[03:12] My http site does not redirects to httpS. Is there anything wrong with configs? https://pastebin.mozilla.org/9083482 [03:17] nevermind ^ [09:38] jamespage: coreycb: maybe one of you can help with investigating why upgrading UCA from pike to queens seems to have broken upstream tests: https://bugs.launchpad.net/openstack-gate/+bug/1765638 [09:38] Launchpad bug 1765638 in OpenStack-Gate "legacy-tempest-dsvm-full-devstack-plugin-ceph job failing" [Undecided,New] [10:08] does anyone know why i'm encountering the following error trying to deploy ubuntu openstack? [10:08] conjure-up hangs on the first image (one sec) [10:08] and MAAS displays the failure events seen in the second image (one sec) [10:10] https://my.mixtape.moe/cxerkq.png [10:10] https://my.mixtape.moe/jwvier.png [10:48] lilac, are the interfaces on the node correctly tied to the PXEnetwork? I had something similar, the node was discovered/commissioned. But when deploying, the node did the first install (OS), but then it switched the interfaces so, it could not reach the service providing information.. [10:48] could you clarify what you mean by "correctly tied"? [10:49] PXE boot works fine [10:49] seemingly DNS, NAT, DHCP, PXE all work fine on the private/pxe network [10:49] parlos: [10:49] for instance, maas is on 10.30.0.x fabric0, after 'deploy' during install, the 10.30.0.x address is now assigned on fabric1, on the node.. [10:50] hence, unless fabric 1 routes the 10.30.x traffic its going to timeout. [10:51] hmm okay [10:51] i'll take a look [10:51] thanks [10:51] this is what happend to me when I did my first juju charms... (afaik conjure-up is just a frontend...) [10:52] one more hint,. also check storage config. . [10:53] I had cases where the bios boot was set to PXE, then boot drive b... --> but as the normal installs endup on sda (first drive), the reboot complains data, treating drive as raw... [10:53] GL & HF [10:59] thanks parlos [11:01] parlos: where would i be checking this? it looks fine in maas [11:02] I'm doing it in the maas ui, interface and storage tab... ... probably looks fine, [11:03] if you can check the console of a device as it is deployed/powered on, when your deploying/conjure-up [11:11] okay so i SSH'd into the node that i presume is being prepared as the juju controller [11:11] its network interfaces look fine based on MAAS interfaces tab and `ip addr` [11:14] ohhhhh [11:14] i think i noticed a misconfigured machine [11:14] interfaces were connected the wrong way around [12:09] Hi guys! [12:09] what to read bout dovecot? [12:09] is this book good? [12:09] https://www.amazon.com/Dovecot-servers-enterprises-Heinlein-2014-09-19/dp/B019NEB3W2/ref=sr_1_2?ie=UTF8&qid=1524226136&sr=8-2&keywords=dovecot+2014 [12:09] I've got it [12:09] Now i have depression, will read for calm nerves... :( [12:12] couldn't have install mail server... [12:12] dovecot more complex even than postfix [12:15] have you tried mail-stack-delivery? It sets dovecot up for you with postfix [12:15] could be a starting point [12:22] ahasenack: no, I think for a while for VPS enough one postfix, [12:22] sites will able to send mails and nice [12:23] mail server is needed nobody, I want to set up it for myself [12:23] and experience [12:24] I saw video there guy installed DNS server seems on virtual machine and was able to send massage on mail server [12:25] it's not exactly [12:25] I want too install DNS server, create two domain on one domain put server and other... [12:25] something like real model of internet [12:30] maybe you are trying too much at first. Even with a quickstart package like mail-stack-delivery, you would be able to poke around once it's setup, and see how things are done, and use it as reference when installing a fresh one by hand, package by package [12:31] ahasenack: ok, I'll save this mail-stack-delivery in file and will try it too [12:32] ahasenack: just don't have common picture about dovecot [12:32] but have many questions [12:32] after that book, if read it fast you would know all futures [12:33] if you really want to understand things, there are not shortcuts [12:33] each one of these things have their own book to read [12:33] postfix, dns [12:33] ahasenack: in #dovecot channel very active support, people very fast react if compare with #postfix :) [12:34] imap/pop, etc [12:35] ahasenack: I know IMAP - internet mail access protocil, POP - post office protocol. [12:35] yes, books help, but bad that book is written