:15] hmmm not familiar with ATAOE I will look into it. Thanks [14:17] I think I'll stick to iSCSI though it appears ATAOE is faster. It seems more standardized. [14:23] hello, does anyone have any guide that was tested on how to setup suphp or suexec in ubuntu? [14:24] So I get the error message that apache could not determine the fqdn when it starts and I've used to solve this by entering something in my /etc/hosts file but I think I read something somewhere that suggested another solution so now I'm confused. What is the correct way to solve this error message? [14:24] X-Sleepy-X: http://www.howtoforge.com/perfect-server-ubuntu-10.04-lucid-lynx-ispconfig-2-p3 [14:25] do the 4 lines starting with echo server1.example.com > /etc/hostname [14:25] actually.. follow pretty much everything on that page that's related to network config and hostnames [14:26] amstan: thanks :) === register is now known as Guest34581 [14:54] anybody noticing slow ubuntu package repository servers today? [14:55] Yes. You're not the first to report such an issue. [14:55] any recommended mirrors? [14:57] hey guys... [14:57] got a quick question about virtual box and ubuntu server [14:58] kpettit: closest to you [14:58] :) thanks. Seeing what I can find online. I'm in USA, Houston [14:59] T3CHKOMMIE: what's the question? [14:59] kpettit: Check https://launchpad.net/ubuntu/+archivemirrors [14:59] i have nightmares of my server crashing.... [15:00] thanks Pici [15:00] my hardware doesnt support hypervysors.. but... could i run server from VB and back up the vdi? [15:00] would that be a decent practice? [15:03] anyone? [15:04] T3CHKOMMIE: i'm not sure what you're trying to accomplish. are you teying to run vb on ubuntu server? [15:04] T3CHKOMMIE: you can't back up the vdi while running [15:04] no, [15:04] if you use LVM you could take snapshots [15:04] what is "VB" [15:04] virtual box [15:04] so im trying to get accurate fast recoverys... we have an OLD server [15:04] oh, with virtualbox I do snapshots nightly [15:05] snapshots are nice, they are very quick [15:05] let say i run ubuntu on my servers hard ware... [15:05] then I do full-backup one weekend [15:05] then i run ubuntu server from virtualbox, and back up the HOLE vdi [15:05] kpettit: are VB snapshots the entire system? [15:05] I thought that was only for recoverable errors [15:05] snapshots are whatever is changed. kind of like a diff [15:06] then... system crashes... reinstall ubuntu.. install VB... upload server .vdi [15:06] any ideas? [15:07] diago, I use snapshots alot for windows dev stuff. I set a snapshot, load lots of crap software and go back to snapshot to get prestine system [15:07] as far as i understand ubuntu VB and all that, i think it would work.. but im still new at this :S [15:07] I do snapshots before I load anything basically. Works nice to go back quickly [15:07] kpettit: right to go back, but I think he wants full backups of the system [15:08] diago, thats correct.. like copies of the Virtual Disk Image. [15:08] I'm not sure the best way to do full backup's. The best way I've found is to pause then copy VB setup [15:08] kinda like making a swappable hdd backup? i guess? [15:08] T3CHKOMMIE: the only way to do it live is LVM snapshots [15:08] for what I understand [15:09] I know trying to copy a running system doesn't work to well. But you can do snapshots to get a live system or pause it to do the full backup's [15:09] is that possible whi [15:09] gm. LVM im trying to learn that stuff... [15:09] Have a SAN or filesystem take care of backing up the files is the only other option I can think of to get a good copy of a running system [15:10] is LVM kinda like volume shadow copy? [15:11] http://en.wikipedia.org/wiki/Logical_Volume_Manager_(Linux) [15:12] I haven't done much with LVM stuff myself though, always drove me nuts [15:15] im trying to do a whole system backup... so in the event of a catastrophic failure, i can simply upload a file.. boom, everything back where its supposed to be, and everything still configured. [15:15] any ideas? [15:15] i dont mind taking the system down for an hour or two on the weekend and doing a back up. [15:16] if the system you are wanting to backup is a VirtualBox server that's pretty easy then [15:16] Just pause it, copy the files to where you want them to go. That's it. [15:16] and thats going to be the entire system? [15:16] right? [15:17] The entire VirtualBox image. Just make sure you get all it's files. Each server can have multiple disks. [15:17] Usually the config files are in $HOME/.VirtualBox and those files will show the disk images it's using and their locations [15:18] For me, I do snapshots nightly, then do a full-backup using rsync on the weekend's [15:19] and its pretty reliable? our server here gives me nightmares. [15:20] it has been for me. I moved to VirtualBox becuase I couldn't get VMWare to work consistentaly. VirtualBox has been very good for me. Snapshots have been awsome. [15:20] sweet [15:20] and what host are you running it from? [15:21] My big problem child one is a Win2003 server that I used vmware tools to convert from a real server to a vmware image. But vmware itself couldn't run it for more than a couple hours at a time without crashing. [15:21] But virtualbox worked great with it. [15:21] it's been running in production now for 3-4 months now. no problems. [15:22] ah... that sucks. im thinking of running vb on my 2003 server.. then slowly making everything a vdi and running it virtually from the same hardware... [15:22] so good to hear kpettit. thanks! [15:22] I haven't used VB server on windows so much, but on linux it's great [15:22] we run KVM and VB here. VB for winblows KVM everything else [15:23] VB on windows is weird but it seems to work fine [15:23] Windows's guest OS"s work great. I use Win7, WinXP, and WIn2003 quit a bit. Like I mentioned I converted a real running win2003 server to a guest os that now runs inside of Virtualbox [15:23] how easy is kvl to implement and deploy? [15:24] kpettit, how did you convert a real imaged to a vdi? [15:24] KVM? You need virt technology on your CPU to use KVM [15:25] there was a vmware free tool to do it. [15:26] ah gotchya. [15:26] http://www.vmware.com/products/converter/ is what I used. [15:26] funny thing was is the image it created didn't work reliably on vmware server, but did in virtualbox. [15:27] kpettit, thanks! ive messed things up accidentaly too any times, and i really need to learn snapshot stuff and recovering... doing a full reimage after i chmod the wrong folder is really starting to get old :S === alex_jon1 is now known as alex_joni [15:28] for learning it's I'd recommend doing a image, copying it around. Doing snapshots and such just to understand it all. [15:28] kpettit, thanks, im setting up my virtual server right now. [15:28] Takes a bit to learn the way they do things. Ohhhh I realy love there cli remote desktop though. Lets you RDP into any system you run [15:28] I've been running VirtualBox Win2008 server guest (CentOS host) for about a year, its been really stable [15:29] aurigus, thats awsome! how do you back it up? [15:30] backups! what backups! [15:30] this is just an experimental server so I dont have backups [15:30] ah ok [15:30] but I imagine I would do a snapshot daily [15:30] I have set up a local server but i want to create ssh users like alket@myip.com how to do that ? [15:30] although it would use a lot of space and generate a lot of transfer [15:30] VBoxHeadless is great though [15:30] alket, https://help.ubuntu.com/community/AddUsersHowto [15:30] vboxheadless? [15:31] Its a centos box w/o gui, so you use that to control the VM [15:31] thank you kpettit [15:31] T3CHKOMMIE, vboxheadless is it. You can start a image from the cli, the RDP into it. [15:31] very nice, lets you see if there are any bootup problems. [15:31] * T3CHKOMMIE is confused now. [15:31] yes its nice to have rdp for emergencies [15:32] Normally on a remote windowsw machine you can only RDP into it if it's up and running correctly. Network is setup right, etc. [15:32] i love CLI, and thats why ive been sticking with ubuntu so much. but how do you opperate vb in cli? [15:33] with virtualboxheadless it basically wraps RDP arround it. So you can RDP into the booting up image and see it bootup. So even if there were network issues with your image, or some boot issue (checkdisk) or whateve ryou can see what's going on [15:33] yes, you are basically RDPing into the 'server' and not the 'OS' [15:33] AH... like when its taking 50+min to checkdsk and you think you killed your server [15:34] that is the method you can use to do the install too, since you don't have an OS yet [15:34] aurigus, ah, i see rdping into the "server" not the os.. so will it show the slash screen as well? [15:34] yup [15:34] and im assuming that this is benificial cus you dont have to run the vb gui to see it all happen? [15:35] right, so you save on overhead of running xwindows [15:35] also in theory you could script everything out and make automated deployment [15:35] awsome. [15:35] but if you are just getting into it, just experiment with it to see how you like it [15:35] T3CHKOMMIE, yes. And if you are doing this remotly it still allows you to see the guest OS gui, even if network is hosed or other bootup problems exist on the guest [15:36] man, i gota learn more of this. you know of any good wikis or tutorials on that stuff? [15:36] trying to find the cli referance for ya [15:36] the VirtualBox website has a nice reference on it [15:36] is it possible to ssh -x11 with vbox headless? [15:36] or are those two different beasts? [15:36] http://www.ubuntugeek.com/how-to-control-virtual-machines-virtualbox-using-vboxmanage.html [15:37] They are different, but you can ssh -X then start VirtualBox to see the normal GUI [15:37] thanks kpettit [15:37] awsome [15:37] Here is the 'manual' http://www.virtualbox.org/manual/ch07.html [15:38] awesome thanks so much for the help! [15:38] gotta love the opensource community! [15:38] well, looks like ive got some reading to do. thank again, ill probably be back on here soon! [15:38] im looking for Ubuntu's equivalent of the failure function found in Redhat's /etc/init.d/functions anyone know where it could be hiding ? : ) [15:39] hi [15:39] how can i check if my ethernet is gigabit? [15:40] with lspci [15:41] New bug: #591286 in php5 (main) "PHP Deprecated: Comments starting with '#' are deprecated in[...] on line 1 in Unknown on line 0" [Undecided,New] https://launchpad.net/bugs/591286 [15:44] spineau: ping [15:48] !webmin [15:48] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. See !ebox instead. [15:48] * kpettit likes webmin :( [15:50] mathiaz: ping [15:50] hggdh: pong === ouniwann is now known as oubiwann [15:51] looks like I found my answer! log_failure_msg : ) [15:51] spineau: how are things going now? [15:52] hggdh: good since I'm now able to launch instances [15:52] nxvl: hi - o/ [15:52] spineau: :-) how did end the issue you were discussing with ttx? [15:52] mathiaz: does puppet has a web interface? it doesn't right? [15:53] nxvl: look up puppet dashboard [15:53] nxvl: or the foreman [15:53] nxvl: http://theforeman.org/ [15:53] hggdh: the nc log were plenty of errors, one mentioned kvm but vt extensions were enabled [15:54] hggdh: and finally it was not this error msg that breaks the starting process of my instances [15:54] mathiaz: great, thanks! [15:54] hggdh: my eucalyptus.conf (on the NC) was corruptus each time I restarted eucalyptus [15:55] spineau: ?? how? Did you find out? [15:55] hggdh: I found a sed command in this file acting on the value VNET_BRIDGE [15:56] hggdh: I cleaned the file and now its working (again) [15:56] spineau: weird. Let me check mine [15:57] spineau: in /etc/eucalyptus/eucalyptus.conf? [15:57] hggdh: yes on the NC [15:57] hggdh: I still wonder if it's a process or a user that was doing this [15:58] spineau: weird. I do not have it, I do not think this is created by install [15:58] spineau: euca logs a *lot* of error messages, most do not impact us [15:59] spineau: *why* log an useless error, I do not know [15:59] ttx: ping have you uploaded something for the krb5 sru? [16:00] zul: yes [16:00] ttx: has it been accepted yet? [16:03] I don't think so, lemme check [16:03] Fix uploaded to lucid-proposed, ACKed from ubuntu-sru [16:03] k [16:03] not AA-processed yet [16:03] * hggdh gntly nag zul re. the samba apport hook [16:04] hggdh: yep its on my radar for today [16:04] roj [16:08] hggdh: I'm trying to get camucamu up but it seems (according to IS) that eucalyptus is blocking eth0 at startup and finally it is not pingable [16:09] hggdh: I have no idea what because I'm not familiar with eucalyptus. There's a bridge configured in interfaces and something is firing ifconfig to bring up a firewall, but I can't find what's firing ifconfig. [16:11] spineau: do you have access via a serial console? [16:11] hggdh: via kvm [16:12] spineau: is camucamu another box in the test rig (I did not know about it)? [16:14] hggdh: yes, normally it's another nc, according to tasksel menu [16:15] spineau: what is the IP address? I am adding it in my .ssh/config [16:16] hggdh: 10.189.84.154 [16:19] hi [16:21] hi i know i keep asking, but would people suggest their favourite local ubuntu VPS provider [16:23] spineau: give me 5, have to reboot [16:24] hggdh: oups [16:39] I configured an iSCSI target and have a client that sees the drive as /dev/sde . I want my KVM machine to use those as the drives. Is it safe to use them as /dev/sde or should I be using UUIDs? [16:44] diago: you should use iSCSI the way you use any block device IMO [16:44] I just read "Paths like /dev/sda, /dev/sdb, etc are not a good choice because they are not stable across reboots, or across machines in a cluste" [16:46] diago: if it makes you feel better, i use UUIDs for iSCSI devices. never had a problem [16:47] pmatulis: ok, I think I may be going about this wrong anyway. Ultimately it's for KVM but it looks like KVM can manage iSCSI itself [16:48] diago: iSCSI targets are exposed locally as block devices, just like your internal drives [16:48] Right, but how would I configure a KVM .xml file to use a UUID? [16:49] diago: do you use UUIDs in other xml files? [16:50] pmatulis: currently I am using qcow2 images and that is all I have ever used [16:52] it looks like (from reading anyway) I just use /dev/disk/by-path [16:55] diago: http://www.libvirt.org/formatstorage.html [16:55] can anybody recommend a good FTP server that will let me specify virtual users in a text file? [16:56] oh cool, thanks pmatulis [17:04] hi [17:05] just curious to know about searching on ubuntu server mailings lists archives [17:05] how do i search any topic ? [17:05] https://lists.ubuntu.com/archives/ubuntu-server/ [17:12] is it possible to download from a web browser to a folder on a fileserver? [17:13] we have landed a man on the moon [17:13] so I could see that being possible [17:14] unless you think that's a conspiracy... [17:14] can someone comment on https://lists.ubuntu.com/archives/ubuntu-server/2010-June/004297.html ? [17:14] the flag wasn't waving! [17:14] sorry, it *was* === linxeh_ is now known as linxeh [17:14] woh [17:18] i think the conspiracy of it being a conspiracy is a conspiracy... [17:18] brb black hellicopter flying by [17:19] black helicopter??? another conspiracy thing? [17:19] lol yeah [17:20] is there any CA management software in the repos? eg webbased that can issue certs from a CA etc ? [17:21] so how could i go about mounting my fileserver so i can see it when browsing for a download location? It shows it as mounted in my window sidebar but i can not find the folder when i try to browse to it [17:21] i am also trying to mark it as a shared folder for a vm OS [17:22] linxeh: xca is an option I use sometime - it's not webased though [17:22] hmm [17:23] I need something that is relatively easy to use for a non-unix person to be able to issue new server and user certs, and probably remotely too [17:23] eBox has a CA module, but I dont want to install that if I can help it [17:29] checking in again for my query ? [17:43] I have a clean install of 10.04 and when I use aptitude I get a segmentation faulty tree. I've searched the forums but the suggestion to clean the cache does not work [17:50] ZenMasta: do you get it with apt-get aswell? [17:51] there was a lot of aptitude segfaults during beta [17:52] Sleepy it does not give the segmentation fault but gives a similar message about dpkg was interrupted [18:15] checking in again for my query ? [18:20] hi guys, i want to install a http proxy... what software can i use? [18:20] *should [18:23] alex88: Squid? [18:24] jpds: isn't it a socks server? :/ [18:24] or can i set as http server? [18:25] i'm talking about the firefox proxy settings === AdamDV is now known as Guest6490 [18:44] hggdh: ping i did the review and it looks ok. I just had to fix up the changelog just a bit [18:48] zul: do you have a list somewhere of everything you're making nightlies of? [18:49] jcastro: the daily-vcs-builds for spec for lucid [18:49] ta [18:49] jcastro: but right now they are disabled because im switching them over to launchpad eventually [18:50] right [18:50] I am just putting them all on a list [18:53] zul: thank you. What should I do now? [18:53] zul: and you name all the ppa's "edgers" right? [18:53] ccheney: ping [18:53] jcastro: correct [18:53] hggdh: nothing...im uploading as we speak [18:53] zul: thank you [18:54] ccheney: by default the maximum space one can allocate on euca is 50G. You can change it on the web interface [18:55] zul: btw, should I prepare also a Lucid branch? [18:55] hggdh: do you think its worth while to do an sru for it? [18:56] zul: I am not sure. I think it is possible a lot of people have not upgraded yet, and this would mostly target new upgraders [18:57] hggdh: bring it up at the SRU part of the meeting today [18:59] zul: ack [19:01] hggdh: yep i found that out, the problem i saw (i think) was that createLogicalVolume can throw but it is not try'd [19:01] hggdh: and if it does fail euca does not unwind the changes (i believe that is proper term) in any case it does not clean up after itself [19:01] ccheney: so the plot thickens :-( [19:01] hggdh, regarding "butterfly", i thought your ~o~ looked like : http://www.zimbio.com/pictures/fDpxr74q8-g/Olympic+Team+Trials+Swimming+Day+3 [19:02] Daviey, what is "doesn't work" [19:02] hggdh: i think createLogicalVolume probably needs to have try catch blocks inside its own function and then another try catch in the function calling it as well [19:02] smoser: it is actually freestyle, you can see my left arm in front [19:02] is this just the kvm kernel regression from lucid of last week ? [19:02] hggdh: because each call inside createLogicalVolume can also throw [19:03] ccheney: ack. I wonder why the developers did not do it -- it should be quite obvious errors could happen [19:04] smoser: kernel bug with a Java library [19:04] Image doesn't get transfered to the NC [19:04] hggdh: yea, also on top of all of this mess is the fact that lvm executables return very poor error codes, it seems pretty much "5" or nothing [19:04] hggdh: and it returns a 5 for transient issues as well [19:05] hggdh: redhat is working on liblvm to resolve some of that mess i believe [19:05] hggdh: which is in lvm2 source and packaged by debian in the current version in sid [19:05] ccheney: yes, sa your posts earlier today [19:05] hggdh: http://pastebin.ubuntu.com/446758/ [19:05] Daviey, so this is a kernel bug ? [19:05] smoser: yah [19:05] hggdh: each of those functions can throw executionException but it is not in try blocks [19:06] what bug number ? i'm just curious now [19:06] java and kernel