[00:16] is there a reason why solr-jetty doesn't depend on openjdk-6-jdk when it won't run without it? [00:16] I don't see any suggests, either.. [00:22] elliotf: what does it depend on? [00:22] elliotf: oh and solr-common depends on openjdk-6-jre-headless | java5-runtime-headless | java6-runtime-headless ... [00:23] elliotf: I suspect it doesn't need the full jdk.. but maybe it does? [00:23] * SpamapS just realized he's terribly late and cannot stay to discuss [00:24] SpamapS: it seems to.. [00:24] SpamapS: it spits out a 500 error complaining it can't find javac [00:25] until you install openjdk-6-jdk... then it shuts up and does it job [00:25] SpamapS: something to do with .jsp files or somesuch [00:25] * elliotf is not a java guy... yet [01:31] Ubuntu 10.04 (Linode) - Postfix/Dovecot Issue - Unable to receive emails - When ever i send a email to my mail server, the logs show that the email was received and no errors where found, when i go to check the account, no emails are in the inbox at all - main.cf http://pastebin.com/ARjhCSFC / master.cf http://pastebin.com/KHLRZt4A / dovecot.conf http://pastebin.com/hfKgB6VW / mail.log [01:31] excerpt: http://pastebin.com/ALqcw0Xb [01:32] JRWR: what MDA? [01:33] Well.. that is a good question [01:33] I dont know... [01:33] Looks like you're using postfix's built-in MDA [01:33] Does /home/vmail/%d/%n exist? [01:33] yes [01:34] Does it have appropriate owner/group/permissions [01:34] yep [01:34] Does the user have a .forward or .procmailrc? Do the logs say anything relevant? [01:34] * twb reads mail.log paste [01:35] I did find out where the mail was going... into Maildir into my home dir [01:35] Looks like you're using amavis as the MDA [01:36] I don't see where you tell postfix to do that, but the logs imply it [01:36] it should be in the master.cf [01:36] thats where i set the content filter at [01:36] Oh right [01:40] twb: here is the full output of postconf http://pastebin.com/SEGfSQGi [01:43] try adding a -n on that [01:43] postconf -n [01:44] http://pastebin.com/Y9KfkhZU === vDubG is now known as vDubG__Gone [01:47] diff -U0 <(postconf -d) <(postconf) | sed /^@/d [01:47] ...that will show only the options you (or Ubuntu) have changed from the defaults, what they are, and what they were. [01:47] why don't you try fixing those warnings first? [01:47] twb, isn't that what -n is for :) [01:48] patdk-lap: that doesn't show you what they were [01:48] twb: already fixed the working [01:48] warning [01:48] twb, heh? [01:49] you removed your domains from mydestination? [01:49] patdk-lap: never mind, man [01:49] patdk-lap: I prefer doin' in my way, but I don't care enough to argue about it [01:54] ok I got the server working [01:55] for the logs, i had survivorzero.com in the mydestination as well as in virtualhost database as well.. this seemed to cause a conflit [01:55] and now has been resloved [01:55] Thanks for the help guys [02:05] Hooray [02:07] atlest i didnt do a I fixed it, and then left [02:07] I know this channel has to be logged :) [02:07] who knows what some poor soul might be goolging for [02:27] hey can you use a reverse break out cable either way? [02:27] Doonz: what is a reverse break out cable [02:28] usually a sff-8087 connect to 4 discrete sata/sas connection [02:28] im wondering if you can go from the controller 4 discrete sata to 1 sff-8087 port on a back plane [02:30] no [02:30] they can only go one way [02:31] it's like the only cable you can't use backwards [02:31] Oh, SAS. Fuck that shit, man. I'm glad I skipped the whole SCSI termination era [02:32] heh, sas is simple, the numbers just lie though [02:32] patdk-lap: sorry your wrong just received my email from lsi [02:32] hmm? [02:32] they work apparently [02:33] hmm, I know there is one cable that doesn't [02:33] can't seem to locate the nice document on it right now :( [02:33] dunno but they sent me the link [02:33] http://www.ncix.com/products/?sku=48798&vpn=C-SFF8087-4S&manufacture=Norco%20Technologies%20Inc%2E [02:34] this one is what i need since i got a raid card with 4 discrete sata/sas ports on it but its going in a case that has a 4port backplane [02:34] so that cable