[01:11] is it worth to protect something else than ssh using fail2ban [01:11] ? [02:35] hey guys I got this error when I try to log in ssh Write failed: Broken pipe === CripperZ- is now known as cripperz [03:36] does someone has an idea on how to fix that error. I get it when I try to log in ssh /bin/bash: No such file or directory connection xx.xx.xx.xx closed [04:46] whats the recommended way to fix the *** glibc detected *** ... munmap_chunk(): invalid pointer: 0x00007fb46a278460 *** ? [04:47] ssh to remote host fails, apt-get update fails .. on 10.04.4 LTS [04:50] its failing since AUg 5 [04:50] Aug 5 [04:50] 2014-08-05 06:28:47 status installed libc6-dev 2.11.1-0ubuntu7.14 === MikaT_ is now known as MikaT [04:57] I'm back I disable the chroot jails. does someone has a link for chroot jail that also allow ssh an sftp login ? === arrrghhhAWAY is now known as arrrghhh [07:42] hi all [07:43] every week I want to backup for files (tar them) and delete them [07:43] my idea was to use cron to call a script to handle everything [07:43] I can easily specify which folder to backup, but while it's doing it, the folder changes [07:44] what would be the best way to delete only the files that I put into archive? [07:44] to delete only files where date created < date of script run? [07:45] or to somehow get a list from tar and delete those files? === ewook_ is now known as ewook === arrrghhh is now known as arrrghhhAWAY === JanC_ is now known as JanC [08:27] FrEaKmAn_: you can get a list of files from tar, or move the files to be backed into another dolder and do backong up from there === rvba` is now known as rvba [08:47] Morning all [08:48] o/ [08:48] I need help as to finding out whether or not I have fixed the issue of my smtp relay HELO-ing as localhost.local domain, as it doesn't receive andy emails and just sends how can I test this? [08:51] telnet [08:52] miceiken: according to CBL TELNET will not show the banner [08:53] hm [08:53] brb [10:52] Hi there, I have a problem with a cifs mount on ubuntu server 14.04: I want to add a samba share to /etc/fstab, however I get "mount error: cifs filesystem not supported by the system". I have cifs-utils installed. [10:58] gnaddel: have you tried it again following a reboot? [10:59] cwhy1: Yes, tried a reboot === CripperZ- is now known as cripperz [11:05] gnaddel: run 'dpkg -l cifs-utils' [11:05] what output do you get? [11:06] cwhy1: http://pastebin.com/HdcMYQ1i [11:11] the fstab entry is as follows: //winfs-uni.top.gwdg.de/MYUSER$ /home/localuser/GWDG-P/ cifs credentials=/home/localuser/.smbcredentials 0 0 [11:16] gnaddel, try adding cifs to /etc/modules for a test [11:17] and check in proc/filesystems if it is supported after loading the module [11:18] ok, rebooting [11:21] ogra_: Doesn't seem to work: http://pastebin.com/5M75kCdV [11:22] this is weird ... also that you dont have anything in /etc/modules by default [11:22] lp and rtc shoul still be in there i think [11:24] I certainly have lp and rtc; though why anyone needs lp these days... [11:24] well, you cant auto-load it for some reason ... so to use it it must be there ... even if only a fraction does actually use it [11:26] It's a pretty vanilla 14.04 server installation, the only things I added so far are fail2ban, auto-updates and some python stuff [11:26] and I don't think it would be a non-standard setup either (It's a VPS from 1&1, German hoster) [11:28] oh, you didnt tell its a VPS [11:29] might be that they use a kernel that doesnt even have cifs as module [11:31] They call it Ubuntu Server 14.04 LTS minimal system [11:43] well, check "uname -a" [11:44] (and also see if /lib/modules has the coressponding modules there) [12:37] arrrghhhAWAY: good morning. bets on if/when the upgrader goes live today? [12:39] jamespage, zul: keystone is ready for review - https://code.launchpad.net/~corey.bryant/keystone/2014.1.2/+merge/230290 [12:40] coreycb, ack === ^rcaskey is now known as nodejs [13:06] coreycb, working through your branches now btw [13:07] coreycb, I'll upload anything that is