by author of dovecot, and it seems not good because author can write using difficult words. Better when it is written by ordinary users [14:07] I am having problem with keyboard layout I choosed when Installing ubunt-server. It types all the alphabets and alphanumeric characters the same that are given on my keyboard. But when I type symbols they all are different. So,How to change keyboard layout? [14:12] Hello, Is their anybody? [14:12] patience is a vritue you know [14:13] pankaj_: yo might want to look at the answers on https://askubuntu.com/questions/342066/how-to-permanently-configure-keyboard - one of the answers might help. [15:05] Hi all, I've run into an issue with ubuntu package manager being able to install packages where a package has a dependency and it doesn't get those packages automatically [15:06] I'm trying to install keepassxc right now, but I remember it happening when I was trying to install another package in the past [15:08] that usually happens when you change your sources to ones that have incompatible packages [15:13] qman, so the issue becomes whoever is creating the package I'm trying to download? [15:14] if you are using unofficial sources, then either the source is not for the version of ubuntu that you have, or the source is broken [15:27] I am trying to figure out which version(s) of apache solr will be included in the upcoming 18.04 LTS server release. Can someone point in my the correct direction or if you know...? [15:44] digs: if yuo mean solr-jetty or solr-tomcat as listed on https://wiki.ubuntu.com/Solr, then I believe 3.6.2 is the base version with additional package-level specific changes from Debian in it. [15:44] not 100% sure of this being what you need, but that's the closest answer I think to what you need. [15:45] digs: "rmadison $PACKAGE_NAME" will tell you which versions ship with which Ubuntu release [15:46] digs: seems like 3.6.2+dfsg-11, see https://paste.ubuntu.com/p/3sQmvQMdRY/ [15:51] Maxel: Are you, by any chance, installing a downloaded package via dpkg instead of from a repository via apt or apt-get? dpkg does not install dependencies. [15:51] ScottE, no, I am doing an apt-get, granted I had to add a private repository to enable it [15:51] this is what I was trying to do: https://keepassxc.org/blog/2017-10-25-ubuntu-ppa/ [15:52] Then your problem is likely with the third party repository. Ubuntu developers don't support third party repositories. [15:52] You should take it up with their support channels. [15:52] OK, just checking Maxel as that does come up sometimes [15:54] rbasak, I did talk to people on the irc channel for keepass. they wanted me to install the appimage myself, which I could but I like having the management left to the package manager for updates and such [15:54] I also ran into this same problem with a different application, eclipse che, and after seeing this sort of thing happen multiple times wanted to understand why it is happening [15:55] It's very difficult to maintain external repositories. [15:55] It's fundamentally a broken model. [15:55] External _apt_ repositories, that is. [15:55] Nevertheless, it's often possible to fix, but the fix generally has to be in the third party repository, which Ubuntu developers have no control or say over. [15:56] The specific problem is usually different every time. [15:56] It's compounded when users add multiple different third party repositories, since they tend to be uncoordinated and the combinations can cause problems. [15:58] Ubuntu uses snaps to solve this general problem. AppImage and Flatpak are different approaches to the same general problem. [15:59] hmmm, so either manage the applications myself manually or just do without, or convince the developer to go through the approved ubuntu process [15:59] jamespage: taking a look at networking-arista. it's not building in bionic b/c it's py3-onlly. [16:00] and needs python3-neutron [16:01] Maxel: to reach Ubuntu users a snap would be best. Then it'd just be another app in the software center. [16:01] (and be visible by default, etc) [16:01] I don't know what keepassxc needs from a security sandbox standpoint though. [16:01] Maxel: and it's a tractable problem for a single person to ship an assembled app with snaps. [16:01] (eg. to talk to a Firefox plugin or whatever) [16:02] rbasak: system access to where the database that it uses to store passwords is. which in a sandbox can be... tricky... to activate [16:02] there is a