=== led_ir23 is now known as led_ir22 [00:16] How hard is it supposed to be to get iommu to work 100% successfully. Allready added iommu=on and intel_iommu=pt in kernel config, and update-grub, rebooted [00:16] But still Error starting domain: unsupported configuration: host doesn't support passthrough of host PCI devices is the happy message ive got [00:17] libvirt kvm === led_ir23 is now known as led_ir22 === led_ir23 is now known as led_ir22 === led_ir23 is now known as led_ir22 [05:35] good morning [06:24] can some one help me with some basic (???) problem ,i am trying add parameter in file /sys/bus/pci/devices/0000:01:00.0/driver/unbind but i am not able to edit /write to file even file permisson is 666 and it give message/error when i type file unbind [06:24] unbind: ERROR: cannot read `unbind' (Input/output error) [06:25] i am trying this on 16.04.4 and beta 18.04 [06:25] what wrong i am doing [06:26] samba35: What is your end goal with this operation, what are you trying to do? [06:26] i want to configure pci passthrough where i want to add device id to this file [06:26] i try echo command also [09:30] Hello everyone, imagine that I have an ubuntu xenial server with 2 net itfs which are on separate networks and for which 2 DHCP servers answer. Any clue why the "secondary" DNSes always take precedence over the ones received from the 1st/main itf ? [09:31] are there any options I need to set in the DHCP servers so that one always take precedence ? any other way to solve this ? [09:39] olivierb-: should be possible to configure the dhcp client to disregard dns entries from one of the interfaces [09:41] olivierb-: looking at the dhclient.conf manpage, theres interface { ... } section where you could supply different options for different interfaces. [09:56] blackflow many thanks, will definitely have a look at this [10:41] olivierb-: you can ignore dns-entries in your dhclient-config. === rmk` is now known as rmk === Kamilion|ZNC is now known as Kamilion [13:24] coreycb: oh, great, I've been meaning to ping you about neutron-dynamic-routing releases but it looks like you're already on it https://bugs.launchpad.net/bugs/1750121 [13:24] Launchpad bug 1750121 in neutron-dynamic-routing (Ubuntu Bionic) "Dynamic routing: adding speaker to agent fails" [High,Triaged] [13:25] frickler: yes. i may not be able to upload until friday though as bionic is 1 day from release and in final freeze. [13:25] thedac: fyi ^ [13:26] coreycb: I'm more interested in the Pike update anyway, but no need to hurry, I do have local builds for now [13:28] frickler: ok. it needs to land in pike+ releases first. definitely in progress though. [14:01] udebs, are they only used in the installer environment? Or somewhere else? [14:01] maybe in small devices, routers, that kind of thing? [14:37] ahasenack: https://wiki.debian.org/udeb [14:38] ahasenack: Looks like they're solely for the installer's use. [14:38] I haven't seen them employed anywhere else, though that's hardly conclusive. [14:38] ok [14:38] I asked for more information in the bug report I got [14:38] about a missing symbol/link step in a udeb package [14:58] nacc: git-ubuntu sync is scheduled in 90 minutes. Are you interested in attending these still? [15:24] teward: you are famous: https://news.ycombinator.com/item?id=16921452 [15:29] I am? What've I missed o.O [15:29] teward: top comment on there is rbasak. :) [15:30] ah, nice. [15:30] :) [15:39] teward: added an entry into the release notes, if you wanted to review or wordsmith: https://wiki.ubuntu.com/BionicBeaver/ReleaseNotes [15:52] hey!, Is there any up-to-date vagrant ppa? Just installed 1.9.1 from the repos and current version is 2.0.4 [15:54] dannax: bionic has 2.0.2 [15:55] dpb1, thx! [15:56] rbasak: i can't today, I don't think [15:56] rbasak: in the future, yes [15:57] nacc: OK. I think you should have an invite to your personal address. [15:57] rbasak: received thanks === Bilge- is now known as Bilge === AndyWojo_ is now known as AndyWojo [16:49] I wonder