[00:04] I have been trying to install LSI SNMP Agent/SAS SNMP on my Ubuntu server box. LSI provides support for "Linux", but their support is an RPM file, and since ubuntu is deb based, well, there is the issue. I used alien to convert it to a deb file, but when I install it along with some other snmp tools, I get the following: http://pastebin.com/p0p44ZKV and I was wondering if anyone has had success with this type of install before, or if anyone could lead m [00:04] e in the right direction. [00:05] GH0: can you pastebin /etc/lsi_mrdsnmp/sas/install ? [00:05] (or at least the first 40 lines or so?) [00:05] Sure, give me one moment [00:06] hrm, the next error is line 206; maybe the whole thing would be best.. [00:06] is there a more convenient way of pastebining an entire file besides copying and pasting then? [00:06] GH0: the pastebinit tool is handy :) [00:06] I recall a cli pastebin tool, just don't remember the name. [00:08] fwiw, it's been years since I've used alien but I was impressed with the results [00:08] http://paste.ubuntu.com/7703283/ [00:08] I mean thats fine, I just don't really see any other way to install this. I really would like e-mail notifications if something happens with the array, and I am trying to keep tabs since I am seeing weird performance on the array as well. [00:12] [ -f /usr/bin/snmptrap ] && rm -f /usr/bin/snmptrap [00:12] cute guys [00:13] I take it I am better off searching for something else. :D [00:13] Especially if it wants to remove a binary file lol [00:16] this script appears to do only two real important things; add a MIB to a snmp config file, and write NoOfInstalledController kinds of things to their own configuration file, but that bit is just incomprehensible.. [00:16] it also configures a new service to run the system-supplied snmpd but only for LSI controller cards. I think. [00:17] nagios-plugins-contrib claims to have a check_raid service that does a lot of LSI checks, but I don't know if any of them are recent.. [00:18] GH0: and the mpt-status package claims roughly the same thing [00:19] Alright, I will look into those instead [00:33] sarnold, mind helping me with one more thing related to LSI/MegaRAID? http://pastebin.com/MupdnL81 :/ [00:34] If I try to shut the framework down, I get the following: Shutting down Framework: [00:34] sh: 0: Can't open /Framework/shutdown.sh [00:37] GH0: guh. that's annoying :) [00:37] Yup :/ [00:37] "cd: /usr/local/MegaRAID Storage Manager: No such file or directory" [00:38] .. and that's practically terrifying [00:38] Don't know why they can't just create .deb files. Would make this so much simpler. [00:41] Actually, I take that back, dpkg --force-all --purge seems to have gotten rid of it, but, I am not sure it removed it cleanly. [00:41] .. and I don't see any dpkg options that would ask it to skip the scripts.. [00:41] GH0: oh! hooray. [00:42] GH0: and half that stupid script is to work around now knowing how to use rpm packages to specify dependencies. annoying. [00:43] they'd do better if they didn't try as hard; if they juts had a tarball and README then surely the suse packagers and red hat packagers and debian packagers would have just picked it up and done the right thing and it'd be less annoying all around [01:09] Hm, the smtp part of that package is still broken after upgrading to a new version. Or it just doesn't like relaying data through Gmail. However, postfix is installed on my machine as a relay to gmail, and I am not sure if I want to open it up to allow a single program to use it like that. Unless I can limit what IP postfix listens for it's relay on. Never really done much of this stuff before. === veebull_away is now known as veebull === FreezingAlt is now known as FreezingCold === thumper is now known as thumper-afk [05:23] Just installed apache playing around with it, it's working and i can connect to the host but i can't figure out how's it's running $ netstat -an | grep :80 [05:23] tcp6 0 0 :::80 :::* LISTEN [05:24] i don't see it listening on regular tcp, yet i can connect to it by going to http://127.0.0.1 [05:47] Good morning. === veebull is now known as veebull_away [06:15] Bind mounts used to ignore -o ro, unless you did a -oremount,ro. I vaguely recall that issue got fixed eventually, but I can't find any evidence. Did I dream it? [06:17] Seems like I did -- even on my laptop running 3.14 it's still read-write by default, though it at least warns you now. [06:58] hi, my server is being restarted for a reason I dont know, I dont see anything in the logs and the vps provider says they dont restarted anything [06:58] how can I investigate this? [06:59] syslog is empty about restarting [07:02] When did this start