keepassxc snap if I remember correctly, but it's got its own set of issues [16:02] it's really not for most apps and debs (into the ubuntu archive), or at least will take you a very long time before you get to that point. [16:02] teward: a snapped version of this app should store it in the correct XDG directory. So the location of the database should not be a problem. [16:02] rbasak: by 'database' i mean a flat file [16:03] Sure [16:03] so if user is trying to use a db they already have in the snap it's not able to find it on, say, external media or such [16:03] (I had the same issue with keepassxc's snap a long while ago) [16:03] There's an external media interface. [16:03] but either way, it's beyond our control to fix issues :P [16:03] Maxel, or just wait a short while: keepassxc will be included in bionic :) [16:03] https://packages.ubuntu.com/bionic/keepassxc [16:04] *returns to beating LXD with a pipe to make it behave* [16:04] But the "difficulty" is sort of the point. Other snaps shouldn't be able to get to your password database. So making a snap for the password manager necessarily involves putting the database in a special place. [16:38] rbasak: dpb1: nacc: Just to keep you all in the loop: Perl autopkgtests were holding up the nginx 1.14.0 migration out of proposed, those got fixed (I think) and NGINX 1.14.0 is now available in Bionic. Which means we're on the stable track of NGINX again, for the LTS, without a post-release MRE. [16:39] (yay! we didn't have a repeat of the 16.04 cycle where NGINX stable was released after we had already released 16.04!) [16:39] many thanks to you guys for helping out with the merge from Debian, that was the biggest hurdle in getting this up to date and ready to go (I was super busy heh...) [16:40] \o/ [16:41] great news teward [19:07] so I actually tried to just do a apt-get upgrade and I'm getting unmet dependencies error [19:14] ahhh shoot, I think it is because my boot mount is full [19:15] that one's never fun to debug [19:15] but I wouldn't have expected unmet deps to result [19:15] full /boot happens... a lot [19:16] yeah, not sure what I should be doing. I used default settings when installing [19:16] it looks like it doesn't even have 500mb on there [19:16] can I just allocate more space somehow.... [19:18] /boot is kind of a tricky special case as it tends to live outside of the volume manager [19:18] so resizing it is not trivial [19:23] what I don't understand is why we're still seeing full /boot .. I thought apt grew knowledge how to keep N kernels around, where N was somewhat reasonable, back ~precise era [19:23] I dont think I've seen that working unless I've been doing "apt autoremove" once in a while myself. [19:24] on yum, this just happens with kernels ;) [19:24] $ dpkg -l 'linux-image*' | grep ^ii | wc -l [19:24] 5 [19:24] I can assure you I don't go to any effort to manage my kernels manually [19:25] I'm a very lazy person [19:25] ohhuh. my other rig's got ten installed. [19:25] perhaps it depends on how you update, like update-manager or apt directly? [19:26] I've not looked too closely at configuring this [19:27] oh possible, I don't dist-upgrade that machine by hand all that often [19:27] it does seem like a lot of peoples /boot's dont get managed properly for whatever reason [19:27] and my gut feeling is that this is a bigger problem on ubuntu than in redhat/fedora land [19:29] heh, apt-get autoremove brought that down to .. six. that still seems like a lot. :) [19:30] it may also depend on how you installed some of the kernels, ie they are flagged manually installed on not by dependency. [19:30] ineffable magic. got it. [19:32] s/ on/ and [19:34] yeah, in my experience that has never worked correctly [19:34] I wrote my own script to deal with it [19:34] removes all kernels except the current running one and the latest version [20:07] sorry, my internet seemed to disconnect at the critical time of discussion over boot [20:08] so in my case, am I best off just trying to clean up my boot partition? [20:15] jamespage: beisner: xenial-queens-proposed passes smoke tests. i'll plan to promote to -updates first thing monday. [20:16] ack thanks coreycb [20:16] Maxels: you have to apt-get remove the extra kernel packages that you don't need [20:16] Maxels: if you can't due to not enough space to even do that, you can overwrite some of the files you're about to remove with zero-length files and then apt-get remove them [20:17] Maxels: be very