=== Guest69248 is now known as rcj === rcj is now known as Guest46616 === phunyguy is now known as phunyguy-zombie === phunyguy-zombie is now known as phunyguy [01:34] Hello, may I ask help with postfix + dovecot here? [01:35] dmiroshn: welcome to irc; it typically works best if you just ask questions and hang around for a while :) asking if you can ask just adds delays.. [01:47] I used https://help.ubuntu.com/lts/serverguide/postfix.html for postfix and dovecot setup. I pretty sure that I did what that page suggests. However, when finished I cannot neither receive nor send emails. Is it complete guide? Should I try something else? Should I install dovecot-imapd as well? [01:49] dmiroshn: email is very complicated these days, it isn't as easy to run as it once was [01:49] dmiroshn: break the problem apart into small steps -- telnet to the smtp port, try to deliver mail by hand, use tail -F to watch the log files while you do it, and verify that the mail is accepted ... [01:50] dmiroshn: .. and take small steps, focus on one problem at a time, either sending, or receiving, or retrieving, and keep an eye on the log files. === markthomas is now known as markthomas|away [02:23] dmiroshn, generic email pointers: MX records, reverse PTR records, SPF records, DKIM/DomainKeys, blacklist check === zz_DenBeiren is now known as DenBeiren [03:30] When I am trying to send email via "openssl s_client -crlf -connect :465". I got the following error "Recipient address rejected: Access denied". May one check my "postconf -n" settings http://pastebin.com/WQp8k6X5 to help figure out what's wrong? [03:40] dmiroshn: it means the email address you are trying to send to is not on the mail server you are connecting to [03:52] ikonia [03:52] ikonia: Do you want say that recipient email is incorrect? It should not be the case because it's my gmail account that I frequently use. [04:00] dmiroshn: right, so your server won't host gmail [04:01] dmiroshn: I would assume you're machine is not setup as a relay host [04:02] ^ Please ignore my last (fixed) question. [05:41] anyone here dealt with apache openmeetings? === liam_ is now known as Guest14772 === kickinz1|afk is now known as kickinz1 === Sachiru is now known as TheOneAndOnlyRea === TheOneAndOnlyRea is now known as The1AndOnlyKiwi === The1AndOnlyKiwi is now known as Sachiru === Lcawte|Away is now known as Lcawte === bilde2910|away is now known as bilde2910 [09:05] Good morning. [09:05] EHLO [09:06] I have some pent up rage in me today, apache openmeetings is a booger to install [09:07] yet my googly client insists that we must have this pronto === Lcawte is now known as Lcawte|Away === kickinz1 is now known as kickinz1|afk === kickinz1|afk is now known as kickinz1 [10:58] hi folks [10:58] I'm trying to install ubuntu-server on a system with EFI boot off a hardware raid [11:04] kblin: should work... === kickinz1 is now known as kickinz1|afk === kickinz1|afk is now known as kickinz1 [11:50] RoyK: sorry, was dragged off to lunch :) [11:51] for some reason the installer keeps installing the boot loader into the partition that has /boot instead of using the MBR [11:51] and then the system fails to boot [11:53] hm... not good [11:53] don't know why, though [12:36] hi@ll, i like to know how to get the fingerprint of my SSL-certificate via SSH .... SHA256 and SHA1 .... what are the commands?? [12:37] oopaine: "openssl x509 -fingerprint -noout -in foo.pem". Or something like that. [12:37] See the x509 manpage for details. [12:37] it's one of the x509 options, for sure [12:40] okay, thx for this faaast answer. :) [13:11] bb. [13:13] ok, even with default settings for everything I can't get that stupid EFI install to boot [13:25] Ok Hi, Guys I just installed Trusty yesterday and I went through the steps of connecting to a Wi-Fi Network when I did I had to attempt 3 times to connect to my personal wifi network with the right password. Now, This morning I tried pinging google as a test if the internet is working and it wasn't. I even checked ifconfig, Nothing [13:28] optrusty: Can you paste the output of `ip addr`? [13:29] Odd_Bloke I will try ssh via ethernet [13:34] Odd_Bloke So I searched and I couldn't get an IP for wlan0 [13:35] Odd_Bloke Let me try checking /etc/network/interfaces [13:36] Odd_Bloke In the network interfaces there is only the loopback network [13:39] optrusty: Server, or desktop? [13:40] Odd_Bloke Ubuntu Server [13:40] Odd_Bloke I would be in #ubuntu by now [13:45] optrusty: Oops, didn't notice which channel I was in. [13:45] Odd_Bloke No Prob [13:46] optrusty: What steps did you take to try to connect to the network? [13:51] Odd_Bloke I am going to try this http://ubuntuforums.org/showthread.php?t=318539 === Guest46616 is now known as rcj === rcj is now known as Guest97390 [14:16] jamespage: can you take a quick look http://paste.ubuntu.com/9698561/ (im concenered about the debian/control file) [14:19] zul, you need Breaks/Replaces; python{3}-oslo.i18n (<< 1.3.0-0ubuntu1) as I think it still provides oslo.