[09:37] Hi. Any idea, why the latest Linux HWE kernel is not installed in my Ubuntu 16.04.7 VM? [09:37] https://paste.debian.net/1162732/ [09:39] !info linux-image-generic-HWE-16.04 xenial [09:39] linux-image-generic-hwe-16.04 (source: linux-meta-hwe): Generic Linux kernel image. In component main, is optional. Version 4.15.0.115.116 (xenial), package size 2 kB, installed size 12 kB [09:39] lotuspsychje: Thanks. I remember why I removed it. [09:40] It depends on the microcode update packages, and *thermald* gets installed too. :( [09:40] *I remember why it was removed. (As I wanted to shrink the image and removed some of its dependencies.) [09:40] Do you have a solution for that? [09:43] It looks like, I have to manually keep it updated: [09:43] sudo apt install linux-image-4.15.0-115-generic linux-modules-4.15.0-115-generic [09:43] … if I do not want the dependencies in a VM. [09:43] *QEMU (KVM) VM [09:56] PaulePanter: linux-virtual-hwe-16.04 [09:57] or linux-virtual rather [09:58] there appears to be no hwe support for virtual in xenial [09:59] but you should probably plan to upgrade soon, anyways. [12:32] hi [12:33] What are my options for 12.04? - Can I downgrade from 16.04? - I'm working from the 12.04 iso and am trying to turn it into a custom cloud image. [12:33] * SamuelMarks needs 12.04 for an old version of Open edX; then he'll restore his backups and upgrade to the next version, and so on, until I'm at latest Open edX and Ubuntu 18.04 [12:46] SamuelMarks: you can still download a 12.04 cloud image from http://cloud-images.ubuntu.com/precise/current/ [12:46] !12.04 [12:46] Ubuntu 12.04 LTS (Precise Pangolin) was the sixteenth release of Ubuntu. !End-of-life was April 28th 2017. See https://lists.ubuntu.com/archives/ubuntu-security-announce/2017-April/003833.html for more information [12:46] yeah I've downloaded it [12:46] * SamuelMarks is giving up and running a VM on his laotop [12:48] yes, you'd definitely want to do this in VMs. [12:48] Yeah it's more that I was trying to get it running on IBM Cloud; but even after converting it to qcow2 it still didn't work there [12:57] Hmm, lots of hash errors on the apt-get update [12:57] * SamuelMarks tries `sudo rm -rf /var/lib/apt/lists/*` [12:59] those archives probably no longer exist where it's looking for them [13:00] old-releases.ubuntu.com [13:01] actually, precise is still on archive.ubuntu.com [13:05] Mmmm yeah that worked, ok proceeding with the bootstrap; hopefully won't error [13:10] gpg keyserver receive failed. Hmm, is pgp.mit.edu not around? [13:11] hopefully no more [13:11] unless they reinstalled it [13:11] there's keyserver.ubuntu.com [13:12] Hmm [13:13] nc -vz pgp.mit.edu 11371 suggests it's listening, though [13:13] * SamuelMarks is running this script https://github.com/edx/configuration/blob/open-release/eucalyptus.3/util/install/ansible-bootstrap.sh [13:14] Oh good, new error "key 69464050 not found on keyserver" === th4ntis_ is now known as th4ntis [19:59] good morning [20:12] so i have an interesting issue... colleague nuked the partition table on a hdd using parted on the wrong disk [20:13] good thing is, the disk in question is just a part of a raid1 data array, not part of the OS disk.. [20:17] so system layout is 1 disk for OS, 2 disks for data in software raid1 using mdadm. LVM for swap and /root on /dev/sda. 1 LVM group for data array. so we destroyed the partition on 1 disk on the data array, but that seems to be preventing the server from booting, lvm errors [21:20] fretegi: is this solved, yet? [21:20] if not, can you still boot to recovery? [21:33] tomreyn, i did get it solved, rebuilding the array as we speak, however i am curious why it wouldnt boot [21:34] tomreyn, the affected vg was seperate from the VG that contains / and the vg's are on separate physical disks too [21:34] fretegi: i suspect you had some file system backed by LV(s) in fstab [21:36] * unavailable LV(s) [21:36] oh yea both vg's has volumes that were auto mounted via fstab. in fact that is exactly how i got the thing to boot. i just #'d out the affected group, booted right up. thankfully was raid1 was was able to just re-assemble the raid degraded, remount everything and re-add the affected disk to md0. rebuildng as we speak [21:36] so any unavailable LV could cause LVM to just panic altogether? [21:37] panic? no. fail to boot fully? yes, i think so. [21:38] i think you can mark mounts as optional in fstab, but then you'll give up on a good way to detect errors you need to fix. [21:39] hmmm... i guess the initramfs may not like the unavail. filesystems [21:39] and you probably wouldn't want, as an example, logs to be written to the root file system rather than the separate /var/log file system, just because the latter failed to be mounted. [21:40] which ubuntu release is this aynways? [21:40] yea thats a good pt. fortunately this is a very simple server, i was able to thoroughly track down what that disk was doing almost immediately so with a little fiddling was able to solve it pretty quick. but i just didnt expect the lack of boot [21:40] tomreyn, good question... lemme check [21:41] 16.04 [21:42] tomreyn: Thank you. BigBlueButton 2.2 unfortunately requires Ubuntu 16.04. BBB 2.3 is going to be based on Ubuntu 18.04. [21:42] * PaulePanter does not fully understand that design decision, but that’s the way it is. [21:56] fretegi: you should upgrade [22:03] PaulePanter: there are test builds for 18.04 now, if you'd like to try: https://github.com/bigbluebutton/bbb-install/issues/37#issuecomment-661347668 [22:04] but they're certainly late on this. [22:04] tomreyn, yea little old eh haha [22:05] 16.04 still receives free support, but just until april [22:07] yea still, your right, should upgrade [22:15] tomreyn: unless ESM is used which adds another 3 years of security updates. One can use ESM for free up to three machines. [22:17] Ubuntu members get up to 50 machines by the way [22:30] Maik: for non-commercial use, yes. fretegi was talking about their "collegue", which makes me think commercial. but sure, they can buy. [22:30] or just upgrade. or both. [22:59] tomreyn: thanks for clarifying, i must have missed reading that. My bad. === Luke_ is now known as Luke