will allow me to go card to back plane no problem [02:34] the case is a norco 4220 [02:34] ya, that will work [02:34] but you can't just flip it around and use it backwards [02:35] you can't go from a motherboard with a sff8087 to 4 sata drives [02:35] hmm, 42MB/second going over Broadcom gigabit, from ZFS. [02:35] according to lsi you can [02:35] that is why they make sff8087 -> 4sata, and sff8087 -> 4sata reverse [02:35] What's likely the limit: zfsonlinux, or the Broadcom card? [02:35] Broadcom 5723, specifically. [02:35] DanaG: THOSE ARE GOOD RATES [02:36] LOL [02:36] dammit caps [02:36] That's from that server to my laptop, with an Intel 82567L. [02:36] what switch in the middle [02:36] Whatever's built into a WRT350N. [02:37] I've tested from SSD to SSD, and gotten 80MB/second. [02:37] And that was with 100% CPU usage on the Pentium M with Broadcom ethernet in Windows. [02:37] thats asus or dlink [02:38] whats a direct link between server and laptop get? [02:38] That's Linksys. [02:38] Hmm, haven't tried a direct link. [02:38] Oh yeah, and iperf showed... I forgot what, but it was pretty high. [02:39] try direct and just see if there is a difference [02:39] Easier thing to do, for now: try the Intel card. [02:40] yeah i would put money on the switch not being able to handle full load [02:40] Interesting: the Intel is getting 52. [02:40] Now I'll try direct. [02:42] oh, the sff8087 is fine, it's the sff8088 that you can't reverse [02:43] how strange [02:47] Hmm, direct with Broadcom is 47MB/second. [02:48] I'll try direct with Intel. [02:48] And notably, the source partition is zfsonlinux. [02:48] Since btrfs isn't yet usable, in terms of having a fsck. [02:48] As soon as I can get btrfs, I'll switch. [02:48] Er, as soon as it's more useful. [02:50] wait, what's wrong with the btrfs fsck? [02:50] Oh, I see. the fact it doesn't _have_ one [02:54] Interesting... I did get a kernel oops from e1000e. [02:54] And I'm seeing that only via the remote access card. [02:54] All networking is dead. === ckrailo_ is now known as ckrailo [03:01] weird... from "booting a command list", to the first kernel log messages, it takes a good 30 seconds of blank screen with blinking text cursor. === Datz_ is now known as Datz [03:07] WOOHOO! Linux 3.0 officially pushed [03:21] Okay, so, link-local from Intel to Intel: BSODs my laptop. [03:22] And with laptop in Ubuntu, I'm getting 26 MB/second. [03:22] That's lower than what I got with Broadcom in Windows. [03:22] Actually, it's probably limited by ZFS, mostly. [03:23] Now to try link-local on Broadcom. [03:23] nah, I've actually had enough experimenting for now. [03:23] Is it expected for a Broadcom card to outperform an Intel card? [03:26] weird [03:27] I'm going to try some iperf again. [03:28] Even while that Intel link-local transfer is going on, I'm getting 930 megabits on iperf to Intel. [03:29] ah, yeah, the file transfer does seem to be limited by ZFS. [03:29] Come on, btfsck... we need you! [03:29] Okay, why the _heck_ is AoE so .. bloody .. slow on ARM? [03:29] Is this all CPU bound? [03:29] Age of Empires? [03:29] ATA over Ethernet [03:29] iSCSI hurts a bit too [03:30] Okay, iperf with broadcom on switch: 84 megabits. [03:30] Now to try Intel on switch and Broadcom direct. === ckrailo_ is now known as ckrailo [04:55] hello. i'm looking for help running ubuntu server 10.04 lts from a thumb drive (the actual os installed on the thumb drive and booting to it) [05:21] idlemind324, what about using the live ISO (& convert isolinux to syslinux)? [05:22] live iso is only on the desktop version [05:22] (i think) [05:25] idlemind324, yes - forgot about that [05:26] i'm trying a tool called unetbootin now [05:26] idlemind324: good tool [05:27] we'll see if it does the trick =) [05:27] unetbootin is stinky and crap [05:27] http://archive.ubuntu.com/ubuntu/dists/lucid/main/installer-amd64/current/images/netboot/ [05:27] Just use boot.img there === DanaG_ is now known as DanaG [05:30] twb forgive me i just copy the boot.img file to my thumb drive and leave it completely blank then? [05:31] gunzip it, then cat boot.img >/dev/uba, where uba is your USB key [05:31] Note *not* a partition on the key, e.g. uba1 [05:32] twb: sudo cat boot.img > /dev/sdb [05:33] twb: gives me "bash: /dev/sdb: Permission denied" [05:33] idlemind324: the redirect is happening in your normal user shell... [05:34] idlemind324: cat boot.img |sudo tee /dev/sdb >/dev/null # will probably work [05:34] Or just sudo -s and then type the cat command in a root shell. [05:35] The first ommand gives me a no medium found error [05:35] adding the -s in front of the original command still gives me a permission denied error [05:35] sudo -i is better. [05:35] Are you sure the USB drive is /dev/sdb ? [05:35] yes [05:35] Though then you'll need the full path. [05:35] DanaG: Explain how it makes a difference to this command? [05:36] I usually use /dev/disk/by-id/ to be absolutely sure it's USB. [05:36] Yeah, -i doesn't matter here. [05:36] jmarsden: UUOC [05:36] DanaG: I usually check /proc/partitions [05:38] http://pastebin.com/rH4E3qz2 <- ls of my /dev/disk/by-id/ folder [05:39] twb: Guilty as charged... sudo tee /dev/sdb /dev/null should work too. I just have the "something |sudo tee filename" approach in my fingers and didn't adapt for the "something" being cat. [05:41] well this is encouraging [05:41] sudo dd if=boot.img of=/dev/sdb went through [05:42] my system on the other side is loading up now off usb drive but it's just loading the netboot img [05:42] i'm going to pick this up in the am need some sleep thanks for getting me pointed in the right direction [06:21] New bug: #814436 in postfix (main) "package postfix 2.8.2-1ubuntu2.1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/814436 [07:27] can someone help me with this? http://www.howtogeek.com/howto/ubuntu/change-ubuntu-server-from-dhcp-to-a-static-ip-address/ [07:37] philipballew: sure - what are you having trouble with? [07:43] jamespage, haha, the whole thing [07:43] :) [07:44] and i need to figure out whats a good local ip to choose so i dont interfeare with my router [07:44] philipballew: well that would be a good start - sometime routers only give out DHCP IP addresses for specific ranges [07:44] or at least can be configured todo that [07:45] the router i have now is kinda lame [07:45] 2wire by att === smb` is now known as smb [07:47] it has a built in modem therefor i cany add my ddwrt routers [07:47] but thats irrelvent [07:51] jamespage, [07:51] philipballew: so most of the network configuration will be much the same as you get using DHCP [07:51] so the following commands might help you out with some settings [07:52] ok. ill try them. i just need to figure what to put for all the nu,mbers to fill in [07:52] ip route [07:52] 'default via XXXX' is what you need to set your default gateway to [07:53] that does in the gateway XXXX entry in /etc/network/interfaces [07:53] ip addr [07:54] philipballew: so are you using a wired or wireless connection? [07:54] wired [07:54] OK - so look at the information related to eth0 [07:54] you should see a 'inet XXXX' entry - which will give you the current DHCP IP address and the broadcast address for your network [07:55] http://pastebin.ubuntu.com/649772/ [07:56] so 'gateway 192.168.1.254' is good then [07:56] http://pastebin.ubuntu.com/649773/ [07:56] alright ill construct this file on my laptop now [07:57] that should give you most things - netmask 255.255.255.0 [07:57] network 192.168.1.0 [07:57] broadcast 192.168.1.255 [07:57] infact pretty much as in the article - aside from the gateway IP addresss [07:59] do you think their ip addres is safe? [07:59] oh - and take a look in /etc/resolv.conf - it should contain the information provided by your router and that needs to stay the same [07:59] so with regards to which ip address - its hard to say as everyone's network has different stuff plugged into it [08:00] running this command will tell you roughly what is plugged into your network: nmap -v -sP 192.168.1.0/24 [08:00] well my laptop now is 192.168.1.66 [08:01] here we go http://pastebin.ubuntu.com/649775/ [08:04] philipballew: I think you have the same laptop as me :-) [08:04] how do you like it? [08:04] :) [08:05] philipballew: works well TBH [08:05] had it just over a year now [08:05] do you have the backlight issue? [08:05] i used the backlight ppa [08:05] hrm - not that I am aware of [08:05] I had some sound issues when I first upgraded to natty - but that is resolved now [08:06] overall this is pretty stable [08:06] ubuntu and this get along [08:07] any idean on the ip from your end? [08:08] looking at your nmap scan picking something high (like .200) would prob be safe [08:09] or it may be that your router gives out addresses from .65 onwards - so something less that that might work as well [08:09] i can try 200 for now [08:10] set and ready to save the file [08:11] * philipballew crosses his fingers and presses ctrl + x [08:12] philipballew: hey - you can always put it back again [08:13] now to the sudo vi /etc/resolv.conf part [08:13] i have no idea what to do here [08:16] what does it look like ATM? [08:16] http://pastebin.ubuntu.com/649782/ [08:19] leave it as is - should be fine [08:20] alright. ill restart networking now [08:20] see if ssh stays connected [08:21] it would probably drop me and id have to re connect [08:22] it works! [09:13] ok i made a big mistake and screwed up all my permissions [09:13] is there a way i can revert? [09:13] like go back in time haha [09:13] goddard: this is called "an object lesson in the value of backups" [09:14] twb so im screwed [09:21] hello all ubuntu geeks. I have a question [09:25] i have some backups [09:26] just an aptitude backup i think though [09:46] hi [09:49] I would like to know enter proxy entries for server [09:52] I would like to know how to enter proxy entries for server [10:02] hello [10:05] can you hear me? Hello, is it late there ? [10:18] hello === zz_ng_ is now known as ng_ [10:24] is there graphical software that helps manage a server === ng_ is now known as zz_ng_ [10:26] New bug: #814536 in openssh (main) "package openssh-client 1:5.3p1-3ubuntu7 failed to install/upgrade: subprocess installed post-installation script returned error exit status 2" [Undecided,New] https://launchpad.net/bugs/814536 [10:45] helo [10:46] how to shut down from a normal user ? [10:50] Tommy_nmw: `shutdown now` [10:50] EricJ: that is only working for root [10:51] Tommy_nmw: Have a look at policykit. [10:51] Pretty sure you can set more finely-grained permissions there. [10:51] EricJ: how to ? [10:52] I think I'm using policykit to give shutdown/reboot-permissions to my non-root default user on my HTPC. [10:53] EricJ: bro, I am just using on plain CLI command mode server . so how can I get policy kit? [11:05] EricJ: polkit on a server? EW> [11:05] setpcap, sudo or chiark-really [11:05] twb: so what would you do? :) [11:06] I don't want a bloody XML IPC framework in my privilege escalation code, thankyouverymuch :-/ [11:07] The main reason I use even sudo at work is because I can replace sudoers with LDAP objects. [11:07] But that's because I have dozens of boxes [11:08] I'm not following. You'd simply make your users sudoers instead of use polkit? [11:09] Well, for example, my users want to be able to use the burner in the CD burning machine [11:10] http://paste.debian.net/123757/ [11:10] Now they can "sudo -g adm cdrecord foo.iso" but not do anything else with escalated privileges. [11:10] Interesting. [11:10] Now, that's pretty damn complex [11:11] Especially for security infrastructure, which should be as simple as feasible [11:11] And polkit is WAY more complex than that, simply so that users can do "sudo -g adm" from within a GUI [11:11] yeah, polkit is sort of a mess... [11:11] That is, so that the GUI just request privileges from the privilege-escalation daemon without having to create pop ups and things [11:12] Now, if you look at chiark-really, that's basically what people mostly use sudo for: "give fred and sally full root access" [11:12] And (hopefully) really is substantially simpler because it doesn't try to be anything else, which makes it easier to audit and less likely to contain problems in the first place (less code). [11:13] OTOH, because sudo is so widespread, HOPEFULLY people (in general) audit it more... [11:14] Taking a totally different tangent, I'd prefer to see some kind of RBAC infrastructure a la krb [11:15] Allegedly grSec can do that, but I haven't cared enough to roll patched kernels [11:33] SpamapS: yo [11:46] New bug: #814569 in augeas (main) "Segmentation fault in augtool and augeas.rb" [Undecided,New] https://launchpad.net/bugs/814569 [13:09] good morning [13:15] I used apt-get install linux-headers-$(uname -r) to install kernel headers, vbox Guest Additions still complains that the correct kernel headers are not installed. What am I doing wrong? [13:15] pythonirc1012: do the linux-headers match your installed and running kernel version? [13:16] joschi: how do i check? [13:16] Why doesn't vbox have a dkms by now [13:16] useless oracle schmucks [13:16] twb: they do i thought? [13:16] pythonirc1012: then use that instead [13:17] twb: it should, but only for the OSE -> virtualbox-ose-dkms [13:18] twb: and virtualbox-ose-guest-dkms respectively [13:18] Well, anyone running proprietary software deserves to be given the runaround [13:18] was missing dkms, that fixed it! [13:18] was missing dkms, once i installed it, things worked...if dkms is the problem, why print out as "headers dont match error"? [13:20] pythonirc1012: dkms is magic that takes care of the whole problem for you [13:36] smoser: ping [13:37] here [13:38] smoser: http://pastebin.ubuntu.com/649965/ [13:38] smoser: include-once implementation :] [13:40] lynxman, hm.. [13:40] smoser: shoot [13:40] you can't write it to /tmp [13:40] as that wont be there on reboot [13:40] smoser: which path would you recommend then? [13:40] need to cache to /var/lib/cloud/instance/ [13:40] smoser: cool [13:40] but you dont have that path in UserDataHandler easily i dont htink [13:40] you'll have to pass it in or something. [13:41] oh, and please update documentation also for 'include-once' [13:41] smoser: I will, just wanted to double check if you were happy with this implementation before continuing [13:42] and also add a mime-type [13:42] yeah, it seems reasonable [13:42] smoser: You think an include-once mime-type would be beneficial? How? [13:42] smoser: I kinda discarded it in my mind, that's why I'm asking [13:43] ah... yeah, the one for #include would work fine for it. [13:43] is that what you were saying? [13:44] smoser: Yeah but the #include mime-type will not include-once ever [13:44] smoser: I think that for completeness it should be there indeed... [13:44] smoser: Thought of this feature as a one off inside a script [13:46] <_johnny> hi. for compatability issues i "need" to run fuse zfs on a disk(array) on a ubuntu server (it's temp, so don't go preaching about the issues with that setup, heh). in relations to GELI on freebsd, what would an "equivalent" be in ubuntu? crypto? LVM is not a good option is it? [13:46] smoser: will do, the implementation is actually just two lines === med_out is now known as medberry [14:21] all right, it just turned pitch black outside, so if i stop answering, power msut have gone out :) [14:22] hallyn, tornado? [14:22] Here, it was the Telstra tech [14:22] ppetraki: might be [14:26] hallyn: if you live close to a girl and a little dog, move. quickly. [14:26] raubvogel: he said it was black, not black and white [14:26] can i save myself by kicking the dog out? :) [14:27] lol [14:36] nagios [14:36] !nagios [14:36] anyone here use it..and if so got any good front end management suggestions? [14:37] I got someone here at work that is interested in it but..well are totally frightened by the cli aspect of adding monitored clients to it. [14:37] grendal_prime I think the nagios website lists a couple of good ones [14:37] He wants to use it to monitor the vmware servers. [14:38] ok.. [14:39] At any rate, look here for a list of GUI's. for larger deployments I stongly suggest looking at Centurion. http://www.ubuntugeek.com/nagios-configuration-tools-web-frontends-or-gui.html [14:39] erm, centreon* [14:43] thanks man [14:45] hey in apt how would i go about enableing automatic security updates? [14:46] I know it askes on system build..but I got a tone of vm's on this thing i need to update..im thinking it would be better for these to just update themselfs. they are just for testing anyway. [14:47] im looking around