[00:31] is there a way to do a network install from the generic ubuntu server cd (14.04 LTS)? [01:08] well, I ended posting it on askubuntu anyway http://askubuntu.com/questions/708356/how-can-i-install-from-network-sources-using-the-generic-ubuntu-server-cd [01:46] 3/win 3 [02:10] Synthead: do you mean making sure you're up to date during the install? or do you mean loading hte installer components over the network? === CiPi is now known as cipi === Lcawte|Away is now known as Lcawte === cipi is now known as CiPi [09:51] Hi [09:52] http://oi65.tinypic.com/r8s2ux.jpg I have enabled headers and getting correct CORS in headers (access-control-allow-orig... *) but still I cannot access the inner dome or page height of an iframe in side a parent page. It says "Error: Permission denied to access property "document"" any clue? === cpaelzer is now known as cpaelzer_afk === cpaelzer_afk is now known as cpaelzer [12:01] hi all i have a crontab task (with mysqldump command) when run the task thhe backup was generated in 0 bytes, but when run same command manually it generate the correct backup, any advice how to know why? thanks [12:08] jak2000: Any stderr output, captured by mail? Perhaps some PATH or other environment issue? [12:10] jak2000: Unless you have a local MTA setup, perhaps you can modify the script to write STDERR to a separate file? [12:34] q/win 13 [12:37] hello world [13:25] andol! [13:25] i have full paths [13:27] see please: http://pastie.org/10625270 [13:38] andol? [15:44] how can i determine why a 12.04 server spontaneously rebooted? [15:44] logs look normal one moment, and then everything is starting up the next [15:49] Disk failure? [15:50] would that be logged anywhere? [15:51] Well that might be a problem ;) Suppose you want to write to the disk, but the disk failed... [15:51] the logging volume didn't fail [15:52] Hmm, I'd check SMART anyways. [15:52] yes, thats a good idea, thank you [15:53] power-loss would be the obvious candidate [15:53] TJ-: yes, but to only one server of many in a rack? [15:54] i'm not sure how the rack is powered, but i'd thikn there would be other issues too [15:54] i'll keep that in mind [15:55] but how could i tell if that was the issue? [16:03] LzrdKing: local PSU issues possibly. Does the server have IPMI? could remote-hands of accidentially nudged cables on that rack - or is it caged? etc [16:03] lordievader: Device does not support SMART [16:04] TJ-: yes it does [16:04] LzrdKing: anything in the logs of the *other* servers at the time that server died? [16:04] i'll look [16:04] good morning guys, happy friday. I have a question, I have a VM that takes long to boot up. It hangs saying the following: Waiting for network configuration, Up to 60 more seconds for network configuration .." [16:05] I use /etc/network/interfaces to configure networking [16:05] TJ-: IPMI might be able to indicate better what happened? [16:05] LzrdKing: Huh? Is it a non-standard disk? [16:05] not sure what could be causing this, Ubuntu Server 14.04 LTS [16:06] lordievader: could be behind RAID controller [16:06] jge: so eventually it does come up w/ networking? [16:07] jge: that times out if the network isn't available; check the config, are you trying to automount remote network file-systems? [16:07] yes, its an SMC RAID [16:08] Ah, that explains it indeed. [16:09] nacc: yep it comes up fine [16:09] TJ-: nope [16:11] this is what my /etc/network/interfaces looks like:http://pastie.org/private/jcat5qnhb3debzh4fodxuw [16:11] very simple, I'm wondering if it's because the next hop for that static route is on a different subnet [16:12] I'll remove it and test [16:13] ha, that did it! [16:13] freaking booted in seconds :D [16:16] i have a laptop that hangs there for a while too, i should look at that [16:18] jge for the reasoning of that message see /etc/init/failsafe.conf [16:25] TJ-: will do, thank you. === cpaelzer is now known as cpaelzer_afk === cpaelzer_afk is now known as cpaelzer [17:11] When ever I try to boot into ubuntu I keep getting this error http://postimg.org/image/yi2biav9n/ [17:12] pod_: and it doesn't boot? or do some devices not work? do you have a USB hub? [17:13] I am using a desktop and only using the usb ports that are on my motherboard. [17:14] Ubuntu is already installed on my hard drive [17:14] pod_: is your concern the messages themselves, or is something specifically not working? [17:16] I am just trying to fix