[01:26] Bug #1590121 changed: [xenial][maas beta5] [arm64] system tries to enlist when I commission. [01:26] Bug #1617044 changed: [2.1] View Full History on Events doesn't show anything [04:33] Bug #1662111 changed: when deploying 'virsh' type of machine(KVM), it repeats deploying constantly [04:33] Bug #1673854 changed: Commissioning selects wrong boot drive on 3-disk KVM === frankban|afk is now known as frankban [10:18] Bug #1691434 opened: Can not apply stage final, no datasource found! Likely bad things to come! [10:27] Bug #1691434 changed: Can not apply stage final, no datasource found! Likely bad things to come! [10:36] Bug #1691434 opened: Can not apply stage final, no datasource found! Likely bad things to come! [12:07] Hi, quick question [12:07] How do i connect to a deployed machine with ssh ? [12:08] I get permission denied publickey [12:13] int-0x21: if you have added your public key to your account in maas, it should be possible to ssh to ubuntu@ [12:15] Ah i might have confused it with the api key [12:18] Hmm adding the key i however get invalid key. [15:56] Hello everyone - running MaaS on Ubuntu 16.04 - 2.2.0~rc3 (upgrading right now). Running into an issues that I'm unable to find the answer to - we are having an issue with our network and DHCP helpers currently. As a result of this, we have discovered that with this DHCP Helpers issue, when we reboot a node that has been provisioned by MaaS, the node will not boot from the drives. Console says that no UEFI devices found. [15:57] Why would this cause an issue? Does MaaS indeed have to be online and fully accessible at all times for a server, already provisioned, to be successfully rebooted? [16:04] alanmac: no. That's a known bug [16:05] alanmac: https://bugs.launchpad.net/curtin/+bug/1680917 [16:08] roaksoax: Thanks! Is there a work around? Not seeing it in the ticket. [16:09] alanmac: the fix is here: https://code.launchpad.net/~blake-rouse/curtin/uefi-clear-reorder/+merge/323875 requires redeployment [16:09] alanmac: for work around, would be to fix the boot order [16:09] or get maas back up, get machines to boot, fix the bootorder to boot from PXE/IPv4 first [16:13] we will probably just do a re-deploy once our network is up and running properly again (new environment with new SDN vendor). PXE/IPv4 is set as first boot. [16:16] roaksoax: its worked well, as did the fix code pre rc4 [16:17] there will be a new bug soon about image upload via MAAS [16:17] we did not submit any info yet [16:25] roaksoax: Could we just update grub on the existing working systems? adding --no-nvram to "GRUB_CMDLINE_LINUX_DEFAULT="? === frankban is now known as frankban|afk [17:30] alanmac: sorry, was on the phone. I dont think that would work. --no-nvram would keep whatever is set on the bios which i think it will try to pxe boot, and not fallback to boot from the disk [17:31] ok. reason i ask about modifying grub is we have a few nodes that are set for OpenStack (openstack ansible - have a few vital components on these servers). [19:45] oddball question for the day has anyone created a preseed file to do a core os and maas install? [19:52] Yes [20:14] alanmac: right, so I had to change my efibootmgr parameters manually to get them to work around the issue really [20:15] ok... hmmm, probably just easier to install the signed kernel then? [20:17] alanmac: the kernel is installed [20:17] alanmac: but you just need the right boot order to fallback to disk after attempting to pxe boot