happening? What changed between then, and when it last worked? Is there a pattern? Can you trigger it reliably? Is this happening for all hosts you run -- if not, what's different about this one? [07:03] Consider exfiltrating the logs to a host that isn't crashing. Consider traffic analysis (e.g. tcpdump) leading up to the reboot. Consider exfiltrating dmesg output to a reliable host (keyword "netconsole"). [07:06] Hi [07:06] I have rsyslog and haproxy running on the system. I need to enable debug logs for HAProxy. Any clue on how to set it? I am running Ubuntu 11.10 === _ruben_ is now known as _ruben === veebull_away is now known as veebull [08:47] hello everyone, is there some one can help me about MaaS I've just one question: I've realized an infrastructure using KVM as Hypervisor and 3 VM: 1 VM for Region Controller (RC) - 2 VM for Cluster Controller(CC), each one with 2 or more nodes installed and manage via virtio. Each VM use different Virtual Network 2.2.2.0/24 (RC),3.3.3.0/24(CC) 'n 4.4.4.0/24(CC) configured on virsh-manager, I've also added and accepted 2 CC on [08:47] RC. but in the RC's web UI I see only 2 CC without the nodes associated on them. But which is the Region's goal? seeing just the CC and not manage their nodes? thanks a lot [08:48] in this case I 've only 2 CC each ones with 2/4 nodes and it's easy to manage that because the nodes are flew, in the case we have more of them How can manage that??? RC should make this function, I mean I think that...... [09:04] in this case I 've only 2 CC each ones with 2/4 nodes and it's easy to manage that because the nodes are flew, in the case we have more of them How can manage that??? RC should make this function, I mean I think that......' [09:04] anyone can help me?....ops sorry I've pasted that === veebull is now known as veebull_away [09:18] anyone can answer me?is there a way to manage the nodes of each single CC via RC? or RC can view only the CC on its web ui? === ndee_ is now known as ndee [09:56] Hi. How do I add a PPA on 12.04? I can't find or install the command add-apt-repository? [09:56] !ppa [09:56] A Personal Package Archive (PPA) can provide alternate software not normally available in the offical Ubuntu repositories - Looking for a PPA? See https://launchpad.net/ubuntu/+ppas - WARNING: PPAs are unsupported third-party packages, and you use them at your own risk. See also !addppa and !ppa-purge [09:57] !addppa [09:57] Since Ubuntu 9.10, a !ppa can be added using a single command « sudo add-apt-repository ppa:user/ppa-name » See https://help.ubuntu.com/community/Repositories/Ubuntu#Adding%20PPAs for more details [09:58] cfhowlett, WHats the repository for add-apt-repository? [09:58] the package is software-properties-common: [09:59] Ah, I found it: python-software-properties [09:59] yes, do you know 'apt-file' [09:59] I've used it, a long time ago. I'll use that next time :-) [10:00] Thanks! :-) [10:00] with apt-file you can search a file through packages, so if you need to find a command that is not the package name, it is usefull [10:00] ok, np === Guest90121 is now known as hxm [10:44] mdeslaur: please could you take a look at bug https://bugs.launchpad.net/ubuntu/+source/php5/+bug/1334572? I'm not sure if this is a dupe or expected behaviour. [10:44] Launchpad bug 1334572 in php5 "php5-fpm upgrade broken after deleting sample config" [Undecided,New] [10:44] (and the reporter has cut the message short) === kickinz1 is now known as kickinz1[lunch] [11:03] Question: Ubuntu Server 14.04 64bit. Is running on KVM with 4GB memory. But top/htop shows only 2Gb. What can be wrong here? [11:05] hello, i'm trying to install ubuntu 14.04 in raid0(stripe) during the install it complained about unable to install grub so i tryed to follow this http://askubuntu.com/questions/334012/a-guide-to-install-ubuntu-13-04-using-a-raid-0 but that does not seem to work for 14.04 [11:16] rbasak: well, he deleted the conf file... [11:16] mdeslaur: yeah. What I couldn't see was the reason for the presumed subsequent failure, since he didn't explain that part. I presume his unattended upgrade would do some default thing, but I don't know what. [11:17] mdeslaur: do you want to comment or shall I? [11:17] rbasak: I gather the failure is because the package upgrade put the conf file back [11:17] I'm commenting [11:17] Thanks [11:17] mdeslaur: I don't think an upgrade puts a deleted conffile back. NOt sure though. [11:18] hrm, let me check === stephank_ is now known as stephank [11:28] rbasak: right, it shouldn't replace it...how odd [11:37] morning === Beliq_ is now known as Beliq === kickinz1[lunch] is now known as kickinz1 === veebull_away is now known as veebull [12:11] ws2k3: you cannot install grub on a raid0. [12:13] it tryed to install grub on /dev/sda and sdb [12:13] and i saw people on youtube doing that with 13.10 in virtualbox [12:14] then how should i setup so i can install ubuntu on my raid0 [12:14] ws2k3, you can create a little partion in raid1, and an other in raid0, on raid 1 you put your /boot, and grub will install fine === eagles0513875 is now known as greenrice === greenrice is now known as eagles0513875 [12:14] ws2k3: grub needs more than just the MBR. And those files cannot be located on a RAID. [12:15] *RAID0. [12:16] https://www.youtube.com/watch?v=lTAh53AXSXA#t=394 then how this guy did it [12:17] he has one raid0 and installed grub just fine [12:17] 5:29 you can see him picking raid0 [12:18] * kickinz1 watching [12:19] I dont care about videos, since technically, it cannot work using a RAID0. [12:26] When I build a kernel with Ubuntu server 14.04, even with same config, even if I use the same initrd, it won't boot. It fails to mount root, and the initramfs prompt it gives does not work... keyboard input doesn't show up, and ctrl+alt+del doesn't work. What did I do wrong? [12:26] kickinz1 did you view the video ? [12:26] peetaur2: How do you build the kernel actually? [12:27] bekks: first I tried a simple "make -j8 deb-pkg", and also tried this howto https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel (which is dreadful and out of date) [12:27] yes I'm trying, to install to see, but it must not work, normally grub works only with raid1 or hardware raid [12:27] not clear: yes I've seen the video, and I'm trying to install to check [12:28] bekks: in both cases, I ended up with a .deb file that I could install, but they can't mount root fs, either by lv path (/dev/mapper/vg-lv) or UUID (root=UUID=...), and the initramfs prompt hangs like that [12:28] I was hoping to ls /dev to see what was there ... [12:29] peetaur2: And whats the exact error message? And why are you tryong to compile your own kernel? [12:30] bekks: I have 3 machines where it randomly crashes due to a known bug with this fix https://lkml.org/lkml/2014/5/8/275 [12:30] peetaur2: So which kernel are you using currently? [12:30] bekks: and I'll try to get the message in a sec [12:32] bekks: when I compiled with "make deb-pkg" it said it was 3.13.11.2, and when I do the wiki procedure it says 3.13.0-29 [12:33] peetaur2: And which ubuntu kernel (non self-compiled) are you using? [12:35] kickinz1 did you view the video ? [12:35] 3.13.0-29 [12:36] ws2k3: Did you implement the proposed solution yet? [12:37] bekks: last part of boot screen http://bpaste.net/show/WwVEyrGOZ6WAA93VfXk6/ [12:37] the make one volume with raid1? no cause i see somebody else doing what i want but people here says thats not possible that confuses me [12:38] ws2k3: Well, the people in here actually do know what they are talking about. Why dont you just trust them, instead of believing a random video? [12:38] raid0 is stripping, so data is part on disk 1 and part on disk B, so it is not possible (from what I know) for grub to get data in the good order. [12:39] kickinz1: Thats correct. [12:39] so you put a raid1 little partion and mount it as boot, so grub copies all neede in that. [12:39] i know. md0 is my raid1 but why i cannot install grub just on /dev/sda and /dev/sdb [12:39] how big does /boot need to be ? [12:40] ws2k3: About 512M. [12:40] raid 1 is mirroring, so partion on diskA equal what is on disk2 part 1, so grub understand it even if raid is still not mounted [12:40] ws2k3: I usually have it at 256M [12:44] bekks: I just tried booting the stock kernel, but edited the boot line so it has the wrong root device, and it does the same except the initramfs shell works fine. So what does that tell us? [12:45] whousld that 512 mb just be ext4 ? [12:45] peetaur2: that your kernel is broken. How did you create the config for your own lernel? [12:45] ws2k3: Yes. [12:46] bekks: when I did the "make deb-pkg" method, I just did cp /boot/config-$(uname -r) .config so it should be the same as the working kernel. [12:46] peetaur2: And did you use make oldconfig after copying the config? [12:46] bekks: yes [12:47] I know what I'm doing when building the kernel... just not whatever extra undocumented thing I need to do with Ubuntu apparently. [13:05] i have made a raid1 with /boot but still the grub installation fails [13:06] it tryes to install grub on /dev/sda en /dev/sdb [13:06] Whats the exact error message? [13:07] executing grub-install /dev/sda failed [13:07] this is a fatal error [13:07] So switch through the TTY and see which more detailed error you can find. [13:08] as root? that's all it says? [13:08] I guess you can only mean in the