the error [17:18] pod_: I believe that is indicating an issue with a USB hub, including possibly the one on your mobo. Could be insufficient power (if it was external, e.g.) for the load, etc. Does it happen w/o anything plugged into the USB ports on the motherboard? Do any of the USB ports on your machine work? [17:19] Yeah I tried booting back into Ubuntu when only the keyboard plugged in and still get the same error. [17:20] When ever I use debian all the ports work fine [17:20] pod_: ok, and do your devices work? e.g., your usb keyboard? [17:20] pod_: what version of Ubuntu and what version of Debian? === cpaelzer is now known as cpaelzer_afk [17:21] ubuntu 14.04 and debian 8.1 [17:25] pod_: 14.04.3 or 14.04.0? that is, which kernel probably matters, as 8.1 uses 3.16.7 (I think) and 14.04.0 used 13.13 (and each .x update has bumped that, so for instance, 14.04.3 is on 3.19 [17:25] if Debian 8.1 worked without the errors, I would guess it's just a kernel fix and I'd try the latest 14.04.3 release [17:25] if possible [17:26] Okay I'll wipe the older ubuntu and install newer ubuntu [17:27] pod_: so you were on an older ubuntu before? shouldn't be ncessary to wipe it, just boot the live image and see if it works? [17:27] TJ-: ilo has nothing useful logged either [17:28] I upgraded from 12.04 [17:29] nacc, it only bumped it if you started with 14.04.2 as the install base [17:29] not if you upgraded to 14.04.3 [17:29] is there a way to run a raid utility without rebooting? [17:29] patdk-wk: good point, sorry [17:31] not sure why a raid utility would needa reboot [17:31] unless it requires dos/windows [17:31] Software raid [17:32] patdk-wk: my understanding was that 14.04(.0) was 3.13 based, 14.04.2 was 3.16 based, 14.04.3 was 3.19 based, and 14.04.4 will be 4.2 based? although those may be specific packages you pick to install from the HWE stack [17:33] i don't know, i've only ever seen raid info come up during a boot when it says siomehting like "press F2 to configure" [17:34] nacc, that is true [17:34] but you have to INSTALL 14.04.2 from a iso for HWE to be enabled [17:34] patdk-wk: ah i see what you're saying [17:34] patdk-wk: yep, makes sense [17:35] or enable it yourself [17:35] for me personally, hwe has been nothing but problems [17:35] patdk-wk: like what? [17:35] due to kernel dmks build issues [17:35] ah [17:35] external modules? [17:35] external to the kernel [17:35] yeah, sorry, that's what i meant [17:35] the packages dkms packages don't maintain kernel compatability with HWE kernels [17:36] so using open-vm-tools, xtables, ..., can cause you lots of issues [17:36] though, open-vm-tools is better lately though [17:36] well, the (upstream) kernel doesn't guarantee API or ABI compatability release-to-release; that's probably the biggest issue? [17:36] yes [17:36] I had to fix a broken part of the kernel from a old verison of ubuntu to fix error 71 [17:38] and I had an issue with libc and hwe kernel [17:38] but that is so esoteric, I didn't dig into it much, once I found it was a libc issue [17:39] but that is the other kernel api/abi problem [17:40] patdk-wk: interesting, i haven't personally seen those problems, but i bet a bug would be appreciated :) [17:41] I thought about it, but reproducing it, by not me, would be impossible :) [17:41] so I didn't bother [17:41] heh [17:42] it's a closed source binary, that I have heavily hacked [17:42] ah [17:42] and is failing on an exec call to run sendmail [17:42] or rather, it works, but stdin isn't working right [17:42] but in all other cases, it works === sarnold_ is now known as sarnold === Lcawte is now known as Lcawte|Away [19:14] hey folks! I would like to test the Unbuntu Cloud (MAAS, Juju, Landscape etc) on a couple of lased bare-metal servers. note that these servers don't offer a vlan, so all I have is external IPv4 IPs. can I work this out, or do I need to have everything in the same place? I'm reading about MAAS right now, and from my understanding it needs to be all in the same network, as MAAS is basically [19:14] setting up LAN IPs via DHCP? [19:15] Razva: indeed, maas does pxe via dhcp. probably your ISP won't be impressed if you start offering dhcp over the internet :) [19:16] sarnold yyyup. :) ok, so what's the solution in my case? [19:16] I really don't have 5 PCs at home in order to test it out, nor a powerful PC to install/run eveyrthing virtualized. [19:17] so I was thinking to