in /etc/apt but i dont see anything about automatic update [14:47] https://help.ubuntu.com/10.04/serverguide/C/automatic-updates.html [14:50] hey :D [15:07] what is the easiest way to install tomcat7 on ubuntu-server? [15:08] hallyn, https://bugs.launchpad.net/bugs/814222, you could very easily reproduce on amazon. m1.large will give you 2x440G drives for $0.32 [15:08] Launchpad bug 814222 in qemu-kvm "kvm cannot use vhd files over 127GB" [Low,New] [15:10] smoser: really? cool, i'll try that! [15:11] hallyn, launch an instance store instance, oterwise you have to explicitly state that you want the additional storage [15:11] http://ubuntu-smoser.blogspot.com/2011/02/getting-ephemeral-devices-on-ebs-images.html [15:13] smb/hallyn: i just uploaded xen 4.1.1 [15:13] smoser: thx. I've got two of those sitting aroudn persistently for kernel testing, following your instructions :) [15:14] zul, Cool. I likely will jump on it next week [15:14] zul: that's for domU support? [15:15] hallyn, dom0/dom* [15:15] hm [15:15] er...domU [15:15] its basically the hypervisor and userland tools [15:15] And should be for HVM working [15:17] zul, Did you happen to see/fix the things from other notes too? Like /usr/share/qemu-linaro for keymaps and /var/lib/xen for hvm [15:18] smb: yeah im slowly getting there [15:18] zul, Cool. No hurries, things just happened to get spread a fair bit... [15:19] i noticed ;) [15:19] what are the changes between xen 4.0.1 and 4.1.1? [15:20] Dori922, go have a look at the changelog [15:20] nobody got a quick summary off the top of their heads? ;P [15:25] Dori922 : There are enough changes that 'off the top of their heads' isn't really efficient [15:25] ah awesome :D worth upgrading so :D [16:38] Alright I think I've tracked down my server issue to either the file system or the raid array. [16:39] Basically anything that reads from the filesystem can often wind up blocking for 30+ seconds [16:39] even a simple 'tab+tab' for command line auto complete. [16:43] what kind of raid array? [16:43] cause that sounds like an sata disk going bad [16:43] a raid shouldn't do that [16:48] On 10.04LTS I've always used -virtual kernels with KVM. Now in a VM, grub is ignoring linux-image-2.6.38-10-virtual claiming it is Xen only. Is grub on glue? [16:49] na, grub is right, that is a xen kernel [16:50] sounds like you need a newer grub === ckrailo_ is now known as ckrailo [16:55] hmm odd the VM host has grub-pc=1.98-1ubuntu12 and grub-common=1.98-1ubuntu12 but the VMs have grub=0.97-29ubuntu60 and grub-common=1.98-1ubuntu12. [16:58] Looks like this is fixed in Grub 0.97-29ubuntu61. I just have to find where that is hidden... It's urgency should be upped since it's absence is blocking kernel updates... [16:59] guess you hadn't updated in a long time [16:59] I think 10.4.2 got grub 1.98 [17:00] No I keep those boxes current. LTS supports the 0.97 series. 1.98 migration isn't always straight forward. [17:01] on my lucids, 10.04.3, both grub-common and grub-pc are 1.98-1ubuntu12 [17:02] hmm, same version as my oldest lucid box [17:04] I'll back port 0.97-29ubuntu61 from Natty, but I'm surprised it isn't already there... [17:05] smoser: alright, ensemble/orchestra fully bootstraps and deploys a machine! [17:05] smoser: testing with your devenv [17:05] woowoo [17:06] that is awesome. [17:06] i thought i might talk about that in my cloud week session next week for kim0|vacation [17:06] smoser: hehehe that's my talk :) [17:06] smoser: i though about talking about that too [17:06] well carp [17:07] how are you going to talk about it? [17:07] i wanted to demo, but didn't know how you could really do that in irc classroom [17:07] smoser: let's do something, let's put our two sessions together [17:07] smoser: and do a 1 two hour session [17:07] smoser: to demonstrate that [17:07] smoser: cause the dev env will take a while [17:10] RoAkSoAx, ok.. sounds good to me. [17:11] i can talk about devenv and you can talk about orchestra-ensemble if you'd like [17:11] smoser: ok cool [17:12] RoAkSoAx, instructions? === klaas_ is now known as klaas [17:14] RoAkSoAx: woot woot [17:14] RoAkSoAx: reading backscroll. :) [17:14]