=== freeflying_away is now known as freeflying [01:04] Hello everyone, somebody can tell me how i could know if I must to load some module of iptables or if not? [01:55] hi all [01:55] anyone want to talk ? [01:55] I have a plan I would appreciate some input [01:55] :) [01:56] I am thinking about purchasing an old server for $100 that supports 4 drives [01:57] four sata drives.....and run software raid between them [01:57] I would install ubuntu-server [01:57] and install the necessarily packages for a 4 X 2TB mythtv backend [01:57] :D [01:58] minus the loss space due to raid 5 or 6 [01:58] any thoughts? [02:05] mattwj2002, you'll get much better performance with raid 10 [02:06] Question: What package provides the facility to cache unlocked SSH keys? As in, I SSH, it prompts me for password, and then I can use it for a certain timeout without retyping the password? [02:06] qman__: [02:06] what is the drive lose with raid 10? [02:06] 1 drive? [02:07] I mean how many can you lose and still recover? [02:07] :) [02:07] you can lose any one, or a particular two [02:07] qman__: it might be faster [02:07] but.... [02:08] you lose too much data with raid 10 [02:08] this isn't overally critical [02:08] this is a hobby box :) [02:08] I'll probably wipe it from time to time [02:08] you're no less protected than with raid 5 (in fact you're better protected with raid 10) [02:08] right [02:09] but 4 X 2T in raid 5 gives you 6 GB [02:09] in 4 X 2T in raid 10 gives you 4G [02:09] that is quite the loss [02:10] 2 drives basically [02:10] actually I might go a collection of drives [02:10] and say screw raid all together :) [02:11] but I am thinking raid 5! [02:11] if you don't care about your data, go right ahead [02:12] qman__: did you see what I wrote before? [02:12] my data is important to me, so I do what I can to protect it [02:12] I hear you [02:13] if it was more important....I would protect it more [02:13] :) [02:13] basically I record tv shows using mythtv......and I want more storage space [02:13] raid 0 will give you the most performance and will give you the full capacity [02:14] but provides no redundancy at all [02:14] yeah [02:14] I think I'll go raid 5 [02:14] :) [02:14] some redundancy [02:14] raid 5 will be slow and will not offer much protection with consumer grade drives [02:15] for a good explanation of why, http://www.zdnet.com/blog/storage/why-raid-5-stops-working-in-2009/162 [02:15] mattwj2002: You do realize that if a drive fails and you replace it, moving around that much data between drives in RAID 5 is going to take a while? [02:16] mattwj2002: To the point where another drive might fail in the rebuild process? [02:16] of course [02:17] basically, unless you buy server class drives, you're looking at coinflip odds on a successful rebuild [02:17] O.o [02:17] to be a 100% [02:17] this is a home thing.....and I all ready have a couple of 2TB drives I have had for a while [02:18] 100% honest I mean [02:18] yeah, you really shouldn't do raid 5 with that [02:18] so raid 6 versus raid 10 [02:18] do raid 10 if you want redundancy, or do raid 0 if you don't [02:18] what do you recommed? [02:18] you recommend 10 over 6? [02:19] raid 6 is even slower and suffers (to a lesser extent) the same problem as raid 5 [02:19] mattwj2002: Recommendations are useless, it depends on what you yourself need. [02:19] good point [02:19] I have a raid 6 that I'm in the process of decommissioning [02:19] mostly because it's too slow [02:20] I like to mess around with this stuff [02:20] :) [02:20] I peak around 35MB/s write, with hardware that individually is capable of 120MB/s or more [02:21] I actually just purchased a bunch of new 3TB drives to create a raid 10 to move my data to [02:21] with 4 disks why would you do raid6? raid10 gives you the same space with better protection [02:21] good point [02:24] I guess I am not sure what is more important.... [02:24] it is obviously a personal choice [02:25] more storage with less protection? [02:25] more less storage with less protection [02:25] it isn't like I am saving medical records or something [02:26] this is just a server in my house to recording a crap ton of tv shows from my basic cable :) [02:26] 8 TB would get me up to about a couple months of straight video recording [02:26] I have a single 2 TB drive now [02:29] if the drives last that long [02:29] and mythtv has the functionality to put one program on one drive and another program on a different drive [02:30] qman__: you are saying raid 10 then [02:30] for my application [02:31] yes, it's what I would use [02:33] okay cool === freeflying is now known as freeflying_away === freeflying_away is now known as freeflying [05:27] hi guys [05:27] I have a crazy idea [05:27] anyone want to listen? [05:28] they sell a 15 bay SATA storage nas [05:29] I was thinking if you could put 15 3 TB drives in a raid 60 setup [05:29] that is like 33 TB [05:29] :) [05:30] Hi, Could you please let me know how can I check what graphic card can I use with a known motherboard? [05:31] does this help mojtaba lspci -v | less [05:33] mattwj2002: Actually I want to buy a second hand computer and I want to replace its graphic card, but I do not know how can I check what kind of graphic card could be used with that motherboard. [05:33] oh okay [05:33] I am not sure [05:34] hello [05:34] Hi, Could you please let me know how can I check what graphic card can I use with a known motherboard? [05:34] sorry [05:36] mattwj2002: thanks [05:36] worries [05:37] :) [05:37] mojtaba: have you been in this room for a while? [05:38] mattwj2002: You mean today? [05:38] yeah [05:39] no, just from 10 min ago [05:39] why? [05:39] I want to build a nas [05:39] a custom built nas [05:39] :) [05:39] I know that there are some linux distro specially made for NAS [05:40] the software I think I'll have undercontrol [05:40] because it is going to be a mythtv system (might do that separate not sure) === freeflying is now known as freeflying_away === freeflying_away is now known as freeflying [06:05] anyone have experience with bind9 / webmin === level15_ is now known as level15 [07:19] hi guys [07:19] anyone here? [07:20] I just looking into a custom build nas [07:20] $30 [07:20] oops [07:20] $300 including a ups [09:49] hello i have placed this command in my rc.local Xvfb :99 -ac -screen 0 1024x768x8 > /tmp/xvfb.log 2>&1 & but it is not runned on boot what am i doing wrong? [10:15] ws2k3: did you put the full path to the Xvfb command in there? [12:34] roaksoax: (part 1) http://pastebin.ubuntu.com/6083298/ [12:46] which package is it that includes the 'file' util? [13:15] roaksoax: (part 1) http://pastebin.ubuntu.com/6083298/ === marcoceppi_ is now known as marcoceppi [13:30] RoyK: file [13:31] dpkg -S `which file` [13:49] roaksoax: https://code.launchpad.net/~zulcss/nova/2013.2.b3/+merge/184579 [14:21] roaksoax: https://code.launchpad.net/~zulcss/neutron/2013.2.b3/+merge/184583 === rdw200169_ is now known as rdw200169 [14:47] so, if upon upgrading a server instance, it is hung on selecting a PAM profile and has deleted everything , like apt, what can I do? [14:48] have I done the unthinkable? === gigantor is now known as raininja [14:50] so, if upon upgrading a server instance, it is hung on selecting a PAM profile and has deleted everything , like apt, what can I do? === raininja is now known as Guest1234 === Guest1234 is now known as raijin [14:54] smb: ping [14:54] zul, yo [14:55] smb: the FFE for xen went through can you get me a list of stuff that you need me to do and ill do it this afternoon [14:55] 36 [14:56] zul, can, do. Well, step one is check/sponsor the xen-4.3 src pkg on