buy 5 cheap (kimsufi) bare-metal servers and start playing [19:17] can someone help me with kvm? I have a 64 bit 14.04.3 install but kvm keeps making 32 bit vms. --arch is set for amd64. Any thoughts on what to look at? [19:18] dlb: are you perchance running 32bit guests by accident? [19:19] I am trying to build the guests. When I build them, they are 32 bit. I expected 64 bit. [19:19] Razva: you may wish to try in #maas .. my best guess is something like set up an openvpn or ipsec vpn over your five cheap hosts, and virtualize a machine on each or something. it's not a great idea. hehe. [19:19] The host is 64 bit [19:20] sarnold already did, nobody is replying right now. :| [19:21] is MAAS mandatory? [19:21] Razva: no [19:22] Razva: maas can make deploying openstack more convenient, since you can use juju to deploy the openstack software to the hardware; but you can certainly install openstack by hand on the machines [19:23] Razva: I hope kimsufi has some way to firewall those machines off from the internet; I slightly doubt openstack services are intended to be exposed on the public routable internet :) [19:24] sarnold it's just some test machines, I won't store any data [19:25] yeah so the whole idea was to use ubuntu cloud as I'm an OpenStack newb :| [19:25] Razva: good start; I'm just not convinced thta the services are sufficiently defensive about their inputs :/ [19:26] will Autopilot work, or is that MAAS dependant? [19:28] I think it requires maas [19:32] darn [19:32] so at this point basically ubuntu cloud is "useless" in my case? [19:32] because Autopilot is the main reason to use the ubuntu cloud vs "standard openstack" [19:40] Razva: that might be the case. perhaps you can still use e.g. juju with the charms to deploy openstack if you use the ssh provider to juju.. [19:42] autopilot may not work for you but you might still be spared the difficulty of setting up openstack services manually :) [19:51] sarnold true... is there any manual of some sort, that will spare me of setting up? [19:53] Razva: not that I know of; I think it's mostly assumed that if you're interested in openstack, you've got the hardware on site to make it happen :/ [19:54] Razva: so putting together something a bit piecemeal out of cheap servers already on a cloud provider is a bit off the well-worn path. [19:54] Razva: this looks like The Guide for using juju with the ssh provider: https://jujucharms.com/docs/1.24/config-manual [19:55] Razva: you may need to read the autopilot script to figure out which specific settings it's using for which services, and you can find the service charms at e.g. https://jujucharms.com/q/?tags=openstack [19:57] I'm curious, how do you handle mutiple physical locations in this case? [19:57] you'd define different availability zones [19:58] ok, but how can MAAS allocate IPs into a totally different physical location? [19:58] you need to have vlans between physical locations or...? [19:59] I think each availability zone would have its own MAAS configurations; if one is flooded or vaporized or whatever, you wouldn't want services in the others to die as a result [19:59] aha, so you have more MAAS configs into the same Landscape/Autopilot? [20:01] hmmmmm, I wonder, I forgot that autopilot is now integrated into landscape. :/ I wonder if I've been on the wrong foot this whole conversation. (I'm sorry, it'd been ~six months since I last looked into autopilot...) [20:02] I wonder if the "new" autopilot still requires brand new machines or if you can use landscape to add existing machines into a cloud.. [20:04] any idea where should I ask? [20:04] http://askubuntu.com/questions/708608/playing-with-ubuntu-cloud-on-different-machines-without-vlan [20:05] jcastro_: http://askubuntu.com/questions/708608/playing-with-ubuntu-cloud-on-different-machines-without-vlan ^^^ [20:14] oh, so I can run it in vSphere locally on my "medium-sized i7 PC"? [20:15] on it [20:15] yeah, on it [20:16] just installed vSphere (free) so brb reboot [20:16] maybe; I saw int eh virtualized bit that there's a 8gigs memory suggestion/requirement for the machines.. and while that probably doesn't actually mean 40 gigs, it might not run great if you've only got e.g. 16 available .. [20:16] sarnold I just want to see how it works...that's all... [20:22] I think these days we can smoosh it onto like three servers [20:28] zul: heya, is all your lxd/openstack stuff landed somewhere? That