careful doing this, though, as to not overwrite or remove files that you need [20:18] is there a guide that describes exactly what I should be doing? [20:18] not that I know of [20:18] last time I did this I think I deleted my document that maps drives, can't remember what it's called [20:19] what you should do first is figure out which kernels you want to keep [20:19] you can then compare this against what's installed to know which ones to remove [20:19] https://askubuntu.com/questions/345588/what-is-the-safest-way-to-clean-up-boot-partition [20:19] this looks like what you're describing [20:19] I typically would keep the actively running kernel and the latest kernel for each type of kernel you have [20:20] https://pastebin.com/w4xu4pXj [20:20] so there's what I've got [20:20] once you know which ones to remove, you'd apt-get remove linux-image-3.something-generic [20:21] ok [20:21] sudo apt-get remove linux-image-[67]?-generic [20:21] that's regex to remove everything in the 60 and 70 versions [20:21] er [20:21] sudo apt-get remove linux-image-4.4.0-[67]?-generic [20:22] if that works, you'll be okay, and just continue removing kernels that you don't want [20:22] yeah must be what you warned about, wont let me because it says it has unmet dependencies [20:22] so I should zero out one of these packages bytes? [20:23] how many bytes are free on /boot ? [20:23] 0 [20:23] ok, yeah [20:23] so, ls -lh /boot [20:24] then zero out some of the files that exist specifically for the older kernel versions you want to remove [20:24] such as [20:24] do you mean make a new file with the same name with no content? [20:24] echo "" | sudo tee /boot/vmlinuz-4.4.0-68-generic [20:24] yeah, there's lots of ways to do it [20:25] the key is that the file has to exist by the same name, otherwise apt can't remove the package [20:25] ok, now these images don't look like the images listed from the prior command [20:25] https://pastebin.com/b0GUu2BD [20:26] ok, looks like some of them have already been removed manually [20:26] which is going to be a problem trying to use apt [20:26] yeah that is probably from my failed attempt to do this [20:27] so, pick one of those files with the version in it to do this with to free up some space [20:27] then touch each file that should be there [20:27] so a vmlinuz for each version that's installed, an initrd for each, a system.map for each, etc [20:27] then apt should be able to remove them again [20:27] how can I see which should be there? [20:28] they follow the same naming convention [20:28] so you notice there's like 5 files for each kernel version [20:28] so touch files by the same name but with the versions from your earlier paste, the versions that apt sees [20:28] the uname -r version is 4.4.0-81 though, which isn't listed here [20:29] the initrd files are the biggest so I'd suggest zeroing one of those [20:29] yeah, it was probably already removed, unfortunately [20:29] so don't reboot [20:29] ahh, ok [20:30] ok, zeroed out a couple files [20:31] ok, so now you should be able to touch files for each version so that all the filenames are there again [20:31] then you can apt-get remove as above [20:31] I have to just touch a file in the boot and it'll recover the image I need? [20:31] it won't recover the image [20:32] it will just create a file that exists by the right name so apt can remove the package [20:32] after you remove the package, you can then install it again [20:32] but only after you free up enough space [20:32] by removing all the ones you don't want [20:32] oh, you mean make the placeholder file for each of vmlinuz, system.map etc.... [20:32] yes [20:33] I see 5 of em in total [20:33] does that seem right? [20:33] yeah, 5 different kinds of files looks correct [20:35] https://pastebin.com/L6Czpjgv [20:35] alright, I think I've got placeholder files for each [20:36] yeah, like that, but you have to do it for all the installed versions [20:37] how do I get a list of installed versions? [20:37] is it just every initrd.img file here? [20:39] in the apt or dpkg output from before [20:39] https://pastebin.com/w4xu4pXj [20:39] each version in that list [20:40] ahh [20:40] ok, a bit of work I gotta do here [20:49] qman, does this look like what you meant? https://pastebin.com/iUHTaSKs [20:50] yep [20:50] you should be able to apt-get remove stuff now [20:50] if you still can't, might need more apt troubleshooting [20:50] can I remove these old versions? [20:50] to free up the