[05:27] Quick question [05:27] would apt-get update and apt-get upgrade cause my host key to change? [05:29] good morning === noxs_ is now known as noxs [06:50] halp: regular apt-get update en apt-get upgrade should not cause changed hostkeys [06:52] halp: upgrade between releases, eg 12.04 to 14.04 or 16.04 could add support for new hostkeys like ed25519 [06:53] henkjan: I got the answer already in #ubuntu but ty [06:53] I've wiped and locked down the machine now. It had nothing of value anyways === _dpawlik is now known as danpawlik [09:53] sarnold, Oh, hey. I did figure out what might be going wrong. It seems like my ISP is indeed stepping in and blocking requests to port :25 [09:54] Because I was able to connect to the same server from a business network no problem. [09:55] If I feel bothered, I'll see if I can bug my ISP to lift that limit else devise a different solution. [09:55] Either sanely by having it go to a proper service. [09:56] Or tunnel to a work server to facilitate the request. === JanC_ is now known as JanC === soahccc_ is now known as soahccc [11:10] Hi! when I run blkid I see these - /dev/sdb: UUID="LSI M-^@M-^F)%" TYPE="ddf_raid_member" [11:11] How do I make these hdds show up as normal hdds? in etc/fstab? [11:11] The hdd Im booting from is showing this -/dev/sda1: UUID="055b6e93-b53f-4020-96d1-8ad49aad2779" TYPE="ext4" PARTUUID="14b09c11-01" [11:26] Capprentice: isn't that a fake raid setup? [11:26] Capprentice: so would mounting it individually not be the wrong things to do? [11:30] cpaelzer, I do not know. I just used fdisk to format the disk partition type to msdos and then create one single partition of ext4. [11:31] Capprentice: that might even have killed some of the disk data - the "ddf_raid_member" means that this is a fakeRaid [11:32] cpaelzer, How do I make it non raid? [11:32] Capprentice: TL;DR the bios/hw does part of the raid setup, but would rely on special drivers to do so correctly [11:32] Capprentice: If you don't want it to be special you need to go to your bios and/or storage controller setup [11:32] cpaelzer, The SATA controller is set as AHCI. [11:32] RAID is not selected there. [11:33] And when installing I passed the argument nodmraid [11:33] Yeah also the borads/devices doing so get less and less - so it might be safer to go without [11:34] yet your blockid reports a ddf type device [11:35] cpaelzer, How do I fix it? [11:35] haven't had such a system for years - I'm thinking (slowly) ... [11:37] Capprentice: since I understand that you want to kill the data on it anyway to install fresh or something like it maybe wipe the old signatures and reboot so the kernel stops detecting it as such? [11:37] Capprentice: dmraid -x or so? [11:39] cpaelzer, The /dev/sdb/c/d can be wiped. I have no data on them. [11:39] Capprentice: there are more things to wipe it - here is a discussion you might want to follow and try https://ubuntuforums.org/showthread.php?t=1322108 [11:41] dmraid -rE seems to be the consensus in this thread, but since I don't have the device or used it be carful Capprentice [11:42] cpaelzer, I dont even have the package dmraid installed and I selected the AHCI mode. How or why it beacame lke this? [11:42] :( [11:42] Capprentice: I think the kernel tries to detect [11:43] Capprentice: but that is only half way done - either "kernel detects fakeraid, userspace consumes and sets it up" OR "wipe it all so it boots without being detected as such" [11:44] Capprentice: it might have metadata form an old installation I'd guess [11:44] Capprentice: there are plenty of similar references around like https://askubuntu.com/questions/91167/un-raiding-a-two-disk-fakeraid-array [11:44] Capprentice: but they all seem to resolve around wiping the unwanted metadata [11:49] cpaelzer, ERROR: ddf1: seeking device "/dev/sdd" to 1024204253954048 [11:49] ERROR: writing metadata to /dev/sdd, offset 2000398933504 sectors, size 0 bytes returned 0 [11:49] ERROR: erasing ondisk metadata on /dev/sdd [11:50] cpaelzer, This had no previous installation. This is the first installation on the server. I dont know why and how this got attached and now giving me error on trying to remove it :( [11:50] :-/ [11:51] Sorry Capprentice I can't help more, I think you now need to hope to find somebody with the same or a similar device [11:52] and while searching for him continue trying to find how to fully wipe it [11:52] Capprentice: what storage controller is that exactly, so that somebody else here might chime in if he has one [11:58] cpaelzer, I have disabled the controller in BIOS. I dont remember its name and lspci shows Sunrise Point-H PMC [12:43] Capprentice: Just coming in, but did you do anything to nuke metadata on the disks? [12:44] A not-horrible suggestion prior to using disks that aren't brand new/unused is to dd if=/dev/zero over them. [12:45] Or maybe if=/dev/urandom if you're going to do crypto on the devices later. [12:55] sgdisk -Z is also interesting, as it also takes care of the partition data at the end of the disk iirc [13:08] Is there a minimal iso for ubuntu server? [13:10] M3mphiZ: How much such an iso differ from the regular mini.iso? [13:12] andol: i thought the mini.iso is for desktop, no? [13:13] M3mphiZ: Well, Ubuntu is Ubuntu. Just that some installers provide different defaults. [13:14] andol: thats what i thought, so ubuntu server is preconfigured for server and the rest for desktop, thats why i thought the mini.iso is preconfigured for desktop.. [13:17] Unsure exactly what defaults the mini.iso gives you, but I do remember having used it to install a server. Just set the low debconf level and you will be fine. [13:17] ...and if you want to be really sure, start by testing it out in a virtual machien. [13:19] andol: what do you mean with lob debconf level? [13:20] andol: or better, how do i set it to low? [13:27] ahasenack: Does that know about fakeraid metadata? I think it just wipes partition data, which would leave the fakeraid stuff there. [13:28] mason: don't know, give it a try [13:29] ahasenack: You miss my point. I wasn't asking to ask, I was asking to tell. :P [13:30] where is the fakeraid stuff written to? [13:30] -Z destroys gpt and mbr data structures, it's what the manpage says [13:30] Looks like the last 512K on disk: https://djlab.com/2013/07/removing-raid-metadata/ [13:31] Right, but... that's not going to touch the metadata. [13:32] Put another way, -Z very cleanly and precisely destroys the partition structure. [13:41] coreycb: hey - the sync process for the UCA is currently broken - a duplicate build snuck into one of the ocata or pike pockets for qemu [13:41] coreycb: resolving that now but it currently means that no updates are getting from the build PPA's to the real UCA [13:41] jamespage: ack [13:56] coreycb: OK I think I've poked it enough and things are now syncing again [13:56] jamespage: ok cool [14:04] Hi, I have not used IRC in 10 years I guess that makes me a newbee... [14:05] rather a unpracticed oldie :P [14:06] thought so, how is it referenced now days [14:14] ah [14:15] IRC. :) [14:16] what software is recommended for server virtualisation would like to run two ubuntu servers at once ? [14:19] redvic: KVM [14:28] redvic: or lxd, for speed and density, unless you specifically need VMs. [14:29] I'll have to try lxd someday [14:31] i use odoo accounting software on ubuntu server 12 lts would like to run two version and one test server [14:32] hey, I'm having problems with a hdd turning to read only mode after a random amount of time. I have already run fsck and can't figure out the problem! has anyone had to deal with shit sort of thing with external drives before? [14:35] IShavedForThis_, usually means the drive has problems. use SMART and see [14:36] glad you shaved, btw [14:39] hmm okay. is there a command for that? [14:39] hahah thanks. [14:45] smartctl -a /dev/xxx [14:45] you have to replace the xxx with the drive [14:45] pastebin.com the results [14:46] may have to use sudo [14:49] how do i setup kvm? do i load ubuntu server then install kvm and setup my virtual servers from there? [14:50] yup. do you have a desktop installed? [14:50] redvic: https://superuser.com/questions/229084/does-displayport-carry-sound-as-hdmi-does [14:50] Um, that didn't copy properly. [14:51] https://help.ubuntu.com/lts/serverguide/virtualization.html [14:51] that one [14:51] So, if you want libvirt, for example: sudo apt install qemu-kvm libvirt-bin [14:55] what version of LTS ubuntu server is recommended 12 14 ? [14:56] we use 16.04 LTS here [14:57] i know its not server related what IRC client is recommended here ? [15:01] 16.04 [15:07] is it possible to setup a server between my wifi access point and internet connection in order to control access to network ? i am helping a school with wifi in all the class rooms for teachers only but students hack wifi and gain access to data [15:08] wifi password is a problem as students have access to teachers devices [15:09] can ipfire do something like this? [15:10] how would i secure data on a wifi network? [16:34] cpaelzer: triaging bug 1685332, I think it would be reasonable to say that non-experimental NVMe support for smartmontools is a wishlist request, so Triaged/Wishlist. What do you think? [16:34] bug 1685332 in smartmontools (Ubuntu) "does not monitor NVMe drives" [Undecided,Incomplete] https://launchpad.net/bugs/1685332 === aruns is now known as indistylo [17:33] powersj: poke. [17:33] teward: sup [17:33] powersj: thou hast forgotten nginx special triage rules [17:33] doh [17:34] "Invalid", not "Incomplete", because "Not a bug" [17:34] Actually I thought I did mark it invalid [17:34] ah, your canned response and the status don't match [17:34] that's whyt [17:34] *tosses phone into /dev/null* [17:34] yeah ;) [17:34] why* [17:35] powersj: i think we'd need to update the canned comment - it references "incomplete" [17:35] FRAKKK stupid tor >.< [17:35] there we go... [17:35] that's fair. I get into copy and paste mode, but did remember you prefer invalid to incomplete [17:35] powersj: yeah, i get too many of those "Hey, this doesn't start" "Stop running the other webserver ya ninny" situations [17:35] just got tired of them :) [17:36] I hear ya [17:36] back in a bit. IT never sleeps, nor do we get a freaking break xD [17:46] hey folks, we just released a new version of DBD::Pg; wondering how we get libdbd-pg-perl updated accordingly [17:47] endpoint_david: it's currently sync'd from Debian (with rebuilds as needed), so I'd guess if Debian updated, we'd pick it up [17:48] and debian won't update before debian 9 is released (this year, probably) [17:49] jelly: good point [17:49] endpoint_david: if there's a strong reason we should update ahead of Debian, I'd file a bug [17:49] unstable is "slushy", no new upstream version are coming in before stretch release [17:49] the primary reason is supporting PostgreSQL 10 [17:50] which breaks a few things as far as versioning, etc, internally [17:50] now, that hasn't been released yet [17:50] but that means that anyone using Pg 10 in the future is going to have to update DBD::Pg out-of-band AFAICT [17:50] endpoint_david: ok, i'm not sure we're updating to pg10 [17:51] yeah, I'd expect that isn't going to be in the base dist for sure [17:51] endpoint_david: it's on, i think, our roadmap for 18.04 [17:51] endpoint_david: let me check my e-mail (i, along with cpaelzer, are the primary packagers for the pg updates) [17:51] ok, when is the freeze for that? [17:52] endpoint_david: for 18.04? well, it's not even open yet :) [17:52] nacc: cool, good to meet you [17:52] then I haven't missed the deadline there... ;) [17:53] endpoint_david: i'm *guessing* that when Debian brings pg-10 out of exp to unstable, we'll probably see them pick up the new version of dbd-pg too [17:53] endpoint_david: it's probably best to file a bug in debian so they are aware of the need to update [17:53] nacc: so how would this work in the meantime? generally people will use the PGDG repo, so do we need to package a custom version of libdbd-pg-perl there as well? [17:54] since it's still pre-release it's kind of moot at this point, just trying to make sure things work out easiest in the future [17:55] endpoint_david: not sure I follow what the issue is (yet). The version we have in 17.10 (I am assuming) works with the PG in 17.10, right? When we do the migration to PG-10 in Ubuntu (next cycle most likely), there will be a transition tracker (as many packages will need to be rebuilt, updated, etc.) [18:03] nacc: the problem would