=== koolhead|H is now known as koolhead17 === slank_away is now known as slank === ariel__ is now known as Guest77318 [00:43] can sombody tell me if lamp supports ajax [00:43] Guest77318: it does if you write your application that way... [00:44] do you know javascript [01:16] hello! [01:16] I'm running ubuntu server 12.04.1. I just moved my hardware to another box (new mobo and cpu, same hdd's and same install) and now it won't find the network. [01:16] I have the server set to DHCP with an IP reservation from the router. Eth0 is failing to come up and I'm not enturely sure why. [01:17] interfaces config file looks normal... auto eth0 iface eth0 inet dhcp [01:17] ifconfig only brings up lo, but lspci shows the realtek gigabit onboard ethernet in the listing. [01:17] roasted, /etc/udev/rules.d/70-persistent-net.rules (edit or delete) [01:18] the old NIC macs are persisted [01:19] PryMar56: thank you! I knew there was something I was forgetting. [01:19] Oddly enough I knew it had a 7 in it for some reason... I just forgot... everything else :/ [01:19] PryMar56: we're up and running now. I owe ya a beer. [01:19] roasted, cheers [01:20] :D [01:33] howdy === slank is now known as slank_away === Koheleth is now known as gimp4ever === gimp4ever is now known as koheleth === smb` is now known as smb === yofel_ is now known as yofel [09:31] Today's server installations fail because linux-headers-server have unmet dependencies (linux-headers-generic (3.8.0.0.13) [09:32] who should i report the bug against? [09:32] Daviey: jamespage ^^ [09:32] psivaa, linux kernel please [09:32] its probably transitory [09:35] jamespage: ahh ok thanks. normally the mismatch in d-i installer and the archive give a different error message, i'll confirm with release or kernel then [09:39] jamespage / psivaa: You'd hope that with the proposed usage now, we'd not get this! [09:43] Daviey: thanks, ill report a bug anyway and see how it goes [09:45] psivaa: could i ask you to let me know how it turns out please? [09:48] Daviey: infinity is fixing the 'proposed migration' and then the images will have to be respun . the issue seems to be in the desktop images too [09:49] psivaa: yeah, i just caught it in -release.. thanks :) [09:49] appreciate you following it up [10:07] Daviey, I uploaded hp3parclient to NEW in raring yesterday if you have time to review and accept [10:07] yolanda2, ^^ fyi [10:08] jamespage, so we could add it as a dependency when it's done? [10:08] yolanda2, yes - but lets hold off until after g2 is out in the distro [10:08] yolanda2, it will need to go through MIR [10:10] jamespage: ok, chuck asked me to do 2 others first... i'll do them today for sure [10:10] Daviey, great - thanks [10:30] Hello [10:30] I have a server (Ubuntu 10.04 LTS) [10:30] which will not boot, it drops to busybox with the line "Could not open /dev/mapper/-root" [10:31] however, the block file exists, and I can mount it and move around the filesystem.. (it's definitely my filesystem, not initramfs) [10:31] I assumed it was pvscan taking a while, so I've tried adding rootdelay=60 [10:31] but that's not working either. [10:32] after it goes to busy box; a ^D will take me to another screen "Waiting for device / press S to skip or M to manually mount the device" [10:32] pressing "M" takes me to a shell that's mounted in ro mode. [10:33] I can remount,rw and get networking up. [10:33] but the machine will not boot into init 5 [10:47] http://superuser.com/questions/256856/ubuntu-alert-dev-mapper-localhost-root-does-not-exist-dropping-to-a-shell [10:59] Hi, I've setup samba on a remote Ubuntu server but when our Windows server tries to connect it says that the password is incorrect? Connecting to \\ip\backups as user 'backups' who is in write users in the samba.cfg. [10:59] root works fine. [11:04] I have a complete disk image (dd if=/dev/sda of=foo.bin) that contains several partitions and LVM logical volumes on which there are ext4 filesystems. How do I mount one of the filesystems on this diskimage? I assume I have to use a loopback device somehow, but I am not sure how to do it when I have the entire device as an image and not just one filesystem. [11:11] tdn: look into kpartx, you may need to use it recursively to get into lvm. [11:12] (if your lvm volumes also have partition table) [11:12] kpartx -a path/to/foo.bin; vgchange -ay should do it, fi you don't [11:13] Hi, I have a user 'backups' trying to access a samba share \\ip\backups from a Windows machine but can't get authorized. It's using the correct password and is in write users for that share. root is too and that works. [11:24] xnox, ok [11:46] I found another sysadmin with my problem. [11:46] update, reboot -> no longer boots [11:48] I'm going to try updating to a newer version of the OS using do-release-upgrade [12:15] i am planning to upgrade/install (from source ) libvirt version on ubuntu 12.04.1 from 9.13 to version 1.0.x? willl that create a problem ? [12:49] so my install is now almost completely broken [12:49] I'm going to reinstall using centos. [12:49] after copying all my VM's off. [13:23] Hi all! [13:24] RoyK, sarnold, my server "survived" his first attack xD [13:24] :) [13:24] the server was unaccessible, or ultra slow, but alive, and once the attack stopped, was still there! [13:25] CSF bloqued a ton of IPs [13:25] blocked* [13:25] no messages of "dst cache overflow" in the logs [13:29] I think i'm going to reduce the synflood parameters in CSF [13:33] http://i.imgur.com/m1oly.png [13:34] still a lot of very similar IPs. It would be great if the firewall could block the entire range when 3 or more similar IPs are already blocked === slank_away is now known as slank === slank is now known as slank_away === slank_away is now known as slank [13:55] jamespage or yolanda2: https://code.launchpad.net/~zulcss/cinder/cinder-g2/+merge/142900 [14:01] zul, approved === motmot is now known as liva === cpg is now known as cpg|away [14:20] jamespage: https://code.launchpad.net/~zulcss/glance/glance-g2/+merge/142906 [14:29] zul, approved === chuck_ is now known as zul === cpg|away is now known as cpg === cpg is now known as Guest53868 [14:43] jamespage: next https://code.launchpad.net/~zulcss/horizon/horizon-g2/+merge/142908 [14:55] jamespage: next https://code.launchpad.net/~zulcss/keystone/keystnoe-g2/+merge/142911 [15:08] jamespage: i thought you added alembic to quantum? [15:10] I did [15:12] jamespage: nm i suck [15:24] jamespage: https://code.launchpad.net/~zulcss/quantum/quantum-g2/+merge/142916 [15:32] zul, all approved [15:32] jamespage: thanks there is a couple of more to do...but im making my way through it [15:40] jamespage: https://code.launchpad.net/~zulcss/python-glanceclient/grizzly-g2/+merge/142919 [15:43] https://bugs.launchpad.net/ubuntu/+source/lvm2/+bug/596554 [15:43] Launchpad bug 596554 in udev "Unable to mount root LVM partition" [Undecided,Confirmed] === megha is now known as firewall [16:05] Hey, I've got an issue with tomcat taking more than 20 minutes to start on Ubuntu Server 12.10. I've written up the problem in detail here: http://serverfault.com/q/466190/45607 [16:06] Any suggestions? [16:10] thompcha: check logs and available resources (cpu, memory, disk) [16:24] Ahoy. [16:30] pmatulis: I'm trying to make the logs output more. [16:33] hm, I have an idea, but I don't know if it is possible or how can I implement it [16:34] capture example, from the last attack: http://i.imgur.com/m1oly.png [16:34] is it possible to count the number of SYN packets from each /24 IP block? [16:34] for example: [16:35] 192.132.209.0/24 -> 27 SYN packets [16:35] 105.75.136.0/24 -> 2 SYN packets [16:35] and so on [16:35] if a /24 block has more than X packets, it gets blocked [16:36] is it possible? [16:37] zul, Daviey: bah - openvswitch dkms package is not liking 3.8.0 kernel [16:37] jamespage: meh [16:40] zul, Daviey: I really think we should drop whoopsie from the kernel seed [16:40] server seed rather [16:40] jamespage: how come? [16:42] zul, it does *nothing* in server world AFAICT [16:42] jamespage: im ok with it [16:43] thompcha, tomcat7 does some key/crypto init on startup - if you have poor entropy on your system it may take some time [16:43] thompcha, installing haveged might help [16:49] thompcha, hmm - its not that "INFO: Creation of SecureRandom instance for session ID generation