boot space? [20:50] with apt-get I mean [20:51] yes [20:52] yeah, still getting the unmet dependencies problem [20:53] ok, paste the apt output [20:54] https://pastebin.com/WT5FUFdJ [20:55] sorry, that was incomplete: https://pastebin.com/mXtbnaZj [20:56] ok, try the apt-get -f install [21:00] getting a couple messages taht say no space left on device that are concerning [21:00] still doing stuff though [21:01] total output: https://pastebin.com/zBR3phqA [21:01] ok [21:01] so it's trying to install more kernels and failing [21:01] because there's not enough space [21:02] so I'd zero out more of those old initrd files and try again [21:02] well it just downloaded some of the ones I had 0'd out before again [21:03] it's trying to install 108 and 119 [21:03] so any other ones you zero out shouldn't be overwritten [21:04] hmm, try this [21:04] dpkg --configure -a [21:04] then try to apt-get remove some of them [21:05] oh, I tried running apt-get -f install [21:05] should I kill this? [21:06] no, let it go [21:06] see what it does [21:07] looks like the same output: https://pastebin.com/sq0knzEG [21:08] ok, then I'd wipe out more files and try dpkg --configre -a [21:08] and yeah, it re-downloaded those initrd.img files [21:08] ok [21:10] At this point, I'd be tempted to move /boot to the / and fix the apt situation [21:10] then move /boot to it's own partition [21:11] Maxels: it's something you might want to consider ^ [21:12] I've run into this problem many times, you don't have to mess with your partitioning to fix it [21:12] but it does take some doing [21:12] yeah, I'm not sure exactly how that would work logistically but I can do that [21:12] I ahve a like 8tb raid drive attached to this thing, so plenty of space [21:12] qman: I'm not specifically talking about messing with parts, just a temporary way to get some free space then clean /boot proper and revert the hack [21:13] looks like basically the same result: https://pastebin.com/aLXkBQdR [21:13] sorry for jumping in ;) [21:13] Maxels: the file it's looking for doesn't appear to be there, try touching it [21:14] /boot/vmlinuz-4.4.0-87-generic [21:14] I see 81 and 89 but not 87 in your list [21:15] ahh, I should add that for all of them? [21:15] yeah [21:15] ok, added [21:15] try the dpkg command again? [21:15] yeah [21:17] it ran much faster: https://pastebin.com/YHjCbC5B [21:17] still missing something though [21:17] it ran out of space again [21:17] what's there now, in ls -lh [21:19] I just blanked a couple initrd.img files [21:19] teward and sdeziel - belated thanks re: solr [21:19] trying dpkg again [21:19] digs: np [21:19] looking better so far [21:19] ok, it's trying to process 4 kernel packages so it looks like it's going to try to build 4 initrds [21:19] so you need at least that much space to get them all fixed [21:21] ahh, so close [21:21] ran out of space again though... not sure what else I can remove [21:21] if it succeeded on a couple of them, you can blank those ones [21:21] it'll only try to process the remaining ones it wasn't okay with [21:22] https://pastebin.com/NUJz7CzG [21:22] 103 104 and 108 should be doable [21:22] 101 too [21:23] alright, trying again with those freed [21:25] darn [21:25] did dpkg get any further along or is it still the same packages [21:26] linux-image-4.4.0-83-generic [21:26] linux-firmware [21:26] linux-image-extra-4.4.0-87-generic [21:26] linux-image-extra-4.4.0-83-generic [21:26] linux-image-generic [21:26] linux-generic [21:26] sorry, didn't meant o post each line [21:26] pastebin wont allow any more pastes from me [21:27] hah [21:27] those appear to be the same ones as before [21:27] pastebinit can send stuff straight to a dozen different pastebins [21:27] ok, that's good, it got through some of them [21:27] I'll check it out [21:28] there's about 5 more initrds in there, you can probably get rid of some more and try again [21:28] I appreciate the help, gotta duck out for a bit. I'll see if anyone is around in 20 [21:28] in the 70-90 range [21:29] once dpkg is happy, apt-get remove should work again [22:14] when creating a boot-resource in maas. how do I determine the "name="what name?" [23:36] Hey: "Although the backend supports multiple boot sources, MAAS itself uses a single source. If multiple sources are detected the web UI will print a warning and will be unable to manage images." https://docs.maas.io/2.3/en/installconfig-images [23:36] oh. that's boot source, not boot resource. never mind...