* links for backwards compat right? [14:19] right [14:20] zul, does it? this is quite important - if not then we need to holdup the renames until after kilo-2 releases [14:21] jamespage: there is an oslo.i18n and a oslo_i18n [14:21] so both namespaces are there [14:25] jamespage: better? http://paste.ubuntu.com/9698591/ [14:40] Odd_Bloke ping [14:43] optrusty: Pong. [14:44] optrusty: (Generally it's better to just tell me what you're going to tell me; saves a round trip :) ) [14:45] Odd_Bloke It's a waste of time if you are away [14:45] Odd_Bloke That's why I like to check if some is home :) [14:45] optrusty: Nope, I'll always see the message. [14:45] optrusty: Furthermore, if other people are following along, they may be able to jump in and help you. :) [14:48] Odd_Bloke Anyway I am doing a fresh reinstall after there was a little incident with over writing the interfaces file [14:49] optrusty: :D [14:49] Odd_Bloke My passion for GUIs are gone :( It is really hard to use cli and not be clumsy [14:50] jamespage: ping looked ok? [14:52] zul, that looks OK [14:52] jamespage: thanks [14:52] zul, are you happy this won't break all the kilo-1 uploads? [14:52] jamespage: yeah im sure it wont [15:20] jamespage: just running the tests for kilo-1 with a newer i18n just to make sure [15:20] zul, +1 [15:20] sounds good === Guest97390 is now known as rcj [16:09] jamespage, zul: systemd enablement for ceilometer - https://code.launchpad.net/~corey.bryant/ceilometer/systemd/+merge/245973 [16:09] jamespage, the templates approach is really nice [16:20] jamespage, oh shoot, I see Didier already merged systemd for ceilometer [16:20] coreycb, yeah I spoke with him - just rebase you changes ontop of his and drop the specific bits he added [16:20] jamespage, ok === cmagina_ is now known as cmagina [16:29] I have a problem installing Ubuntu-Server. It keeps installing Grub to the USB as SDA [16:32] CluelessPerson: only grub? is ubuntu installed to the disk? [16:36] Eggs_: it seems only the grub, yes [16:36] Eggs_: I just attempted to do sudo shutdown now [16:37] Eggs_: it's failing to even shutdown [16:37] * CluelessPerson shutoff the computer physically and restarts to see if it boots... [16:37] Eggs_: I ran the command > sudo grub-install /dev/sda [16:37] Eggs_: That seems to have fixed the boot problem [16:38] Eggs_: Now the next problem is that it won't shutdown, it fails something, goes to root user, and waits. [16:38] * CluelessPerson runs > sudo shutdown now [16:39] Eggs_: it goes to the purplish Ubuntu 14.04 splash screen, then lists the normal line [16:39] root@host:~# [16:39] awaiting my entering whatever [16:41] is that the root user just for the install? === markthomas is now known as markthomas|away === markthomas|away is now known as markthomas === markthomas is now known as markthomas|away === kickinz1 is now known as kickinz1|afk === markthomas|away is now known as markthomas [17:40] HI I am using ubuntu 14.04. I do not understand the propose of the /etc/rc.d/rc.driverUpdate? thx [18:07] jamespage, in case you're not eod, can we get ironic and horizon merged (2015.1~b1-0ubuntu1)? [19:31] HI I am using ubuntu 14.04. How can I control sequence of kernel loadable module when they are getting loaded? thanks [19:34] rostam: that sounds like solving a problem that shouldn't exist [19:38] rostam: What is the exact issue you are having which makes you ask this question? [19:53] rostam: What are you traxing to achieve? [19:59] All, we have a video driver from AMD which we need to run a script before that driver is loaded. Is there a way I can do that please? thanks [19:59] The driver name is flgrx. [19:59] fglrx [20:02] rostam: modify the udev rule that loads the module to run a script instead? [20:02] rostam: hmm, it's kind of ugly, but you could stick the fglrx module into a /etc/modprobe.d/ blacklist, then make an initscript that runs the script you want and runs modprobe fglrx manually -- which I believe will ksip the blacklists [20:03] rostam: Modify the udev