bug [19:06] smoser: changing pc, one mo [19:06] hggdh: and if i understand the caller function correctly it only cleans up after itself if the next function after createLogicalVolume fails, but i believe if this one does it bombs out and doesn't clean up [19:06] ccheney: this is really bad, each step depends on the previous, and any can fail [19:07] smoser: bug #588861 [19:07] Launchpad bug 588861 in linux "Instances block in pending state, and don't start" [High,Triaged] https://launchpad.net/bugs/588861 [19:07] Impossible to start ANY instance [19:07] hggdh: i'm a bit rusty on java and exception handling in general but i think i remember how it works :) [19:07] so UEC is quite, er, wet in maverick atm [19:07] ccheney: if you are rusty you are better than I am... hated the whole java thingie ;-) [19:08] hggdh: iirc a function which throws immediately interrupts the execution of the rest of the code after it and it goes to the nearest catch block it is inside (more or less) [19:08] hggdh: if that is correct then yea they don't handle this issue at all [19:08] smoser: Following up on the kernel teams triaging status later today [19:09] hggdh: it does catch executionException but in a useless way for the bug we are seeing [19:09] ccheney: this is how I remember it. But, still, we are left with pieces [19:09] hggdh: uploaded thanks [19:10] hggdh: this is one of the catch blocks it is in: http://pastebin.ubuntu.com/446761/ [19:10] hggdh: if it failed at exportVolume it would be cleaned up, but they forget to do anything about it if createLogicalVolume fails [19:11] hggdh: what is somewhat odd is the error message I get doesn't match up with that string at the bottom, it might get overwrite before it goes to the log file though, i'm not sure [19:12] ccheney: two different exceptions [19:13] hggdh: what i actually see in the log is: [19:13] 22:29:24 ERROR [BlockStorage:pool-10-thread-1] com.eucalyptus.util.EucalyptusCloudException: Unable to create volume group vg-ZR5Pvg.. for /dev/loop10 [19:13] ccheney: probably because of the throw at the end of the block? [19:14] hggdh: that is the proper exception message from createLogicalVolume but not sure why the "Unable to run command:" is not in the log [19:14] hggdh: yea [19:14] I get a Segmentation faulty tree 50% when I use aptitude/apt-get (new install of 10.04) [19:15] ccheney: I think ExecutionException would only be thrown if the command failed to execv [19:16] hggdh: maybe so i haven't looked closely enough to determine how its getting caught [19:17] hggdh: actually you are probably right that it isn't going to that catch block, BlockStorage has a generic catch block after running the code that would log it [19:17] hggdh: http://pastebin.ubuntu.com/446764/ [19:18] ccheney: yeah, and this is a generic catch [19:18] hggdh: if no one tries to fix it before me i may work on it when bored on my leave :) [19:19] ccheney: you *are* on leave, right? [19:19] hggdh: yea i think the generic catch would probably cause it to show up right in the log [19:19] hggdh: yea [19:19] at the moment i'm just sitting at home after being woke up by a severe storm in the area [19:19] * ccheney will be back up to the hospital in a few hours [19:22] * ccheney thinks he may try right now, nothing else to do, heh [19:22] ccheney: OK. Take care of the family :-) [19:24] what is the proper way to update a branch that has been integrated, eg lucid-proposed vs my lucid-sru branch for euca [19:24] should i bzr pull --overwrite or bzr merge? [19:25] it looks like overwrite might be the proper way since its not listed as a merge on lucid-proposed [19:27] * ccheney just did an overwrite, probably not a big issue [19:29] hmm no that is not the right answer, heh [19:31] SpamapS: what's a bug number you're asking for sponsorship (one of your sync/merge)? [19:32] * ccheney found the right way :) [19:32] bug #589566 [19:32] Launchpad bug 589566 in mongodb "Please merge mongodb 1.4.3-1 (universe) from Debian unstable (main)" [Undecided,Confirmed] https://launchpad.net/bugs/589566 [19:33] also bug#586613 [19:38] sorry if this is an faq, but is there a gui to view KVM-based VMs? [19:39] RoyK: Sure. [19:39] RoyK: virt-manager or virt-viewer. [19:39] ah, virt-manager [19:40] forgot about that [19:40] Or virt-viewer. [19:40] just reinstalled my box from opensolaris to something useful (lucid) [19:40] bloody vbox was crashing osol so often the people hosting my server got tired of restarting it [19:41] so moving to something real open now [19:43] replaced a pair of 320gig drives with two 1,5TB drives, connected the 3ware card, installed Lucid and testing now - this better be a wee better than osol and vbox (and xen, which is horrible to configure for anything but hvm, which sucks I/O-wise) [19:44] * ccheney thinks he is missing part of the puzzle on this euca bug [19:46] mathiaz, ScottK, I was following the -backports (I think). I opened bug 589214 and bug 589211 . Do either of you see anthing that i've done wrong, or know what to expect next ? I think i've done as much of https://help.ubuntu.com/community/UbuntuBackports#Backport Process as I really could. [19:46] Launchpad bug 589214 in lucid-backports "backport maverick ec2-ami-tools to lucid" [Undecided,New] https://launchpad.net/bugs/589214 [19:46] Launchpad bug 589211 in karmic-backports "backport maverick ec2-ami-tools to karmic" [Undecided,New] https://launchpad.net/bugs/589211 [19:52] I have the nagging feeling i already asked, but here it is again: in UEC, to register an SC, you need to run "euca_conf --register-nc ..." on the CLC. This mean the CLC has to be able to reach the SC, right? Collateral: the SC, unlike the NC cannot be on a private network (along with the NC and the private interface of the CC) that is not reachable from the outside; ie, not reachable from the CLC? [19:53] mathiaz, ^^^ perhaps you know about that [19:53] EtienneG: isn't the SC only dealing with the CC? [19:54] mathiaz, that is my understanding, but you need to register it from the CLC [19:54] hence it needs to be reachable from the CLC, at least for the registration [19:54] EtienneG: if the CLC needs to know about the SC then yes the CLC and the SC need to be able to talk to each other :) [19:54] mathiaz, ok ... [19:54] I am perplexed [19:55] mathiaz, thanks for the info [19:55] EtienneG: why [19:55] is it possible to just delete all the crm resources and start over :( [19:55] hggdh, because I have to implement a topology where the SC is on a private network with the NC and a the private CC interface [19:55] hum [19:55] hggdh, and it just do