using [SHA1PRNG] took [164] milliseconds." [17:11] smb, still around? [17:15] TheBronx: woot! good work. :) [17:16] under attack again. I've posted a thread in CSF forums: http://forum.configserver.com/viewtopic.php?f=6&t=6134&p=18807 [17:17] that feature in CSF would be great I think. or a script that can do that [17:17] hmm? [17:17] synfloods are pretty well blocked by the kernel on its own, for me atleast [17:17] I got a huge one a few days ago [17:18] but if each different IP only sends 1 SYN packet [17:18] but beyond that, you should have something doing watching and blocking, sure [17:18] both the kernel and the firewall (CSF) will do nothing [17:19] dunno about a csf thing [17:19] Is UFW not enough? [17:19] I use fail2ban and some scripts that upset ipset for banning stuff [17:19] ufw is static, not active [17:19] Fair enough, pardon my ignorance. [17:20] I was actually thinking about adding ipset support to ufw [17:20] he wants some kind of indrusion detecton thing [17:20] Is there a different between denyhosts and fail2ban? I'm under the impression that fail2ban is just more confurable [17:20] configurable [17:20] ipset kicks ass, except you need to install xtables [17:20] Crazy_Spai, same idea, ya [17:21] Figured as much, thank you. [17:22] CSF uses iptables, but it is easy to configure, for example for SYN flood attacks [17:22] you can set a limit, for example: 30/s SYN packets [17:23] but this limit is applied per IP [17:23] Couldn't you use another program to block those IP's [17:23] and what I want is to block the entire /24 block of IPs, as that is what the attackers are using [17:24] yeah, if know one that can do what I want, of course! [17:24] if you know* [17:24] Couldnt fail2ban monitor a log set by csf? [17:24] *I'm not too sure of anything, but from my understanding just trying to learn through brainstorming with you [17:24] fail2ban can monitor anything, and call any script [17:25] I dunno how you detect a single syn, as bad though [17:25] but fail2ban can easily grep the kernel iptables logs and ban ranges based on that [17:25] That's the route I was thinking was possible. [17:26] I guess if you log based on an INFO rule, it would get *LOTS* of info [17:27] but SYN packets are not logged, so fail2ban can't know what is happening [17:28] it is possible to use the CSF logs, so that when an IP is blocked for SYN flood, fail2ban can call a script that bans the entire range [17:28] sure [17:28] but again the problem is that CSF only blocks IPs that send more than 30 SYN packets per second (or 20 or whatever I set in the config) [17:29] Well, you have control over that frequency [17:29] personally, 30 syn per second is acceptable to my http servers [17:29] as that is normal [17:30] less than 30 is just paranoid hehe [17:30] as patdk-wk says, 30 syn is acceptable [17:30] the key is frequency per /24 block, not frequency per single IP [17:31] I personally allow 80/s per ip [17:31] that's the key, as the attack uses a lot of IPs from the same block [17:31] any my syncflood only come in at like 1000/sec [17:32] CSF is not blocking anyone. The logs are clean of banned IPs, until an attack starts [17:32] when an attack starts, CSF starts blocking IPs, that's what I see in the logs [17:32] so it seems 30/s is nice for my server [17:37] well, normal clients, will generate about 6 syn per webpage normally [17:37] unless you use multible hostnames, then 6 per hostname [17:38] I use 5 hostnames, so that gets me up to the 30 range per client [17:48] yeah, I only have one site (and a few small blogs and static sites) so 30/s is enough === alein is now known as a|way === a|way is now known as alein [18:07] http://pastebin.com/pdPNRMnH [18:09] dijit: apt-get install python-gnupginterface first ? [18:10] shouldn't I already have that. [18:10] isn't that like, a dependancy? [18:11] http://pastebin.com/PQBFvhH4 [18:11] one would hope.. [18:11] yeah this is very broken now. [18:11] I tried to run an apt-get dist-upgrade before. [18:12] that might be why. [18:12] dijit: .. to upgrade between distributions? [18:12] well, what else is it for? [18:12] (I'm on 10.04) [18:12] this happened /after/ my issues