[08:17] Bug #1757067 opened: Intermittent failure: test_estimated_runtime_sets_start_time_when_unavailable === frankban|afk is now known as frankban [14:27] Bug #1757153 opened: [2.4, UI, regression] Run time for commissioning/testing scripts is incorrect [14:33] Bug #1757155 opened: [2.4, UI] Storage firmware version is mispelled & should be part of Name | Model | Serial switcher [17:36] Bug #1757210 opened: [2.4, regression] Cannot compose a machine inside === frankban is now known as frankban|afk [18:52] hi all - looking at https://docs.maas.io/2.3/en/installconfig-network-dhcp and trying to configure MAAS to use my external DHCP server; does not look like the dos correlate to my MAAS UI (running v2.3.0) [18:52] *docs not dos [18:54] I was able to configure my external DHCP server to next boot IP of the MAAS server, and used BIOS filename 'pxelinux.0', and was successfully able to netboot a machine from MAAS [18:58] However, got a cloud-init message that the node "failed to enlist system maas server" and powers off [18:58] And (unsurprisingly) no node shows up in the MAAS UI [19:32] wdennis: /etc/maas/rackd.conf make sure it is pointing to an ip instead of localhost [19:39] Bug #1757242 opened: [2.4, UI] Need some space between green "Compose machine" button and the pod details table. [20:22] roaksoax: It was indeed pointing to 'localhost' - corrected to put in IP. Do I need to restart any service etc? [20:33] wdennis: yes, rackd [20:33] sudo service maas-rackd restart [20:58] Bug #1757210 changed: libvirt drops connection when attaching a disk (breaks MAAS pods) [20:58] Bug #1757262 opened: [2.4, pod] KVM pod doesn't fail if it fails to attach disk [21:10] Bug #1757210 opened: libvirt drops connection when attaching a disk (breaks MAAS pods) [22:00] roaksoax: Thx for the info... restarted the maas-rackd service, then retried the PXE on the host, but still not registering... Can you point me to logs (either on PXE client OS or server) that I can look at? [22:04] Found server-side logs in /var/log/mass/*, looking at rackd.log, seeing these errors periodically: https://bpaste.net/show/1de12b924a25 [22:05] Region controller and rack controller are on same node (default MAAS install via Ubuntu ISO...) [22:07] wdennis: /var/log/maas/rsyslog/* [22:07] wdennis: if it is a non-known maas machine, check the logs under maas-enlisting-node// [22:13] roaksoax: No files in /var/log/maas/rsyslog/ on server [22:17] wdennis: do yo ahppen to have a console log ? [22:19] roaksoax: On the client? Unfortunately no [22:28] roaksoax: Aha - looking at 'cloud-init' messages captured in syslog on the client, I see that the server url as used by maas-enlist has the wrong IP address in it... [22:28] Is that se somewhere on the server? [22:29] *set [22:29] wdennis: ok, see if it is the same IP as /etc/maas/regiond.conf:maas_url ? [22:30] roaksoax: Yup, it was the wrong IP there - corrected. Restart a service needed? [22:31] And, any other places I should look for incorrect IPs? [22:33] A grep for my IP block in /etc/maas/* shows the correct IPs now: https://bpaste.net/show/65ba3791ed38 [22:40] OK, looks like that was it - the node successfully registered now :) [22:40] Lesson learned - don't install a new MAAS server on a temp DHCP address, then re-IP it later... [23:20] How may I customize the Ubuntu install that the "Deploy" action lays down on disk? (partitioning, add'l pkgs to install. etc?)