rule to run a script, which loads the module as last command. [20:04] I think bekks has the more elegant solution [20:05] hey, I want my cookie, I said it first :) [20:05] * genii slides mdeslaur a chocolate chip cookie and a fresh coffee [20:05] * bekks hands mdeslaur a cookie :) [20:05] \o/ [20:05] :) [20:07] thank you I will try all your suggestions and will report the outcome. Thanks === markthomas is now known as markthomas|away [20:16] hello, somehow my isc-dhcp-server6 sucks [20:16] when I start it, it uses 100% cpu time [20:17] Aison: check logs [20:17] nothing in the logs [20:17] no errors... [20:17] but I just found this one: 1971953593 Jan 8 20:41 dhcpd6.leases [20:22] what's wrong here?!? [20:22] lolo [20:24] Aison: use diagnostic tools (strace, perf) [20:28] I could swear I thought I read a bug report with an eerily similar description but I can't find the thing in my archives [20:33] hi [20:35] is there a filesystem that can be used to efficiently store incremental changes to files, for a set period of time perhaps, that can be used as an incremental backup solution possibly? [20:39] I deleted the leases file and then I can start dhcpv6 without having 100% cpu load [20:39] a|3x: ZFS using snapshots, compression, and deduplication, but be aware that it requires enterprise hardware and large amounts of RAM [20:40] the leases file contains "endless" many equal entries [20:41] qman__, enterprise hardware? [20:42] a|3x: ECC RAM, TLER disks at a minimum [20:42] qman__, why? [20:43] It is extremely sensitive to flipped bits, data integrity is a core design feature of ZFS and it doesn't work properly without it [20:43] In practice your system will crash a lot [20:44] tler disks are not needed [20:44] but if you expect to have any kind of quality of service, you need it [20:44] qman__, well, i do have ecc ram where i want to use it and raid 10, but i only plan to use it for backups (to be able to look at changes over time) [20:45] ZFS replaces RAID [20:45] It performs its own form of it [20:45] zfs isn't the solution you want then, likely [20:45] i see there are some other solutions here: https://en.wikipedia.org/wiki/Versioning_file_system [20:46] ext3cow, etc [20:46] are any of these good? [20:46] versioning filesystems isn't exactly what you want for a *backup* system [20:46] unless these backups are tiny, and change a lot [20:47] well i want high granularity snapshots basically [20:47] Where I am unable to use ZFS, I manage the backup versions via rsync trees [20:47] E.g. rsnapshot [20:47] but *snapshots* depend on how data gets put on it [20:47] snapshots don't like you puttings *new* backups onto it [20:47] it perfers you to change the existing ones [20:47] otherwise you will be looking at something that is designed to do dedup at it's core [20:48] but we will need a much better definition of what is *backups* in this case, to answer [20:48] patdk-wk, yes, i want to change filesystem over time and have snapshots and able to clone fs from snapshot or something like this [20:49] so it's really not a backup? but the production system? [20:49] or the backups just happen to be lots of tiny files that don't change often [20:49] then snapshots would be fine [20:49] also, would be nice to have a human readable log of which files changed in which snapshot [20:49] oh, now i've got several 100 of these messages per second in my syslog file, caused by dhcp6 [20:49] Jan 9 21:49:01 inetserv dhcpd: Renew message from fe80::3ed9:2bff:fea1:f066 port 546, transaction ID 0xD19A2B00 [20:49] Jan 9 21:49:01 inetserv dhcpd: Sending Reply to fe80::3ed9:2bff:fea1:f066 port 546 [20:49] a|3x you can do similar things in userspace on any FS, rdiff-backup is one way. [20:50] does it have to be on the FS layer? [20:50] rberg-, but is that going to be efficient enough? i would imagine fs layer may be much more efficient with less copies of same blocks [20:51] I suppose a backup will have t copy every block at least once, from there its a delta [20:51] a|3x: it is fairly space efficient but slow [20:51] patdk-wk, it is a backup system, i would like to rsync production system to it and create snapshots for future potential use [20:52] normally rsync and snapshots don't work [20:52] cause rsync will change the WHOLE file [20:52] not just the part that changed [20:52] a|3x: most of my production backups are rsync copies hardlinked together [20:52] but if that is *acceptable* [20:52] if it has to be a FS I would use zfs. while it likes lots of ram it doesnt seem to require