earlier. [18:13] sadly, some larger changes are difficult to do entirely via dist-upgrade, which is why the do-release-upgrade (or whatever the name tis..) was introduced. [18:13] * dijit sighs [18:13] I wish I knew that. [18:14] anyway, this server is a lost cause, it's not booting with LVM anymore because, god knows why. [18:14] I'm just going to rebuild it as a centos vm host or something. [18:14] and I happened to fuck up apt anyway. [18:14] so, I think it's time. [18:14] xD [18:15] sometimes not solving a problem is easiest :) [18:20] can someone explain the purpose or point me towards some documentation covering the precise-server-cloudimg-amd64-loader ? [18:22] I get an qemu error when I try to use it with a libvirt direct kernel boot. [18:26] Also, I'm looking for an initramfs for the Precise cloud image but I don't see any listed on http://cloud-images.ubuntu.com/precise/current [18:26] am I missing something basic here? [19:06] jamespage: re: that keystone charm branch, looks like the swift stuff landed while i was working on it and just needed to be rebased against the upstream trunk. doing that now [19:14] adam_g, great! [19:14] I thought that might be the case [19:29] happy new year all. [19:29] I've just set up a new OS (12.04) from the mini.iso - all looks good, but I don't have any video-out (VGA or DV. I'm able to log in via SSH, but bupkis on the TTY's. headless setup (no X) === Ursinha_ is now known as Ursinha [20:10] is it possible to push my headless/CLI TTY's to all display outputs? [20:13] kieppie1: the console= kernel boot param may let you select something.. or you may need to use netconsole (though I don't think that allows for interacting..) [20:13] sarnold - cheers [20:14] found the issue to my previously stated problem: I have multiple graphics cards on this machine, & it was defaulting to an output that had no monitor connected [20:14] ah! [20:14] I've done that. :/ hehe. [20:14] silly really [20:15] it's going to be a headless resver, no X installed, so it's a bit silly that it's not simply throwing out TTY to all displays :/ [20:20] sarnold - I'm finding references to enabling the serial console (probably not a terrible idea), but not for VGA. got any refs please? [20:20] zul: pecan.. ./pecan/middleware/recursive.py is a different licence to the rest? [20:21] zul: http://people.canonical.com/~agandelman/g2_deps/ <- added, alembic, python-webob, python-testtools, python-testscenarios, testrepository [20:23] Daviey: erp [20:24] zul: I also don't really agree with keeping the experimental changelog entry, as it seems it hasn't been published there? [20:25] Daviey: ok ill have another look [20:25] kieppie1: darn. maybe video= is the better starting point? [20:26] sarnold - cheers. brub/kernel options does seem the way to go at least. I'll keep digging, thanks [20:26] zul: not a requirement for NEW'ing.. but.. is this intentional? "python setup.py test --functional || true" [20:27] Daviey: yeah i didnt spend alot of time with the testsuite [20:30] jamespage: python-hp3parclient looks good.. but i note it has a testsuite that will probably need enabling for MIR.. if we need it. [20:43] zul: python-ceilometer, should it be arch: any .. and there is boilerplate left in debian/copyright [20:44] doh [20:45] zul: ok, i'll reject it for now. [20:45] at least g2 is uploaded except for ceilometer [20:46] \o/ [20:47] zul: i'll reject python-pecan whislt you look into the licence, if it is indeed good.. it can be salvaged from the queue === Ursinha is now known as Ursinha-afk === alein is now known as a|way === Ursinha-afk is now known as Ursinha [21:08] hi [21:08] i need some help [21:09] i replaced the hdd enclosure for my server's external disk [21:09] and now during bootup [21:09] presumably on the fsck stage [21:09] it says HDIO_GET_IDENTITY failed for '/dev/sdb' invalid argument [21:10] I'm trying to add a repository on one of my Ubuntu 12.04 servers with the command: sudo add-apt-repository ppa:openbravo-isv/ppa but I get a "sudo: add-apt-repository: command not found" in response. I did this on another server with no issue. What am I missing? [21:11] subman: it's apt-add-repository not