where I can document in https://wiki.ubuntu.com/BionicBeaver/ReleaseNotes that cifs-utils (i.e., mount.cifs) now defaults to SMB2 or higher instead of SMB1 [16:49] it's a client tool, but desktop isn't client necessarily [16:50] ahasenack: just adding a 'samba' entry would be good enough [16:50] under server [16:51] or maybe under "known issues", and link to the bug I am troubleshooting? [16:55] * ahasenack checks if that changed in artful already [16:56] hi all. I'm back to trying to fix my boot partition. I've been running: dpkg --configure -a and getting part of the files rebuilt until I run out of space and empty older versions of the files in the boot partition [16:57] I'm not actually sure what I'm doing or if I'm making progress, it will download one release newer each time I do this [16:57] Maxel: can you clean up space in general first? [16:57] I'm not sure what I should be tampering with on the boot partition [16:58] I should preface this with I believe some time ago I stupidly deleted everythhing from my boot partition because apt upgrade wasn't working for me [16:58] so I'm trying to get to a good state [16:58] well, ya, that's not great [16:58] :) [16:59] Maxel: these tips: https://askubuntu.com/questions/345588/what-is-the-safest-way-to-clean-up-boot-partition [16:59] luckily I haven't restarted the vm and I am hoping I can recover it [16:59] honestly, read that over, it's good info [16:59] even talks about what to do if you can't do it through apt [17:01] dpb1: TJ- has a script that does the right thing in general [17:01] let me see if i can find the link [17:02] http://iam.tj/projects/ubuntu/apt-autoremove-old-kernels [17:03] * dpb1 posts that as an answer [17:04] nifty [17:04] should I be able to run this in my current state however? [17:04] where I deleted the actively running kernel? [17:05] Maxel: that answer with 'Note: this is only if you can't use apt to clean up due to a 100% full /boot', I would read through and digest it [17:05] oh, yeah down an answer. will do :) [17:09] ok, I believe this is essentially what I am doing. one difference is flickerfly says to delete the old images while I'm just reducing them to 1 byte in size [17:09] not sure if one way right or wrong [17:14] Maxel: no real difference, given the truncated image is unusable [17:17] dpb1: looks good to me (for the release notes blurb). thanks. [17:21] ok, so I am going through the answer on askubuntu, and when I perform the apt-get -f install, it still doesn't have enough space [17:21] Maxel: what are the numbers [17:21] (df -h /boot) [17:22] here: /dev/sdb1 472M 469M 0 100% /boot [17:22] that is indeed small [17:23] I don't remember defining the size on install [17:23] du -ms /boot/* | sort -n | pastebinit [17:23] should be plenty big if autoremove is used [17:25] whaaa, I had no idea pastebinit was built in [17:25] slick [17:25] http://paste.ubuntu.com/p/Wx3jrSFzZV/ [17:26] Maxel: you can also do | nc termbin.com 9999 [17:26] ok, same thing you did to the kernel files, you need to do to the initrd, one at a time, oldest first. [17:27] I've just been running this: echo "" | sudo tee /boot/*-4.4.0-{62,66,70,71,72,75,78,79,87,89,98,101,103,104,108}-* [17:27] to clean everything up [17:28] ok, that's fine, but it's missing your initrds it seems, unless you haven't ran it yet [17:28] since that's where your space is going [17:31] it truncates the initrd files, but when I run apt-get install, it re-downloads the files [17:31] Maxel: after you truncate the initrds, you should autoremove [17:32] then do an install, i think [17:32] just apt-get autoremove you mean? [17:32] sudo apt-get purge linux-image-, or autoremove, if you can. [17:34] ah right, the initrds are being generated again [17:34] with your -f install [17:34] autoremove seems to be working [17:34] basically, you have to tell apt to stop installing those broken packages [17:34] which will always put back all those files [17:34] autoremove gets rid of those unneeded packages [17:38] ah ha, looks like it's