installer... in which the TTY answer is the one [13:09] so what file should i check in the tty [13:09] or alternatively, find a text TTY and run that command yourself. And as a side note, is it a GPT disk? If so, you need a bios_grub partition (not the same as /boot) or it will fail. [13:09] ws2k3: I told you waht to do with the TTYs. :) [13:09] ws2k3: he means try all the TTYs until you find one with a log... and then still look in the rest in case there are more (in some distros one has dmesg and another has more install logs) [13:12] syslog only say grub-install /dev/sda failed [13:12] alt+F4 gives you the logs [13:14] it does comeplay something about gpt [13:15] also it seems he tryed to install grub on the wrong partition [13:15] so if it says you need a bios_grub for GPT and something about refusing to use a hack called blocklists, then it's the error I'm thinking of, and you need a bios_grub partition. [13:16] ws2k3: "something"? Whats the exact error message? [13:16] or regardless of what it says, if you are using a GPT partition table (eg. >2.2 TB disk), then make a bios_grub. [13:16] this system was only ssd's of 200 gb [13:17] its saying warning this HGPT partition label contains no BIOS boot partition [13:17] yeah that's the message.... make a bios_grub aka bios boot partition [13:17] So you need to create a boot partition first. [13:17] bios_grub .. not just /boot partition [13:17] cause i already created a partition in raid1 with /boot [13:18] ws2k3: Which isnt a GPT boot partition. [13:18] parted /dev/sda mkpart ....blahblah... ; parted /dev/sda set $n bios_grub on (where $n is the partition number, and ideally it is before 2.2 TB on the disk, and in number order ... I've seen it fail where the partition numbers are out of sequence, so I just always make it first partition) [13:20] how can i get out of the alt + f4 [13:21] guess at more nubers to find the one you were in before (likely F1) [13:21] text mode is F1 in Ubuntu installers... and maybe non-text is F7 [13:36] ws2k3,bekks, it seems surprisingly possible to boot directly on a raid0 on trusty, I just created a vm, with a lvm on top of raid0 and it just worked. [13:36] kickinz1: with /boot on a RAID0? It doubt that. [13:37] kickinz1, ws2k3: it is documented as supported to use metadata 1.2, raid10, raid0, etc. with grub2. However, it is very unlikely for it to work reliably. (eg. I know that raid10 will just make it fail when degraded, but seems to work fine otherwise) [13:38] yes, just a big raid partition on each disks, a lvm on top with two logical disks, one for swap one for /) [13:38] kickinz1, ws2k3: but for a surefire simple solid system, use raid1, and for metadata I know 0.90 works, and I knwo that on openSUSE (grub2) 1.0 works, and on Ubuntu (grub1.99) not sure if 1.0 works. [13:40] (just checked and Ubuntu 14.04 uses grub 2.02~beta2-9) [13:42] yeah well the reason i wanne do this is cause this systems are part of a cluster [13:42] and i need high io over relayable [13:43] so basicly your saying you cannot use raid0 for the OS ? [13:44] no, just /boot matters [13:44] OS can be on anything [13:45] hmm okay [13:46] so /boot can be on a raid1 right ? [13:46] yes [13:47] and in my experience, metadata 0.90 is a good idea (side effect is that mdadm.conf can't control the /dev/md* number so you'll likely end up with /dev/md127), or metadata 1.0 which I have only tested in openSUSE. And 1.2 is a bad idea. (grub2 officially supports it, but never worked when I tested it). And sadly the installer will not ask you which version you want. (openSUSE is smart enough to set /boot to 1.0 instead of 1.2 though) [13:49] what does this means peetaur2 [13:49] it means that 0.90 metadata (mdadm --create -e 0.90 ...) works best for /boot. [13:50] and after you install, if you care about it, test it. If it doesn't work, you could consider manually changing it. I always do. Others don't care. [13:50] I think it's not well known that metadata 1.2 and grub are buggy together. [14:04] odd i have made a partition and mounter it as /boot in the installer [14:04] but it still does grub-install --force /dev/sda and that is failing [14:06] at least that is what the logs show me [14:07] Hello folks. Any hint what could be the issue when imklog moans at the /var/log/syslog about a bad file descriptor? My rsyslog goes upt o 80% CPU usage. Ubuntu 14.04 [14:11] mswart: grub-install doesn't touch /boot... it touches the MBR and bios_grub [14:12] mswart: update-grub and update-initramfs touch /boot [14:18] during install i cant get it to work [14:18] so i will install ubuntu normale and change to later to raid [14:19] ws2k3: did you make a bios_grub? [14:20] no how should i do that ? [14:21] nearly the first thing I said was it will fail if you don't... [14:21] just when you say "use as: ..." instead of saying raid, or ext4, or whatever, say something like "reserved bios boot area" [14:21] cant recall you sad that [14:21] third thing I said: [15:09] or alternatively, find a text TTY and run that command yourself. And as a side note, is it a GPT disk? If so, you need a bios_grub partition (not the same as /boot) or it will fail. [14:22] and how do i know this is a gpt disk ? [14:22] heh [14:22] you can only guess ;) it stupidly will not ask, but will automatically make it GPT if it is large enough [14:22] or it will leave it GPT if it is already GPT probably [14:22] !gpt [14:23] and you said earlier: [15:14] it does comeplay something about gpt [14:23] [15:17] its saying warning this HGPT partition label contains no BIOS boot partition [14:23] !info gpt [14:23] gpt (source: gpt): G-Portugol is a portuguese structured programming language. In component universe, is optional. Version 1.1-2ubuntu1 (trusty), package size 71 kB, installed size 260 kB [14:23] which are hints that it is gpt. [14:23] cfhowlett: :D [14:24] peetaur2 yeah, that one is obviously wrong [14:24] !info GUID [14:24] Package GUID does not exist in trusty [14:24] !info GUID Partition Table [14:24] 'Partition' is not a valid distribution: extras, kubuntu-backports, kubuntu-experimental, kubuntu-updates, lucid, lucid-backports, lucid-proposed, partner, precise, precise-backports, precise-proposed, quantal, quantal-backports, quantal-proposed, saucy, saucy-backports, saucy-proposed, stable, testing, trusty, trusty-backports, trusty-proposed, unstable, utopic, utopic-backports, utopic-proposed [14:27] back i was dissconnected for a second === hxm is now known as Guest96132 [14:29] so how do i know the installation sees my disk a a gpt ? [14:40] ws2k3: ubuntu cds are retarded and don't have parted, etc. in CLI, ... so just make a bios_grub, install, and look afterwards [14:41] parted /dev/sda print says the type, but parted doesn't work in the install. It should work in rescue mode though. [14:48] How do I compile a .deb package normally again? [14:54] !compile|sander^home [14:54] sander^home: Compiling software from source? Read the tips at https://help.ubuntu.com/community/CompilingSoftware (But remember to search for pre-built !packages first). Also read !checkinstall === Solution-X|AFK is now known as Solution-X [15:18] hi! what is the best way to debug upstart when the service looks happy, but fails with nothing in the upstart log? === kickinz1 is now known as kickinz1|away === krtaylor is now known as krtaylor_brb === RoyK^ is now known as RoyK === Lcawte|Away is now known as Lcawte === krtaylor_brb is now known as krtaylor [17:55] how do we flush/empty ufw preouting rules? [17:55] we've deleted the /etc/ufw/before rules [17:56] but can't get the rules to actually purge [17:56] prerouting* [17:59] ufw disable ; ufw enable ? [18:02] RoyK thanks [18:10] /lib/ufw/ufw-init force-reload [18:10] /lib/ufw/ufw-init flush-all [18:10] whew [18:10] I thought just restarting the thing should do it [18:10] worked for me... [18:16] is there a way to mirror a PPA locally? [18:19] wget -m ? [18:19] since it's still just a debian repo with a web server [18:19] hahaha [18:19] yeah, wish it had rsync or such === tomaw is now known as 5EXABH90N === 5EXABH90N is now known as tomaw [20:16] is it okay/safe to delete old linux kernels? [20:16] Keep at least one older kernel, as a backup :) [20:18] gotcha - i have 11 right now, so i figure i can afford to delete at least a few === elliotd123_ is now known as elliotd123 [20:31] Anyone know of an rsync target for ppa.launchpad.net === kaitanya_ is now known as kaitanya [20:35] Phibs: why not wget? [20:38] RoyK: why not ... drive a tricycle to work [20:39] Phibs: if there's no rsync, well, what can you do? [20:40] truth [21:29] HI I am using 14.04. During boot, the name of ethernet interface is changed from eth0 to device controller name: "systemd-udevd[155]: renamed network interface eth0 to em1" is there anyway I can stop this? thx === maxb_ is now known as maxb === alexisb is now known as alexisb_afk [22:40] hey folks, when i run dmsg i see what looks like a timestamp [1381034.102975] . how can i convert this into a regular date/time stamp? [22:42] xibalba: Using this tiny script e.g.: http://linux-funda.blogspot.de/2013/10/convert-dmesg-timestamp-to-human.html [22:42] sickies [22:42] trying [22:43] bekks, thank you much. worked well [22:44] you're welcome. [22:44] It reminds mind to write an alias for dmesg, printing human readable dates :) === Lcawte is now known as Lcawte|Away === sveta is now known as svetlana