not make any sense to me [19:55] * hggdh goes digging [19:56] that the CLC would need to talk to the SC [19:56] no, it does not make much sense [19:56] I get a Segmentation faulty tree 50% when I use aptitude/apt-get (new install of 10.04) [19:57] perhaps we can manually exchnage keys, but then, how do make the CC notice that it need to make use of the SC? [19:58] EtienneG: yes, the SC needs to be reachable by the CLC [19:59] hggdh, well then, so be it. what make you so confident in the answer? [19:59] is it documented somewhere, ro somesuch [19:59] EtienneG: man euca_conf ;-) [19:59] hggdh, I see you have jedi power! [19:59] I am actually shocked euca_conf has a man page at all :) [19:59] EtienneG: http://pastebin.ubuntu.com/446793/ [20:00] hggdh, that is about the CC [20:00] heh. I *remember* having to manually register an SC when I missed setting up the keys [20:01] EtienneG: no, you *need* a CC also (because the CC/SC are the ones to actually talk with the NCs) [20:01] this is about CLC and SCs [20:01] EtienneG: bloody hell [20:01] EtienneG: yes, you are right [20:02] the SC registration is done at the CLC, but it is the CC that talks with the SC [20:02] hggdh, yep, there we are [20:02] ok then [20:02] so it should work [20:03] * hggdh makes a note to read the whole thing before announcing it [20:04] * ccheney is about to test if euca is as screwed up as he thinks :) [20:04] i got a big issue, this is a unusual request. but for some reason i can't get to SSH on my ubuntu server anymore the router is blocking the port for some reason, so my question can i easily exploit a service to gain root and change the ssh port [20:05] DecryptedChaos: no [20:05] so theirs nothing i can do? [20:06] New bug: #591413 in ipmitool (universe) ".diff file leaves doesn't change timegm to gmtime on line 6782" [Undecided,New] https://launchpad.net/bugs/591413 [20:06] DecryptedChaos: i would investigate why suddenly the connection is no longer possible [20:07] well the portscan show that 22 is no longer open but that dosen't make sence because that machine is phyically open to the interent all ports [20:07] 80, 53, 21, 3306 all remain open [20:08] DecryptedChaos: and who manages this router/firewall? [20:08] its my equipment but its located far away and im trying to avoid going to phyically fix it [20:09] DecryptedChaos: someone either changed the f/w rules or turned off the ssh daemon on the server. seems weird if it's your equipment [20:09] ttx: ping [20:10] hggdh: pong -- but on a call [20:10] the werid part is that i lost access to the router at the same time as ssh it still works i just can't get into either of them for management [20:10] ttx: when you can -- why did you put the samba apport hook back in progress? [20:11] hggdh: we changed the blueprint at the same time [20:11] damn - installing a vm on kvm now - looking good - looks FAST [20:11] KVM rules [20:11] ttx: ah, OK. [20:11] hggdh: I just switched assignee to zulcss so that the spec shows up on our tracker [20:14] i guess my only option is get either go to the site or get someone thats on-site to open a vpm for me to get access to the network? [20:17] hm... [20:18] I installed kvm and all - it seems to have created a virbr0 interface - is this usable for bridging? [20:19] RoyK: sort of. [20:19] sort of? [20:19] RoyK: It's a bridge libvirt uses to connect NAT'ed VM's to the network. [20:19] ok, so for bridged networking, I need to do it manually? [20:20] RoyK: fsvo manually, yes. [20:20] fsvo? [20:20] For some values of. [20:20] as in edit /etc/network/interfaces etc? [20:20] i have another question.. why did ubuntu rip xen dom0 support my server dosen't have VT support and i tried to get xen dom0 running without luck because ubuntu removed dom0 i have been forced to use VMware Server and its terioble [20:21] RoyK: Yes. [20:21] DecryptedChaos: Because Xen is an absolute nightmare to maintain. [20:22] works loads better than VMware server and in that account i belive it was worth keeping [20:22] if not then kvm guys need to implementy PV [20:22] Not going to happen. [20:23] well thats redicilus [20:24] It's really not. [20:25] Xen is not a nightmare to maintain because their developers are idiots. Because they're not. It's a nightmare to maintain because it does what it does. Making someone else do the same, but starting from scratch is also going to be a nightmare. [20:25] im my prospective it is, because it forces guys like me with slightly older hardware to either use some crappy platform or pay thousands upgrading my hardware [20:26] Guys, i followed the instructions in the Ubuntu server guide in 10.04 to enable IP Masquerading, now I can't ssh into my machine? how can i enable this? [20:26] Yes. That was the case before as well. The crappy platform was just Xen instead of vmware. [20:27] let me ask you this about xen though the only problem i ever had with it. ( i installed it on top of CentOs jut to test) the vms ran great however would spike up to 20,000ms ping why did that happen? [20:27] okay, well i figured out using ufw allow 22 worked [20:27] but now i can't ssh within my local network [20:29] DecryptedChaos: Because it's a nightmare to maintain. [20:29] Probably. [20:29] Or a corrollary of that. [20:30] ok, got the bridge up (serial console is nice when not on-site) [20:30] or ssh. [20:30] ssh when configuring a bridge can be troublesome if you do something bad [20:30] i wish i could FTP as root by default because i could edit the ssh config problem solved [20:31] uh [20:31] RoyK: Ah, point. [20:31] who's in charge of this v [20:31] https://wiki.ubuntu.com/ClusterStack/LucidTesting#Overview [20:31] webPragmatist: You are now, because you asked. [20:31] "resource-stickines="100" [20:31] stickines? [20:31] okay [20:31] works like a dream :) [20:31] adhesiveness? [20:31] typo? [20:32] webPragmatist: Does it not work? [20:32] damn - kvm rocks - ubuntu rocks [20:32] webPragmatist: And does it work if you correct the spelling? [20:32] Can someone give me some pointers, I have a headnode running dnsmasq acting as a dns/dhcp server to 8 nodes, i had them all up, but they werent going out so I enabled IP Masquerading, following the server guide, now I can't ssh within my nodes... [20:32] I've been trying to run this shit on opensolaris for some time just to have zfs snapshotting etc, and this is a wee bit easier [20:32] webPragmatist: If you can answer affirmatively to both questions, then yes, it's likely a typo. [20:32] you are smart [20:33] * soren does try [20:33] webPragmatist: I assume you're asking because it doesn't