might be an interesting option [20:32] sarnold: wow, you're not going to believe how simple the answer is [20:32] http://openstack.astokes.org/ [20:39] jcastro_: dude. [20:39] Can someone help me determine why I cannot set up a print server on a ubuntu server? "nmap" from a wireless laptop does not see the open 631 port but "nmap" from the server itself reports that the port is open. [20:40] sarnold: I know right, dude look at the video: https://www.youtube.com/watch?v=1eUoamVBzFI [20:40] geetar: check netstat -lntp, make sure it's bound to a public IP address or * [20:40] openstack on a nuc [20:40] jcastro_: I just watched the w hole thing. insane. [20:40] any idea how much RAM's in that? [20:42] sarnold: quick googling indicate some models have capacity up to 16G, the latest ones seem expandable to 32G [20:43] nacc: it feels like a lot to ask of an 8gig machine but 16.. maybe. hehe. :) [20:43] http://www.intel.com/content/www/us/en/nuc/overview.html [20:43] sarnold: yeah :) [20:44] might try that out when I reinstall my 32G box at home [20:54] sarnold: I guess I am doing it wrong, what do you mean "bound" to a public ip address? [20:55] sarnold: otherwise I do not understand the output. [20:57] geetar: sockets are 'bound' to an pair; that lets you run one webserver locally on 127.0.0.1:80 but a different webserver on 192.168.1.2:80 and a third on 10.0.0.1:80. [20:58] geetar: so this is checking if cups is configured to only listen on e.g.127.0.0.1, in which case it won't be visible off the machine, or if it is bound toa specific IP that you're not talking with, or if it is bound to that port on any IP.. [20:58] geetar: this lets you know if you need to fix the firewall configuration on a machine or router or if you need to fix the daemon configuration [21:07] sarnold: ah, so I don't understand the output. The result of "netstat -lntp" on the server is at http://paste.ubuntu.com/13942789/. Note I can ping the server from the laptop. [21:08] geetar: on my system cups is listening on 631, it looks like the same for yours, lines 14 and 27 [21:09] sarnold: Same for mine. So, could it be the router is blocking 631 traffic? [21:09] geetar: is your router perhaps forbidding wireless connections from communicating with wired? try connecting to port 22 or port 80 to see if this is unique to this service or computer.. [21:10] geetar: are you using iptables or ufw or similar firewall on this system? [21:11] sarnold: port 80 works, I have a wiki on the server. [21:11] sarnold: that I can access from wireless. [21:13] sarnold: I will see if it is iptables or ufw. I was thinking it was the router, but it should not be if I can access the wiki. Thanks for helping me think it through. [21:13] jcastro_, yes we have bundles for them [21:13] geetar: it might still be the router.. hehe. lots of fun places for a firewall to live. ;) [21:14] sarnold: drats. [21:31] zul: I found the single node installer, I'm good [21:31] zul: that thing is brilliant btw [21:35] sarnold: The first problem was the firewall as you suspected. I fixed that and got into the CUPS webpage. Still not printing because no admin access through browser but I think I can fix that. Why did nmap run on the server report that its 631 port was open if it was blocked by its firewall? [21:36] geetar: the firewall probably only blocked packets coming into the ethernet cards or leaving via the ethernet cards; when run locally, they go over the lo interface, which is only rarely firewalled [21:36] sarnold: thanks again [21:37] (well, I don't think I've _ever_ seen lo firewalled; it's possible, though. :) [21:51] when I install fail2ban does it do something out of the box? I only want it to do something custom not SSH, etc. [21:53] soahccc: iirc it does ssh out of the box - but it's easy to configure it to do whatever you want [21:53] well, it will detect [21:53] but if it blocks properly, based on whatever firewall your using [21:53] you might need to adjust it [21:54] RoyK: yeah if you know regex :) I'm afraid I ban all requests [21:54] soahccc: then learn regex ;) [21:54] soahccc: it's not too hard [21:55] soahccc: setup a VM first and try a bit before you go in production [21:56] RoyK: where is the fun in that ;) nginx + path + status code, can't be that hard === Lcawte|Away is now known as Lcawte [22:38] I hope fail2ban tails the file after reading it once :S It takes forever to parse my little 4 GB access log