be if someone installs the PGDG repo to install pg10 but uses the system libdbd-pg-perl (which doesn't work with Pg 10), then anyone trying to use DBD::Pg wouldn't be able to work with it [18:04] so to me it looks like the choices are to either package/distribute the updated DBD::Pg in PGDG repo as well or update system libdbd-pg-perl to support it [18:04] endpoint_david: right, if you have your own repo, then you need to make it self-consistent [18:04] endpoint_david: you package/ship PG itself? [18:05] well, the PGDG project does [18:05] and I'm sure we could get them to include a custom package [18:05] endpoint_david: yeah, that seems like the right place to resolve this, it's not really an ubuntu issue [18:06] dumb question: are external repos able to depend on other packages? we don't want to have to maintain a separate full perl installation for dependencies [18:06] endpoint_david: well, the packages themselves express their dependencies [18:06] endpoint_david: i'm assuming the PGDG repo is meant to be added to ubuntu [18:06] ubuntu or debian both [18:07] endpoint_david: so they could depend on any ubuntu package; but as with libdbd-pg-perl, they need to ship anything they want to control specifically the version of [18:07] endpoint_david: otherwise they are reliant on whatever ubuntu/debian decide [18:07] ok, well that sounds like the way we'll have to go then, thanks [18:07] thanks for your help [18:08] endpoint_david: np [18:10] Is anyone here pxe installing 16.04 server? How did you get the serial console to work? Through 14.04, I was able to just append console=ttyS0,115200 to the boot, and it worked. Not, not so much === JanC_ is now known as JanC === jmarsden|work is now known as jmarsden [20:50] is their any good open source ERP system that i can learn from...i want to learn how to setup an ERP system. is their any good open source ERP systems out their just so i can practice and see how an erp is implemented? === nacc_ is now known as nacc [21:19] !package openssh-client [21:21] sarnold: !info? [21:22] !info openssh-client [21:22] openssh-client (source: openssh): secure shell (SSH) client, for secure access to remote machines. In component main, is standard. Version 1:7.4p1-10 (zesty), package size 654 kB, installed size 4629 kB [21:22] oh! [21:22] sarnold: not sure if that's what you were looking for :) [21:22] nacc: that's it exactly; thanks [21:23] with !package the little guy replies with Sorry, I don't know anything about package openssh-client -- which looks like it might not understand the https redirect on https://packages.ubuntu.com/ :) [21:23] but if you ask the correct thing it still gives correct answers. hehe. [21:25] sarnold: right, because it's interpreting the whole string as a command (in this case 'package openssh-client') and it happens to look like English :) [21:26] "Sorry, I don't know anything about " [21:26] I wish it replied to a help or !help or commands or !commands or something. sigh :) [21:27] sarnold: yeah, i wonder if teh command list is too long [21:44] !info xv [21:44] Package xv does not exist in zesty [21:45] sorry mason, it hasn't existed in a while it looks like https://launchpad.net/ubuntu/+source/xv [21:45] I can hand-build it. But yeah, it's not there in Xenial either. [21:46] I'm thinking of using it to learn .deb packaging at long last. [22:08] mason: what is xv? [22:08] image viewer [22:08] it used to be the only thing we had twenty years ago :) [22:09] sarnold: oh i thought it was some other new thing -- given the reference to 'deb packaging' :) [22:09] nacc: https://en.wikipedia.org/wiki/Xv_(software) [22:09] sarnold: yeah, i've used *that* xv before :) [22:12] mason: xv is probably better (at this point) as a snap [22:12] you'd certainly be done packaging way sooner :) [22:13] :) === db` is now known as db === db is now known as Guest44806 [23:05] Ooh, a snap. That's a good idea too. [23:05] Probably worth learning both. [23:06] mason: for end-user applications, the snap is going to be a lot faster [23:08] Yar. Just saying. [23:40] sarnold: the more you know, huh? [23:40] ★✰⭐⭑⭒ [23:41] *throws sarnold into /dev/null* [23:42] sarnold: mind a PM? [23:42] sure