work? [20:33] no actually i'm asking because i noticed it [20:33] why does this room have no ops at all lol i just noticed that [20:33] DecryptedChaos: it probably does just not "opped" [20:34] no reason to have visible ops around - they op when they need to [20:34] DecryptedChaos: ChanServ managed channels rarely have people who are OP'ed all the time. [20:34] DecryptedChaos: We get OP'ed when we need it and drop them again. It's like sudo for irc :) [20:34] Can someone give me some pointers, I have a headnode running dnsmasq acting as a dns/dhcp server to 8 nodes, i had them all up, but they werent going out so I enabled IP Masquerading, following the server guide, now I can't ssh within my nodes... [20:35] lol yeah. im a old school irc admin irc has changed a lot since i was big in it [20:35] hi [20:35] how can i create a startup for any programm? [20:35] create a startup? [20:35] autorstart [20:36] sry [20:36] se /etc/init.d/skel [20:36] your gonna probably need to write some bash script to execute it [20:36] see, even [20:37] copy the skeleton script to /etc/init.d/yourprogram, link to it from /etc/rc2.d [20:37] it's also mispelled here http://clusterlabs.org/wiki/Dual_Primary_DRBD_%2B_OCFS2 [20:37] huh? [20:37] poop [20:38] and of course, change the yourprogram file [20:38] okay [20:38] hm [20:38] hggdh: i see why i was confused, they call the exit value of a process in SystemUtil.run() 'pid' [20:38] i try it [20:38] ccheney: er, what? a pid is not a pid but the RC? [20:39] hggdh: yea brain damage [20:39] Can someone tell me how to enable ip masquerading on a local network without disabling ssh within the network? [20:39] hggdh: i had to lookup the java lang ref to see what it should be returning [20:39] heh [20:39] hggdh: http://pastebin.ubuntu.com/446805/ [20:39] hggdh: pid there is rc [20:40] wng-: it should be trivial - just google for iptables masquarding [20:40] RoyK: each time i follow the server guide it disables ssh within my nodes [20:40] anyone here use ClearOs Enterprise? [20:40] webPragmatist: I don't know if it's misspelled. Try it. :) [20:40] hggdh: which made sense considering they threw execution exception, but i was wondering why how a pid could be non zero and bad :) [20:40] wng-: just don't turn on ip filtering [20:40] at least for a start [20:41] ccheney: yeah. OTOH, it *does* throw an ExecutionException [20:41] which -- IIRC -- we did not see [20:42] so I have SSH setup with key-based authentication. copied the ssh key to the remote server, and was able to ssh into it. once I reboot the remote server, I get "permission denied (publickey)" until I restart sshd. any ideas? [20:42] ccheney: so, I wonder what, exactly, are the returns from Process.waitFor() [20:43] hggdh: on top of that it appears to call the stdout returnValue if i understand what it is doing [20:43] RoyK: hmm, i think it actually screwed with my NFS mounted home, so the sshd's on the nodes don't see authorized_keys [20:43] hggdh: http://java.sun.com/javase/6/docs/api/java/lang/Process.html [20:43] has anyone ever tried to tar -x a file with http path like http://ip/file.tar? [20:43] hggdh: return of waitFor() is exit value [20:44] hggdh: so basically $? [20:44] seems i have to wget first, then tar -x... [20:44] ccheney: yeah, OK, but what is output.getReturnValue()? A local method? [20:45] ccheney: and error does not seem used at all [20:46] hggdh: yea its not afaict, i'm not sure what that returns yet, I am going to add debug logging to see [20:47] I get a Segmentation faulty tree 50% when I use aptitude/apt-get (new install of 10.04) === amstan_ is now known as amstan [20:47] hggdh: i think i can remove the returnValue.length crap in createLogicalVolume entirely [20:48] ccheney: I do not remember the code you pastebin-ed, but a zero-leght would mark success [20:53] hggdh: hmm not sure what is going on then, it checks if 0 length and considers that an error [20:54] hggdh: once i determine what it is i might have to add it back, maybe its there is no error but no output from the command either [20:54] SpamapS: https://launchpad.net/~mathiaz/+archive/bzr-git/ [20:54] perhaps just a sanity check [20:54] SpamapS: ^^ this is an example [20:54] ccheney: this is indeed confusing. The run() method has only two returns, one with getReturnValue(), and one with an empty string [20:55] ccheney: and -- if I am reading it correctly -- the empty string is succecss [20:56] ccheney: but on createLogicalVolume(), the empty return is an error [20:56] WTF? [21:00] mathiaz: Thanks. I've got a couple of PPA's that I've been using thus far.. its really quite an awesome technology. :) [21:02] hggdh: yep, odd stuff [21:02] hggdh: will know in a few minutes, i rewrote the createLogicalVolume function to do proper error checking (i hope) and then will add logging of the value from run() [21:02] ccheney: so the other create*() methods must do something [21:02] hggdh: 35 line change for one function, heh [21:02] :-) [21:03] still with no docs, lol [21:04] might not be the most pretty way to do the change, i'm not really a java coder [21:04] and its still not perfect by any means due to the lvm2 return code problems itself [21:10] Ok so I have a mailserver that is also a dhcp/dns server. ip is 192.168.50.155, but also 192.168.1.1 for the local network. I have 3ware 3dm2 on each node attached to the 192.168.1.* network, but email notification is not working [21:11] ccheney: if you look at losetup() -- it does errorCode = proc.waitFor() [21:11] so it was just a Real Bad Naming issue [21:12] hggdh: yea [21:13] how would I force ubuntu-vm-builder to use a particular filesystem type for the root partition it makes? === jetole_ is now known as jetole [21:22] Ng: You can't right now, I'm afraid. [21:23] soren: fair enough, I'll just hack it :) [21:24] looks easy enough to make a new suite with the overrides I want [21:24] I'm getting some odd filesystem errors with ext4, so I want to try a guest with ext3 to see if the same thing happens [21:24] Ng: It's surprisingly difficult. [21:24] Ng: Actually. [21:24] Ng: Grab a bzr checkout and run from that. [21:31] mathiaz, did you see well above i asked about -backports ? [21:32] soren, find . -type f | xargs 'sed -i s,ext3,ext4,' [21:32] that should do it [21:32] :) [21:33] haha [21:33] Do most peopel have the universe repository enabled on their server? I'm concerned about the bit about no security review or updates. [21:36] smoser: Certainly. Adding another suite is just less fun that it should be. [21:36] agreed. [21:36] it'd be ace to be able to construct a suite somewhere in a