OK [13:07] jamespage, great, thanks [13:09] how can i tell preseed to delete all the existing lvm. delete everything. [13:24] coreycb, does our MRE cover trove yet? [13:24] jamespage, I'm not sure - where can I check? [13:25] coreycb, MRE wiki page [13:25] jamespage, looking === cripperz is now known as CripperZ- [13:27] jamespage, I don't see trove here - https://wiki.ubuntu.com/StableReleaseUpdates/MicroReleaseExceptions [13:27] coreycb, OK _ well drop that from the update then [13:27] jamespage, ok [13:27] coreycb, we don't have an effect test strategy [13:37] coreycb, horizon and ceilometer uploaded [13:41] coreycb, for future reference, if you need to sync in the archive version of the package to the branch, make sure you pull in the changelog entries in full [13:41] nova - > 1:2014.1.1-0ubuntu2 [13:41] for example [13:41] rather than amending your changelog entry to include that ones patch [13:41] jamespage, ok [13:41] jamespage, want me to update those? [13:42] coreycb, no worries - I've done neutron and nova [13:42] jamespage, ok [13:42] coreycb, if any other ones need updating please do [13:46] jamespage, looks like cinder's the only other one that needs the changelog fixed, so I'll go ahead and do that [13:49] coreycb, awesome [13:51] coreycb, zul, gnuoy, ++ : just in case you had not noticed but PPA builds are now much faster - scalingstack is taking care of them so they are running ontop of OpenStack under full KVM; queue is virtually 0 and they are faster than my SSD quad core.... [13:54] Go team IS! [13:54] jamespage, very nice! [13:54] jamespage, cinder is update [13:54] updated === Solution-X|AFK is now known as Solution-X [14:16] coreycb, glance and cinder uploaded [14:17] coreycb, ditto nova === CripperZ- is now known as cripperz [14:26] coreycb, having severe deja-vu - waiting for keystone to finish building! [14:26] jamespage, lol [14:28] jamespage, there's some sort of db issue with heat that I'm trying to figure out [14:28] coreycb, ack [14:41] coreycb, neutron uploaded [15:04] coreycb, and keystone [15:05] coreycb, the complete set is here - https://launchpad.net/~james-page/+archive/ubuntu/icehouse/+packages [15:05] as well [15:05] jamespage, ack, thanks a lot. just waiting on me for heat now. [15:05] coreycb, ack [15:10] anyone built GRE tunnels between two ubuntu servers behind NAT? [15:10] I set up a tunnel but not sure if they're are up or not [15:11] I do 'ip tunnel' and it's empty [15:11] so I guess they're not up, is this a good way of checking? [15:18] Hi all, I'm running ubuntu 14.04 but for some reason I can't manage to change DocumentRoot in apache 2 [15:18] chriys: Are you modifying the file for your site? [15:18] in /etc/apache2/sites-enabled/ ? [15:19] And if so, what isn't working? [15:20] jamespage, do we want trove here with it not being on the MRE list? https://launchpad.net/~james-page/+archive/ubuntu/icehouse/+packages [15:21] Pici: I want to change the location of the website. Instead of /var/www/ I want /home/myuser/html [15:26] coreycb, meh === cripperz is now known as CripperZ- [15:29] jamespage, maybe it doesn't much matter [15:33] anyone might know why my GRE tunnel between two ubuntu servers is not working? [15:40] coreycb, if you fancy some MIR practice after sorting out heat - https://bugs.launchpad.net/ubuntu/+source/python-pysnmp4/+bug/1349868 [15:40] Launchpad bug 1349868 in python-pysnmp4-mibs "[MIR] new build dependencies for ceilometer" [Undecided,Incomplete] [15:43] jamespage, sure I'll work on that [15:43] coreycb, thanks :-) === arrrghhhAWAY is now known as arrrghhh [16:28] Nivex, :/ I can try to ping the folks who were talking about it in -devel, I just hate sounding desperate. I just wonder what is needed, if they need help with additional testing or something... === Lcawte|Away is now known as Lcawte === Ursinha is now known as Ursinha-afk [16:42] Nivex, check the page now :) [16:42] http://changelogs.ubuntu.com/meta-release-lts [16:42] huzzah! [16:42] root@atlantis:~# do-release-upgrade -c [16:42] Checking for a new Ubuntu