add-apt-repository [21:12] sarnold, Wow, didn't notice that at all! I was just copy/pasting instructions right from the OpenBravo website. [21:12] virt-manager is giving pain on raring.. I keep getting 'Error starting domain: internal error Failed to get user record for name '116': Connection refused', 116 is the uid of libvirt-qemu [21:13] can't start any vm [21:13] sarnold, but that gives the error: "sudo: apt-add-repository: command not found" [21:13] here http://i.imgur.com/SBICT.jpg [21:13] is my error [21:13] ^^^ [21:13] stiv2k: Error: "^^" is not a valid command. [21:13] subman: haha [21:14] subman: that's hilarious. I just typed "apt-" and thought you had it wrong.. [21:14] subman: lrwxrwxrwx 1 root root 18 Sep 28 07:06 /usr/bin/apt-add-repository -> add-apt-repository [21:14] subman: install the python-software-properties package :) [21:16] anyone [21:16] sarnold: ping [21:17] sarnold, why thank you! [21:18] stiv2k: sorry, I don't know much there, except that it _looks_ bad. [21:18] looks like I hit https://bugzilla.redhat.com/show_bug.cgi?id=883547 [21:18] tjaalton: Error: Could not parse XML returned by bugzilla.redhat.com: HTTP Error 404: Not Found [21:18] its not bad [21:19] im sure the disk is fine [21:19] it was JUST working [21:19] all i did was put a new enclosure [21:19] stiv2k: yeah, but your enclosure looks like it is giving results that disagree with the old enclosure about drive size [21:19] stiv2k: it feels like it steals a sector for its own use? or .. why would it report something different? [21:19] i dunno [21:28] Daviey, great - I have the test suite working as of this morning - I'll upload a fix early next week [21:41] is there a server and client like ubuntu one just for syncing files to your own server? [21:41] eutheria: rsync is neat.. [21:42] what sarnold said [21:42] sarnold, rsync requires me to remember to run it [21:42] not if you use a crontab [21:42] eutheria: indeed :) [21:42] eutheria: cron may help [21:42] ubuntu one syncs things for me [21:43] eutheria, there's no client/server system i'm aware of similar to ubuntu one that you can easily set up for your system, except for an rsync set up to run every so often via cron [21:43] fair enough [21:43] and to be honest... [21:43] deja-dup created me a useless backup [21:43] i have that running on 4 servers. [21:43] it would be keen to have an inotify-aware rsync daemon thingy... [21:44] agreed with sarnold [21:44] (piecing something together with imcron and rsync would be _painful_, don't do that.) [21:45] (heh) [21:45] i guess i could use rsync [21:46] maybe there is a task manager to remind me to backup [21:46] or you can just schedule the rsync with cron to run at a given time each day [21:46] or multiple times per day [21:46] it is a laptop [21:46] it might be off at that time [21:47] for a server sure [21:47] as i said, set it up for multiple times per day :p [21:47] or don't its your call [21:47] * TheLordOfTime returns to poking at nginx's source code [22:07] zul: dont suppose you've seen this one: https://launchpadlibrarian.net/128227169/buildlog_ubuntu-precise-i386.testrepository_0.0.11-0ubuntu5~cloud0_FAILEDTOBUILD.txt.gz [22:08] cannot set JAVA_HOME in /etc/enviroment with JAVA_HOME="/home/syncsys/bin/jdk1.7.0_10/" echo give me that but its not a correct java home or i need something else? === gary_poster is now known as gary_poster|away === syncsys is now known as Quest [22:52] please see this http://stackoverflow.com/questions/14287501/java-home-in-etc-enviroment-in-ubuntu any help? === motmot is now known as liva [23:22] Hi, I installed nagios3 through apt-get but on the tutorial that I'm looking at it shows that the web interface should be setup automatically to http://localhost/nagios, is this correct? It's not even listed as running through apache2 at the moment, should I cp the sample conf to /etc/apache2/sites-available and then a2enable it? [23:22] I've tried #nagios, but there doesn't seem to be many signs of life over there at the moment. [23:48] sw__: they'd probably tell you to find details on distribution packaging from your distro.. :) [23:48] sw__: I think your a2enable idea is good, give that a shot..