local part of pythonpath and have it be available [21:36] bpgoldsb: there are updates to universe for security. [21:37] Ng: Hmm.. Yes. [21:37] Ng: Yes, it would. [21:37] * soren ponders [21:37] bpgoldsb: it's just community-driven, so only more popular stuff gets attention. [21:37] kees: So most stuff in it gets updated, just not by the Ubuntu Sec Team? [21:39] i spent the last 4 hours tring to get stupid Modo Rescue to work under 10.04 smoothly with little luck..any other simple single file image backup program out there that supports ext4? [21:39] or and can run on a live system [21:45] jdstrand, around ? [21:45] have you seen http://ubuntuforums.org/showthread.php?t=1468882 [21:46] i'm seeing similar output of 'udevadm' command. under some circumstances i get "custom logging function 0xb9237008 registered" as stdoutput [21:46] or "in stdout" rather. [21:46] I run my own internal DNS at home with bind on hardy. Can I use DNS to redirect queries from certain domain names to a bogus IP? For instance, if I wanted to block cnn.com for everyone on my internal network, I could just change its DNS IP to 192.168.0.2 or something. Is there a better way to do this or am I on the right track? [21:47] smoser: hey. I have not. I *think* kees may have seen something similar, but his issue may have been different [21:47] bpgoldsb: well, universe is very large, but it gets attention. it's best to be aware of what you've got installed from universe and keep an eye on it. [21:47] kees: ^ ? [21:48] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/570692 is also a hit for that string. [21:48] Launchpad bug 570692 in linux "Ubuntu 10.04 Can't open RFCOMM device: Permission denied" [Medium,Confirmed] [21:48] i had thought it might be coming from apparmour or something [21:48] but one way or antoher, that output gets sent, and the tool that parses it is not expecting it. [21:48] jdstrand, smoser: sorry, never seen that before. [21:48] kees: Do USN's get generated for Universe? [21:49] never mind. [21:49] its from udev [21:49] bpgoldsb: no, there there is a very early effort to publish "community usns" [21:49] bpgoldsb: bit it hasn't really gotten off the ground beyond a pilot announcement for clamav. :) [21:49] smoser: hi - I think both bugs are good [21:49] smoser: bug 589214 [21:49] Launchpad bug 589214 in lucid-backports "backport maverick ec2-ami-tools to lucid" [Undecided,New] https://launchpad.net/bugs/589214 [21:49] mathiaz, so as in what do i need to do [21:50] smoser: 589211 [21:51] smoser: hm - I'm not so sure about the versioning though [21:51] smoser: as it's a no-source-change backport [21:52] smoser: given that it builds in your PPA, I'd test it [21:52] smoser: if it works set the state to Confirmed [21:52] i've tested, it works fine. [21:52] but it wasn't clear to me that i was allowed to move it to confirmed. [21:53] smoser: ok - so I'd move it to In Progress and subscribe ubuntu-archive [21:53] smoser: as outlined in https://help.ubuntu.com/community/UbuntuBackports#Backport%20Process [21:54] smoser: hm - I'm not sure though - I haven't done -backports for a long time [21:54] smoser: so you may wanna ask ScottK about it [21:54] thanks mathiaz [21:55] SpamapS: around? [21:57] mathiaz: yes still here. :) [21:57] just enjoying the irony that the Velocity conference website is slow [21:59] SpamapS: :) [22:03] jiboumans: ping re. the internal UEC [22:04] hggdh: otp, will get back to you shortly [22:05] https://blueprints.launchpad.net/bugs/589566 [22:05] Launchpad bug 589566 in mongodb "Please merge mongodb 1.4.3-1 (universe) from Debian unstable (main)" [Undecided,Confirmed] [22:07] SpamapS: http://qa.ubuntu.com/reports/sponsoring/ [22:08] jdstrand, kees if you care... i just opened https://bugs.launchpad.net/ubuntu/+source/udev/+bug/591460 [22:08] Launchpad bug 591460 in udev "udevd daemon run with --debug affects output of 'udevadm info'" [Undecided,New] [22:09] it hopefully describes the issue i was seeing above. [22:15] Hi guys, I have an urgent thing here: I dist-upgraded lucid where linux-image(-2.6.32-22) was upgraded. update-grub however threw this at me: "error: unknown LVM metadata header.". I'm requested to reboot the server, however I'm unsure if its safe. Is it? [22:17] sveinse: have you seen this yet: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/452350 ? [22:17] Launchpad bug 452350 in grub2 "Unknown LVM metadata header" [Undecided,New] [22:19] sveinse: and this https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/385428 , specifically comments #3 and #6 [22:19] Launchpad bug 385428 in grub2 "grub2 boot from lvm Auto-detection of a filesystem module failed" [Undecided,Incomplete] [22:21] hmmm. #3 of bug 385428 didnt do it for me... [22:21] Launchpad bug 385428 in grub2 "grub2 boot from lvm Auto-detection of a filesystem module failed" [Undecided,Incomplete] https://launchpad.net/bugs/385428 [22:23] hggdh: i think i may have fixed the issue, but i'm starting to hit the other communicating one now [22:23] hggdh: actually nm i seem to have lost the setting for 200GB cluster [22:23] not sure how that happened [22:24] ccheney: you mean the CLC reconfigured itself back to 50G? [22:24] SpamapS: Well at least I see the same as the two bugreports you sent me. Thanks. [22:24] SpamapS: I just hope rebooting doesnt offline the server.... [22:24] hggdh: i think so, unless i did it and forgot i had [22:26] vgck and pvscan all reports lvm header checksum error [22:27] * sveinse dont like to be in this fix: To reboot or not to reboot... [22:27] hggdh: grr [22:27] hggdh: it seems to lose its setting when you restart euca [22:27] hggdh: is that supposed to just be a runtime config option? [22:31] AAAhhh! Got it. pvremove /dev/sda1 (the boot/ part) did it [22:31] Thanks SpamapS [22:33] sveinse: you may want to add a comment that the workaround worked for you. :) [22:33] Will do [22:34] sveinse: good luck on your reboot. :) [22:39] hggdh: hmm my patch isn't completely working for some reason [22:40] I get a Segmentation faulty tree 50% when I use aptitude/apt-get (new install of 10.04) any ideas how to fix this? [22:41] * ccheney notices his own typo/braindamage is the cause, lol [22:42] i'm somewhat surprised this didn't just completely explode [22:42] copy/paste gone awry [22:51] heh [22:51] ccheney: I would expect that changes in the web would be maintained [22:51] ccheney: I will restart the rig -- completely cycle it -- and test (since I also changed it) [22:52] mathiaz, smoser: any of you using the UEC test rig? [22:52] hggdh: nope - you can knock it out! [22:52] UEC rig