in a good state finally [17:38] now the kernel I'm currently running is still truncated though, I assume this would be an issue on boot? [17:39] oh, err actually the initrd is there, but the other images are not [17:39] sudo apt install --reinstall linux-image- [17:40] depending on which package it actually is there. you want the full package name though [17:40] on my box, I have linux-signed-image-4.4.0-121-generic [17:41] and linux-image-4.4.0-121-generic [17:43] neat, looks good, now I'll see if I can update [17:46] so doing a apt-get upgrade I'm getting an error. grub failed to install 3 devices [17:47] pastebin if you can [17:49] it's pretty ugly: https://pastebin.com/UcqMkxWQ [17:49] it has a ui for this error message [17:50] Is upgrading to 18.04 as easy as just changing all "xenial" lines to "bionic" and running apt-get update && apt-get dist-upgrade [17:50] Like in Debian [17:51] sec there is a link [17:51] https://help.ubuntu.com/community/Upgrades [17:51] thanks [17:52] mtl_: you *can* do that, but d-r-u is 'better' [17:53] rbasak: how does the importer review look? I might have a few moments this aftenroon to refresh my cleanup MP [17:56] nacc, my virtual server provider provides cloned ubuntu repositories, can you tell d-r-u to use them instead of official mirrors? [17:56] mtl_: no idea, sorry [17:56] nacc, ok. [17:57] mtl_: VPS (by design) seems easier to create fresh, then migrate data then to upgrade in place [17:57] but that's my opinion [17:58] I think d-r-u does a s/xenial/bionic/ more or less, it doesn't change the mirrors. emphasis on think. [17:58] it does remove non standard repos though, so... I wouldn't be surprised if I was wrong [17:59] yeah, i can't remember anymore [17:59] I have the cheapest package and even snapshots are not available [18:00] if it's a VPS then the answer I'd give is 'It Depends on the way it's Virtualized" [18:00] if it's OpenVZ virtualized you should avoid upgrading entirely [18:00] and I have lots of own configs.. :( [18:00] if it's KVM then I'd say "YMMV, I'd suggest 'fresh systems' and then data migration." [18:00] teward: good point [18:01] it's KVM === Guest92578 is now known as devil === devil is now known as Guest47087 === Guest47087 is now known as devil_ [18:11] this looks like the only note on the grub issue: https://askubuntu.com/questions/315207/grub-failed-to-install-to-the-following-devices-error-on-upgrade [18:12] I'm not sure if I should stop the upgrade [18:25] if I choose to not proceed from that grub error, it takes me back to a screen to decide which drives to install grub on [18:25] where if I choose ok, goes back to the error screen [18:36] rbasak: dpb1: anything big come out of the g-u sync? [18:46] how can I figure out which drive is designated boot? [18:49] Maxel: could you pastebin lsblk's output? [18:50] I'm stuck in a loop of error saying it couldn't install grub, and then choosing which partitions to install it on [18:50] I'm worried if I disable the drives I'm getting an error on that I wont be able to boot [18:50] Maxel: you generally don't want grub installed on a partition [19:41] nacc: we skipped it [19:50] nacc: the importer seems to be stuck on qt4-x11 and texlive-extra now. [20:13] rbasak: both are blacklisted now [20:13] rbasak: fyi [20:13] rbasak: so may be worth killing the import subjobs and seeing if it progresses [20:59] nacc: did you blacklist them before, or just now? === mtl_ is now known as mtl === devil is now known as Guest418 [22:15] rbasak: in master before i left [22:16] late-ish on Monday, so after your snap build presumably [22:50] Ah, OK. [22:50] I'll finish fixing the snap build tomorrow hopefully. [22:56] dpb1, rharper: the alternate server installer URL could do with updating in https://wiki.ubuntu.com/BionicBeaver/ReleaseNotes#Ubuntu_Server [23:00] Oh, it's not out yet. [23:01] rbasak: thanks -- yeah let me know if you need anything [23:12] rbasak: yes, there will be a round of updates after they do final image builds (hope that is happening now)