it, at least when I ran it [20:53] rberg-, depends on many things [20:53] but you really want your metadata that you use to fit in ram [20:53] true, I didnt de-dupe [20:53] if it doesn't you kindof get screwed [20:53] dedup will just screw you, no matter what you do :) [20:53] a|3x: ZFS is the only filesystem level way I know of to do dedupe and snapshots well [20:53] patdk-wk, how do you hardlink rsync backups? [20:53] I don't [20:53] There are dedupe layer tools [20:54] ram use did get pretty high when I started playing with zil and l2arc, but I never got OOM kills so it must give the ram back fine [20:54] l2arc can't give the ram back [20:54] oops, qman__, how do you hardlink rsync backups? [20:54] something did :) [20:54] normally using l2arc unless you have >40gigs of ram, is a mistake [20:55] yeah, in my case thats true.. my workload didnt benefit from it.. (also a backup destination) [20:55] a|3x: a custom script that stores backups in directories by date, with the --link-dest parameter [20:55] Rsnapshot does this [20:56] qman__, thanks, this is an interesting solution [20:56] personally I went with md raid / xfs and dealt with the rest in userspace. [20:57] My system at home is zfs and I take hourly snapshots, and zfs send them to a replica server [20:57] I don't use dedup on that one though [21:01] Got almost 30tb and only 64g of ram so it wouldn't work === Batman is now known as Guest1080 [21:03] hello [21:03] can upstart be run outside of init? [21:03] server boot stops after Freeing initrd memory [21:03] there was a serious fsck... [21:03] how can I boot this server again? [21:04] does it not found the root device ? [21:04] lazyPower: Upstart runs all the time. [21:06] bekks: we're exploring some experimental work with juju + docker, and it appears that the local provider has quite a few dependencies on upstart - docker doesn't run a true init, so this is why i'm asking - to ensure that i'm not breaking some upstart dependency by starting it in the dockerfile [21:09] ok, as soon as I turn on IPv6 support on my HP LaserJet Printer, ubuntu server goes down after a few hours [21:10] because the dhcp6.leases and syslog file grows to serveral GB [21:10] and 100% CPU use [21:10] nice one.... === markthomas|away is now known as markthomas [21:41] where can i see the POST logs? [21:41] Alina-malina: Nowehere. You BIOS does not log the POST anywhere. [21:41] oh [21:41] anyway i can log that stuff? [21:41] Alina-malina: No. [21:41] erm [21:49] Hi sorry if this is of the topic could someone help me find out why this udev rule does not work: ACTION=="add", SUBSYSTEM=="pci", ATTR{idVendor}=="1002", DRIVERS=="fglrx", RUN+="/usr/local/test.sh" thanks [22:09] Alina-malina: maybe you can log it on some computers using IPMI / AMT / ... (and in theory UEFI could do logging too, but I don't know of any that implement it?) [22:10] erm [22:10] rostam: is that a typo? should it be /usr/local/bin/test.sh? [22:10] i dont know i thought to just log POST and GET [22:10] rostam: is the file set executable? is the interpreter valid? [22:30] sarnold, yes to all your question and I changed the path to /usr/local/bin/test.sh still no output.... [22:30] rostam: does your script require a specific PATH? [22:31] sarnold, the content is: #!/bin/bash echo "Test" > /tmp/test.txt . [22:31] Ok Guys I just installed Ubuntu Server with Host VM, but I don't know how to use it [22:31] rostam: hah, that should work :) [22:32] sarnold, I wished :) [23:01] hi using ubuntu 14.04. Is there a place I can install a script before any driver is loaded. I can get udev working so was wondering if there is a place in rc script I can place something before any module is loaded? thanks [23:03] Would it be possible to setup a Hotspot on my Ubuntu Server with CoovaChilli (just the login page), FreeRadius and Easyhotspot web with only 1 NIC on the server and a router that has the DD-WRT firmware with Chillispot? [23:03] Or would I need 2 NIC's on the server === bilde2910 is now known as bilde2910|away [23:31] I tried to create an upstart script [23:31] the strange thing is now, when I run the command as root in the console, it works perfectly [23:32] when I put the same command in exec line, then it segfaults [23:32] Today when I tried connecting to my server trough my saved putty profile I get the error " Server refused public-key signature despite accepting key!" How is that possible? I been ssh'ing the whole evening to tesame server without no problems [23:32] then suddenly this error came up