is leaving the building... [22:53] UEC rig has left the building [22:56] hggdh: oh i think i know what might be causing it [22:56] ZenMasta: you mean you get a SIGSEGV in APT/aptitude? [22:56] hggdh: i'm having to nuke the storage db file, which probably has it in there [22:56] ccheney: why nuke it? [22:56] hggdh: since euca gets wedged reloading volumes that it doesn't know about [22:57] hum [22:57] hggdh: euca has a nice 'feature' of reloading volumes that don't show up in euca-describe-volumes [22:57] hggdh: so if they get partially removed they still get reloaded [22:57] ccheney: is this good? [22:57] hggdh: but still never show up in the euca-describe-volumes since they are in partial removed state [22:57] no a bad thing [22:57] definitely bad, then [22:58] yea :\ [22:59] hggdh no i mean I get a segmentation fault [22:59] soren: oddly I ran it from a bzr checkout, made my own suite which set preferred_filesystem to ext3 and it still made an ext4 root, so I just fudged the meaning of ext4 to TYPE_EXT3 ;) [22:59] rerunning test now with the proper reversion calls (no more typos i think) [23:00] and no filesystem errors yet [23:00] ZenMasta: which Ubuntu version? [23:00] is there some known badness with kvm/virtio and ext4? I wondered if it wasn't getting to do a final flush before being "powered" off [23:00] hggdh 10.04 http://pastie.org/997027 [23:02] ZenMasta: it sounds more like apt was happily loading the tables, and got hit by a seg fault -- so it is not "segmentation faulty tree", but "segmentation fault" [23:02] hggdh: i think my new code fixes the create side at least [23:02] ZenMasta: which overwrote the line [23:02] ZenMasta: your best bet is to open a bug for that [23:03] Ng: Which bzr branch [23:03] soren: oh hrm, I just blithely branched lp:vmbuilder [23:03] hggdh: i now don't have any partials left after creation [23:03] hggdh this is a clean install two, happened on desktop and server iso [23:03] Ng: That's fine. [23:03] *too, even [23:04] ZenMasta: even more reason to open a bug ;-) [23:04] where do I do that? [23:04] Ng: I can't guess why that would be. I'd be happy to look at your patch and commandline and find out. Tomorrow :) [23:04] ccheney: are you going to publish it? [23:04] ZenMasta: sudo ubuntu-bug apt-get [23:05] hggdh: will once i get it working well, i still need to do something about deletes [23:05] hggdh: i'm pretty sure deletes are still broken [23:05] hggdh what does that do... because I got a segmentation fault on that too [23:05] soren: k :) [23:05] ccheney: I am 100% sure, I got to clean up the pieces when I ran the delete [23:06] ZenMasta: this is rather strange. You installed clean, or upgraded? [23:07] hggdh installed clean, no dual boot etc [23:07] ZenMasta: another option -- if you have GDB installed -- is to run apt-get under GDB [23:07] of course it would help a lot to have the debugging symbols [23:08] hggdh unless that came with the default server iso then its not installed [23:08] ZenMasta: dpkg -l gdb\* [23:08] i'll run that but pretty sure that's a neg [23:08] nope, no package matching [23:08] hggdh: yep delete is still broken with my code too [23:09] ccheney: heh [23:09] hggdh: its going to take some more thinking about what to do to fix the clean part [23:09] hggdh: i may have to regex the output from the command if its available in the right place [23:09] ccheney: no prob... we are already much better because of your patch [23:09] hrey all [23:09] hggdh: so if it shows as waiting on lock to retry until it removes it [23:10] any other people with suggestions of software such as Acronis for backup in linux, about to give up [23:10] ccheney: OK. [23:10] hggdh: i'll push what i have to my lucid-sru branch for now, but my patch probably will need a full rewrite to be inside the functions to do each thing itself instead of the ones calling them, eg i think createLogicalVolume probably should clean up itself and then throw instead of making createVolume do it [23:11] ccheney: this will be more invasive, but sounds much more correct [23:11] hggdh: right now my patch has the calling function doing all the cleanup with is bit messy [23:12] ccheney: well, they will still have to cater for failures [23:12] ZenMasta: I am stumped [23:12] yea but a bit more clean than currently [23:13] well i need to look at the code more and think about it, will do that later tonight [23:13] * ccheney has to run to dinner and over to the hospital to relieve his mother in law [23:15] Why cpu.out is empty when running this command "top bd1 -p 15436 | grep 15436 > cpu.out" but I see results when running this: "top bd1 -p 15436 | grep 15436 [23:15] I just dont get it === dendrobates is now known as destro === destro is now known as dendrobates [23:21] New bug: #591481 in tftp-hpa (main) "package tftpd-hpa 0.49-1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 127" [Undecided,New] https://launchpad.net/bugs/591481 [23:29] hggdh, another funny one (which I will not report; out of time): after manually installing the eucalyptus-nc package, you can register the service no problem and it will will report resource no problem. However, it will fail running an instance with the infamous "libvirt: hvm not supported" error. You reboot the NC, and then it works just fine. [23:29] this thing is driving me insane [23:29] EtienneG: this is weird [23:30] EtienneG: and ccheney has been working hard on getting volumes to behave nicely [23:30] hggdh, i'm not using rig [23:30] hggdh, thanks $deity for his work on that! [23:30] smoser: too late, already rebooted them ;-) [23:30] hggdh, anyway, not a show-stopper so I will move on, but we can keep that in the back of our brain [23:31] EtienneG: this is the Dell rig? [23:31] and my brain is frying... [23:41] On Hardy, when I create a new logical volume, why is it not automatically added to /dev/disk/by-uuid by udev? [23:41] Do I need to poke something? [23:44] It appears the 60-persistent-storage.rules specifically skips all "dm-" devices. [23:48] Insyte: you still get the /dev/vgname/lvname link though right? [23:49] Yep. [23:49] Insyte: shouldn't present any issues then. [23:50] You're right, of course. I can certainly use those. I'm just curious why the LVs I created when I installed the box are getting the UUID symlinks, but not new ones. [23:51] I can't find any udev rules that would create them, nor anything in init.d. Yet they exist... [23:59] Insyte: maybe the LVM metadata is pulled from the wrong location, which hasn't been updated since installation. [23:59] It seems to be consistent across all of my Hardy boxes...