chinstrap:~smb/4review [14:57] so I am going to have to rebuild this install by hand basically [14:57] using wget and dpkg [14:57] ffs [14:58] smb: and rebuild libvirt right? [14:59] zul, yeah, libvirt just as a no-change rebuild. I have to give you an update for xen-api to adapt the lib paths there [14:59] smb: okies cool [15:06] zul, Ok, mail sent and xen-api src pkg in same directory [15:06] smb: i saw thanks [15:07] so, if upon upgrading a server instance, it is hung on selecting a PAM profile and has deleted everything , like apt, what can I do? [15:07] on 12.04, how does cron send mail? does it have a list of places to look for sendmail? === LargePrime is now known as Guest19152 === LargePrime_ is now known as LargePrime [15:44] so, if upon upgrading a server instance, it is hung on selecting a PAM profile and has deleted everything , like apt, what can I do? === LargePrime is now known as Guest58042 [17:08] hallyn_: im gonig to do a no-change rebuild libvirt for libxl do we need anything [17:12] hi, i have issues whit my ubuntu-server, i can't connect to other gentoo server using a cross-over... [17:14] Richter: did you manually assign an IP address and netmask to both interfaces? did you restart whatever services you want to use after manually assigning IPs? [17:14] yes [17:15] i use the same config of the last server [17:15] sarnold, auto eth1 [17:15] iface eth1 inet static [17:15] address 10.100.100.2 [17:15] netmask 255.255.255.252 [17:15] network 10.100.100.0 [17:15] broadcast 10.100.100.3 [17:17] sarnold, this show in my syslog: Sep 9 13:57:17 huguinho kernel: [ 2118.053008] IPv6: ADDRCONF(NETDEV_UP): eth1: link is not ready [17:17] sarnold, but i am not using ipv6 [17:17] Richter: what's the configuration for the gentoo machine? [17:18] sarnold, config_eth3="10.100.100.1 netmask 255.255.255.252 brd 10.100.100.3" [17:18] mtu_eth3="9000" [17:18] this in my nfs storage [17:19] Richter: check arp -n output on both machines? [17:19] sarnold, this in my old server: config_eth5=( "10.100.100.2 netmask 255.255.255.252 brd 10.100.100.3" ) [17:19] mtu_eth5="9000" [17:19] nops [17:19] 200.143.116.1 ether 00:1b:21:bc:5c:72 C eth0 [17:19] 200.143.116.* ether 00:d0:b7:85:f7:72 C eth0 [17:20] if you're using jumboframes, make sure you use it on all endpoints [17:20] RoyK, the server is in other datacenter [17:20] ok [17:21] RoyK: will the cards auto-negotiate the lowest MTU between them? or will mismatched MTUs cause reliability problems? [17:21] RoyK, i was thinking... maybe the 'dudes' dont are using a crossover cable [17:22] sarnold, probably is default config [17:22] Richter: if even one card is gigabit, it ought to auto-negotiate, right? [17:22] sarnold, i really don't know [17:23] sarnold, but this 2 server uses gigabits cards [17:23] they are not old servers [17:23] sarnold: MTU must be the same on the whole broadcast domain. Beyond that, path mtu discovery should work when you have router hops. Unless you have a misconfigured firewall anywhere. [17:23] sarnold: I don't think MTU is autonegotiatable [17:23] Richter: I wouldn't expect crossover vs not-crossover to make a difference, unless ethtool or similar has been used to force a specific configuration [17:24] maswan,RoyK, thanks. [17:24] Richter: so, check mtu on both interfaces again? [17:24] Richter: no need for crossover with gigabit [17:24] crossover vs not-crossover should be autoneg on gigE or above [17:24] maswan, is a crossover connection... don't have firewall [17:24] Richter: Ok, so the same broadcast domain then. Then MTU needs to match on both sides. [17:24] sarnold, how i check this? [17:25] maswan: even most 100Mbps cards autonegotiate crossover [17:25] RoyK: Yeah, but it isn't guaranteed. [17:25] Richter: ifconfig or ip both report mtu [17:25] sarnold, lol, sorry [17:25] but "link is not ready" shouldn't