release [16:42] New release '14.04.1 LTS' available. [16:43] sounds like I get some lunch and start my upgrade [16:43] I pinged the guy in -devel and he was like "concidentally I just did that" hehe [16:43] except he spelled it correctly. coincidentally* [16:43] and coincidentally I had just hit reload when you said to :) [16:45] and all is right in the world of 'buntu === dosaboy_ is now known as dosaboy [17:22] and here we go! === Ursinha-afk is now known as Ursinha [17:48] HI is there a linux utility to re-read uefi boot efivars which are displayed by efibootmgr -v ? thx [18:04] jamespage: smoser: any objection to http://paste.ubuntu.com/8019264/ ? [18:05] hallyn, the string 'qemu-system-x86' is that relevant ? [18:05] qemu-system-x86.modprobe. ie, is that arch specific ? does this matter elsewhere ? i suspect not. [18:05] i have a system where "init" is extremely busy, using 100% cpu a lot of the time, is there a way i can find out what it's doing? [18:06] hallyn, the only suggestion i hav is: [18:07] rmmod kvm_intel && modprobe kvm_intel || true [18:07] smoser: yes, it is only on x86, however there isn't currently a postinst for qemu-system-x86, so i didn't want to add it. [18:07] that way you dont modprobe if rmmod failed [18:07] smoser: will do, thx [18:07] and maybe just comment that yes, this could fail. [18:07] smoser: if you think it's cleaner to add a x86 postinst, i'll do it, this just seemed less likely to be messe dup by the likes of me [18:07] if kvm was being used) [18:08] hello. I have a first boot script run in rc.local that in some cases asks a question, when it asks I see the question but I cannot see anything I type into the bash read.. does anybody know why? [18:08] good point - i was actually only thinking of in a container [18:08] smoser: feh, maybe i should check the value of /sys/modules/kvm_intel/parameters/nested [18:08] i think i'll add that check, to redue the amount of module load/unloading [18:10] yeah. that'd be good too [18:11] hallyn, isn't there some util to say : am-i-in-a-container ? [18:11] i thoguht there was one [18:11] smoser: there is, but i'd rather not have kvm depend on that, [18:12] bc there are other ways to fail in the same way [18:12] ok [18:13] hallyn, you didn't come up with any magic to allow me to use upstart in a non-network-namespaced container, di dyou ? [18:16] smoser: I'm afraid not. upstart is just not designed that way. I'd simply recommend bridging eth0 and putting hte container on that bridge [18:17] smoser: use systemd? :) [18:17] thanks [18:21] smoser: so this again is for running maas or the openstack network daemon in a container? [18:22] smoser: so i'm going with http://paste.ubuntu.com/8019402/ for qemu [18:22] yeah, thats waht it was. [18:22] your answer is really ok. [18:23] smoser: I really hate it [18:23] having non-network namespace seemd just so perfect. [18:23] as i could basically run the container as an "application". [18:23] and clean it easily. [18:23] but then abstract sockets have pained me since i was first taught about them 10 years ago ("yo umoron, you'r enot handling these") [18:25] to answer my own question stty can re-enable echo [18:38] Does everyone around here run a dns caching server on each of their servers? [18:42] toyotapie, I think everyone uses their servers in whatever way they see fit... if you have a question, feel free to ask :) [18:46] My server was slowed to a halt because the first dns server in resolv.conf was offline, I am looking for a solution where a local dns cacher will check if servers are alive and route dns requests accordingly, that way it eliminates the delays trying to reach backup servers. [18:47] toyotapie, pdns-recursor does this by default [18:48] others, I dunno [18:48] you can also modify your resolv.conf to help [18:48] timout and attempts [18:48] the idea there, set a really low timeout, but attempt multible times [18:49] so you request it from basically all your nameservers, and the 2nd or 3rd attempt will get the results [18:49] default is 5seconds [18:49] and