be it for MTU mismatch, that should rather end up as large packets just being lost. [17:25] gentoo is MTU:9000 [17:26] maswan: I think that's just ipv6 whinging, business as usual there :) [17:26] ubuntu server is MTU:1500 [17:26] lol [17:26] sarnold: Ah, ok. [17:26] Richter: Yeah, start by fixing that. :) [17:26] RoyK: do you get tired of always being right? :) [17:26] Richter: heh - then set MTU on the ubuntu box ;) [17:26] sarnold, i can force this in interfaces file? [17:26] sarnold: hehe - no ;) [17:27] sarnold, : ( [17:27] Richter: 'mtu 9000' .. [17:27] sarnold, where i change this? [17:28] Richter: in the 'iface eth1' section [17:28] sarnold, /etc/network/interface [17:28] sarnold, ? [17:28] Richter: /etc/network/interfaces -- note the 's' [17:29] sarnold, ops [17:30] the MTU is correct now, but still don't work :O [17:30] Richter: what doesn't work? [17:31] RoyK, PING 10.100.100.1 (10.100.100.1) 56(84) bytes of data. [17:31] From 10.100.100.2 icmp_seq=1 Destination Host Unreachable [17:31] ping should work even with an MTU mismatch [17:31] RoyK, hummm [17:31] since the packages sent by ping are so small anyway [17:32] Richter: are these machines on the same network? [17:32] RoyK, yes [17:32] RoyK, its a crossover cable [17:33] Richter: pastebin output of 'ethtool eth0' [17:33] but eth0 is the wan link [17:33] eth1, then ;) [17:33] RoyK, use a eth1? [17:34] RoyK, kkk [17:34] RoyK, root@huguinho:~# ethtool eth1 [17:34] Settings for eth1: [17:34] Supported ports: [ TP ] [17:34] Supported link modes: 10baseT/Half 10baseT/Full [17:34] 100baseT/Half 100baseT/Full [17:34] 1000baseT/Full [17:34] Supported pause frame use: No [17:34] Supports auto-negotiation: Yes [17:34] Advertised link modes: 10baseT/Half 10baseT/Full [17:34] !pastebin | Richter [17:34] Richter: For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imagebin.org/?page=add | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [17:34] 100baseT/Half 100baseT/Full [17:34] 1000baseT/Full [17:34] Advertised pause frame use: No [17:35] !pastebinit [17:35] pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit -b http://paste.ubuntu.com [17:35] Pici: probably sufficient now :) [17:35] 13:35:08 RoyK, how i use pastebin? [17:35] 13:35:19 RoyK, its a irc command or pastebin the online service? [17:36] Pici: I just did a !pastebin to him a minute ago ;) [17:36] I saw [17:36] !pastebinit [17:36] pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit -b http://paste.ubuntu.com [17:37] * RoyK thinks a -q would be right now [17:37] RoyK: he's -q .. [17:37] ok [17:37] right - I misunderstood that [17:37] RoyK, you see the output? [17:37] or misread [17:38] Richter: can you pastebin that, please? [17:38] Richter: not all came in since you were quieted before the paste was finished [17:39] RoyK, can i post line by line? [17:39] RoyK, still lrning how to use irc xD [17:39] !pastebinit | Richter [17:39] Richter: pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit -b http://paste.ubuntu.com [17:40] http://pastebin.com/kpHWpGNL [17:40] Hmm, can you grow a single-drive btrfs setup into a multi-drive one? Or do you have to have created it with a chosen raid style in the first place? [17:40] RoyK, http://pastebin.com/kpHWpGNL [17:41] Richter: you don't have a working link [17:41] try with a straight cable [17:42] RoyK, this link work here... but i will call the datacenter [17:42] RoyK, txh [17:43] Richter: well, it doesn't work between the two machines [17:43] RoyK, ? [17:43] you don't have link, meaning MTU has nothing to do with it [17:44] RoyK, i look the gentoo server [17:44] RoyK, Link detected: no [17:44] bingo ;) [17:44] try another cable before doing anything else [17:45] RoyK, yeah [17:45] RoyK, thank you [17:45] np ;) [17:55] zul: i don't think so - sounds good [17:55] well, i suppose it might be