use rotate [18:50] pdns++ [18:50] I still need to look into how unbound does it [18:51] pdns is different from pdns++ ? [18:54] toyotapie: heh, no, that's just a shorthand for "I also like pdns" :) [18:54] OK ;) [18:54] I use pdns for my ns servers, but not the recursive server. [18:55] Maybe I should use pdns as a recursive server, I was still using the old trusty bind for recursive lookups [18:55] is it a good idea to setup a single pdns instance as a public authoritative server for my domain and at the same time the recursive dns server for my servers ? [18:56] do what? [18:56] pdns is a server only [18:56] pdns-recursor is a recursive server only [18:56] Ah ok [18:56] thanks [18:57] given my success with pdns, I will try pdns-recursor [18:57] thanks guy [18:57] guys* === Lcawte is now known as Lcawte|Away === msx is now known as Yezabel [19:14] would it be save to remove /etc/init/network-interface.conf if i don't have any hotswap network interfaces, or is this necessary for normal boot time network config? [19:14] i guess it executes the ifup so probably not [19:23] damn, I just killed my system [19:24] I did a chown -R on /var mistakenly [19:24] nothing works === Yezabel is now known as Yezalel === Yezalel is now known as Guest99110 [19:30] Aison: boot to a live ISO or the server rescue option, then use a script to read the correct ownership from a good system and apply it to that broken one [19:40] From the live iso for example ;) [19:40] Aison: this works with the broken FS mounted at $TARGET (remove the "echo" inside the loop to have it actually do the chown rather than print it): "TARGET=/target; while read inode b perm qty user group size month day time name; do [ -e "${TARGET}$name" ] && echo chown $user:$group "${TARGET}$name"; done < <(sudo find /var -type d -ls)" [19:41] Aison: You might want to modify that since it only targets directories, and also won't operate on directories that aren't in the reference system. You could sort the result of the find by path-length first, and apply permissions using -R from the shortest path to the longest === Lcawte|Away is now known as Lcawte [19:47] HI using ubuntu 14.04 I got this error: BUG: Bad page map in process sshd pte:01000000 pmd:23ed32067 what this means? thx [20:09] TJ- hi have a question on uefi. I have installed grub on a usb disk. that creates an entry that I can display through efibootmgr -v, but when i reboot sometimes I do not see that entry could be an issue with UEFI implementation? thx [20:10] zartoosh__: That sounds like the firmware isn't reading the USB device correctly/in-time, and so suppresses the boot-menu entry because the device isn't there [20:11] zartoosh__: If the mobo has an EFI shell you could explore from there === Ursinha is now known as Ursinha-afk [20:12] zartoosh__, It does have uefi shell but it does not see the entry at all, It can see the uefi entry from hardisk but not the one created through grub-install ? [20:17] TJ-, i've got the var directory now from my bacula backup [20:17] TJ-, this is how I install the grub, please let me know if you see anything wrong: http://paste.ubuntu.com/8020251/ thx [20:17] TJ-, so maybe it is possible to copy over just the rights? [20:22] Aison: Sure, use a modification of what I posted [20:23] TJ-, just working on that :D [20:23] zartoosh__: looks OK... but I'm a bit tired right now not thinking clearly, so don't rely on me :) [20:28] TJ-, thanks ping you tomorrow thanks [20:39] hm, ipxe currently doesn't build in utopic [20:39] It fails because [20:39] ISOLINUX_BIN=/usr/lib/syslinux/isolinux.bin bash util/geniso bin/ipxe.iso bin/ipxe.lkr [20:40] while that happens in trusty ^ , in utopic ISOLINUX_BIN is empty [20:41] hm, looks like syslinux-common needs to be added as an explicit build-dep now [20:42] yeah syslinux used to depend on syslinux-common, now it recommends it === Ursinha-afk is now known as Ursinha === Guest99110 is now known as msx [22:17] smoser: wtf - default utopic ec2 image doesn't have bridging enabled? === Lcawte is now known as Lcawte|Away