worth looking through the git tree [17:56] Richter: any luck with a new cable? [17:57] RoyK, i call the datacenter [17:57] k [17:57] RoyK, nothing new [17:58] RoyK, they say that use new cables, but the problem continues [17:58] they'll have spare cables around [17:58] RoyK, maybe is the offboard dell card [17:58] RoyK, : / [17:58] RoyK, yes, they use a new one [17:58] RoyK, from the plastic bag [17:58] Richter: are both machines using gigE cards? [17:59] RoyK, yes [17:59] can you ask them to connect the machines to a switch? [17:59] it'll be easier to debug that way [17:59] RoyK, i will ask to connect right to other computes [17:59] computer* [17:59] they can just setup a separate VLAN for the two as to separate them from the rest of the network [18:00] RoyK, they are lazy, [18:00] RoyK, haushaushaush [18:00] RoyK, i am going to the datacenter see the problem [18:01] RoyK, thank you, and sorry for my bad english [19:19] is there a way to check the "host" or "device" status of a given USB? === Ursinha is now known as Ursinha-afk [19:43] I'm trying to do a hands off install of ubuntu 12.04 with preseeding. the preseed works fine except that it hangs for about 10 minutes on `anna net-retriever default`. I tcpdumped and during that whole time the thing is just sitting there trying to RARP itself. Any suggestions? [19:44] xarph: out of curiosity, is that prior to the net console even starting? [19:44] because I'm having a similar delay but only of about 2-5 minutes [19:45] but haven't been able to debug because it's prior to net console [19:45] I'm not using a network console. I also had a short delay earlier because it was trying to hit security.ubuntu.com and our pxe vlan doesn't have internet access. I fixed that and just got a larger delay :( [19:46] this is how I got rid of that one: [19:46] d-i apt-setup/services-select multiselect security [19:46] d-i apt-setup/security_host string repo-01.mydomain.com [19:46] d-i apt-setup/security_path string /ubuntu/12.04-x86_64 [19:47] oh, I don't use network console either, only when I need to inspect the running install [19:48] I just bump over to a vtty [19:48] though that was its own adventure if you're going to esxi via an rdp session as I fount out on friday --; === Ursinha-afk is now known as Ursinha [20:00] rbasak, around ? === lj1 is now known as lj [20:05] I have some configs in /etc/network/interfaces.d/vips.conf that don't seem to be loaded on reboot. I'm running 12.04 LTS. Any advice? [20:14] hi all [20:15] booted a LiveCD of Ubuntu Desktop onn an IBM X5630 server with a Debian installation that just failed [20:15] I want to backup its data [20:15] so I guess I have to mount the discs [20:18] yay! it worked [20:21] izanagisan: how did it fail? [20:29] Hi! I'm trying to setup my isc-dhcp-server to serve these following subnets: 10.0.10.0/24, 10.0.20.0/24, 10.0.30.0/24 and 10.0.40.0/24. But for security reasons we want to run it on the ip address 192.168.0.2 which is another subnet, no problem! We use intervlan routing and DHCP Relay so no worry. But i ran into the error that it couldn't listen onn any interface that it doesn't have a subnet declaration for... Why? Anyway to [20:35] Anyone got 2 mins to help me with a small bind issue [20:43] halvors: add an empty definition for 192.168.0.2 [20:45] That actually works? [21:11] zul, hows it going with keystone + neutron? [21:11] Does it work to just declare a dummy subnet? [21:26] adam_g: keystone is nearly done neutron as well [21:26] cool === thesheff17_ is now known as thesheff17 [21:50] adam_g: just nagging people about oauth2 [21:59] zul, whats up with oauth2? [21:59] adam_g: its blocked on a security review [21:59] and the MIR guys want us to use oauthlib as well [21:59] zul, oh, is a keystone dependency now or something? [21:59] adam_g: yep [22:39] halvors: yes