[00:38] hi [00:59] If I setup DHCP through MAAS will it cause conflicts with my existing DHCP server [05:57] Hello,! [05:58] i use vmware to test maas ... maas requre power type to boot up the node [05:59] So any recommend for this ? i don't which power type should i choose and i don't know even power address or etc === mgz is now known as mgz_ === wojdev_ is now known as wojdev === wojdev_ is now known as wojdev === imthewhe` is now known as cgseller_is_away [13:27] Trying the MAAS trusty hwe-v kernel using the daily images and the kernel panics on boot. See bug: https://bugs.launchpad.net/maas/+bug/1465071 [14:28] dweaver, you rock thanks [14:28] smoser, ^^ [14:41] dweaver, fix is coming :) [14:42] dweaver, and actually, should be fixed i hope. [14:52] kiko smoser, stellar, thank you so much ;) === zerick_ is now known as zerick [15:22] dweaver, seems still busted even with 20150611, but it shoudlnt be. investigating. [15:27] plars, rvba thinks you have encountered a bug [15:27] kiko: ah, ok [15:27] plars, however, we know that some custom images work, i.e. centos [15:29] kiko: plars: tbh, I'm not that familiar with the custom images code so I can't be sure this is a bug (maybe there is just something I'm missing)… and we don't have documentation for this feature unfortunately. [15:30] rvba, does anyone other than blake_r know? [15:30] kiko: roaksoax probably. [15:31] rvba, and other than him? :) [15:32] Nobody else really. [16:05] dweaver, luser error made me think it was still busted. [16:05] i think i booted too early and got old initramfs. [16:05] if it is not fixed for you, and you have [16:05] 20150611 for ubuntu/*/hwe-v/trusty [16:05] then please let me know. [16:06] smoser, that version is in the daily images now? [16:06] yes. [16:06] refresh either from web ui, or cli (maas boot-resources import) [16:07] it is only there for ~ 2 hours though. [16:07] smoser, OK, I'll sync and have another go, if you don't hear anything else from me then you must be awesome and fixed it, thanks [16:09] smoser, OK, syncing now. [16:20] Bug #1465071 changed: Daily hwe-v kernel for trusty results in kernel panic when attempting to mount iscsi partition [16:47] Bug #1466151 opened: MAAS unit tests (at a minimum) are broken when bonding driver is loaded [16:55] plars: kiko: Actually you can't use 'ubuntu' as a prefix for your image name because the suffix isn't a valid series name. [16:55] plars: so just drop the 'ubuntu/' in front of your image. It should then show up under 'Custom'. [16:55] interesting, I'll give it a try in a bit [16:55] thanks rvba [16:56] plars: you can check that it's correctly synced by making sure /var/lib/maas/boot-resources/current/custom/amd64/generic/ contains the name of your image. [16:56] Welcome. [17:16] thanks plars [17:16] and thanks^2 rvba === kiko` is now known as kiko [17:47] Bug #1466162 opened: commissioning error [19:24] Good Afternoon === kevin is now known as Guest81452 [19:28] I am running an ubuntu server 14.04 instance with MAAS v1.5 as a virtual guest in Virtualbox. I am setting the regional controller, cluster controller, and dhcp server on the same server. eth0 is connected to the NAT network and has public internet connection. eth1 is reserved for the dhcp server. Scope for the NAT network is 10.0.2.0/24 so eth0 is 10.0.2.254 and eth1 is 10.0.2.3. [19:31] eth1 is also a host-only network adapter. DHCP was working well though and assigning numbers in the 10.0.2.0/24 scope and I still had internet connection. After I ran dpkg-reconfigure maas-regional-controller and dpkg-reconfigure maas-cluster-controller and setup accordingly I can no long reach the public internet [19:37] what is the best way to troubleshoot this? [20:03] CrummyAdmin, you can't reach the internet from the server, or from nodes provisioned from it? [20:03] also [20:03] don't use 1.5 [20:03] use the stable PPA above [20:21] I can't reach the internet from the server [20:21] I still needed to download images for maas but am unable to currently [20:22] Should I go ahead and upgrade maas even though I am currently having this issue or wait? [20:22] I can just start from scratch again but would like to be able to troubleshoot for the experience [20:31] CrummyAdmin, you should upgrade maas regardless [20:31] it's a bit weird that your uplink went down -- how are your interfaces configured? [20:39] plars, did that work? [20:40] kiko: not so far... it seems to be syncing? [20:40] I ran: [20:40] maas testmaas boot-resources create name=ci/wily architecture=amd64/generic content@=/tmp/wily-desktop-amd64.iso.tar.gz [20:40] and I see it under images, but it's still under generated, not custom [20:40] plars, hmm, and are either doing something? [20:40] and there's a circle next to it, when I hover, it says "waiting for clusters to sync" [20:41] okay [20:41] is there a way to force that? [20:41] plars, is df changing? :) [20:41] no, it should happen automatically [20:41] Bug #1466213 opened: commissioning output should show the output of lsblk and udevadm commands used to gather block device data [20:42] kiko: not quickly if it is [20:42] kiko: this was many hours ago that I ran the create [20:42] plars, and the cluster is healthy, connected, etc? [20:43] kiko: seems to be - it's just a small test system I set up locally [20:43] kiko: just trying to figure out what options maas could provide for testing things based on daily isos [20:45] plars, I am sure we can do things to make things better [20:58] eth0 static 10.0.2.254 gateway 10.0.2.3 dns-search google.com dns-nameservers 8.8.8.8 8.8.4.4 netmask 255.255.255.0 [20:58] eth1 static 10.0.2.3 dns-search google.com dns-nameservers 8.8.8.8 8.8.4.4 netmask 255.255.255.0 [21:01] Destination Gateway Genmask Flags Metric Ref Use Iface [21:02] 0.0.0.0 10.0.2.3 0.0.0.0 UG 0 0 0 eth0 [21:03] 10.0.2.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 [21:03] 10.0.2.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1 [21:04] that is my route -n output [21:05] arp -a = "? (10.0.2.3) at on eth0 [21:23] CrummyAdmin, you are really a crummy admin :-) you can't have two interfaces on the same network without special handling [21:23] CrummyAdmin, what is eth1 supposed to be doing? [21:24] CrummyAdmin, perhaps it's worth taking a step back and explaining your network setup [21:25] I was going to make it a private subnet for the dhcp server but couldn't get internet so someone suggested issuing ip's from it that are on the same subnet so I would still get internet. up until configuring maas it seemed to be working okay lol I am pretty new at this and just doing this to learn [21:26] sure [21:26] so the right way to do that is to use the machine to also route from eth1 to eth0 [21:26] if you shut down eth1 for now do you get your internet back? [21:27] what are eth0 and eth1 plugged into? [21:27] So with virtualbox I have a nat network with dhcp turned off and the Network CIDR as 10.0.2.0/24. this is the connection to the public internet that my maas server has on eth0 [21:27] ah [21:27] you are running maas inside virtualbox [21:27] I see [21:28] No I used ifdown eth1 and try connection again and could not ping google.com [21:28] that's odd [21:28] could a restart be necessary? [21:28] well [21:28] ping google.com is not a good test [21:28] can you ping 8.8.8.8? [21:29] I have 3 physical servers at home I plan to deploy to but learning to configure in virtualbox while on the go seemed to be my best option for now [21:29] right [21:29] so most simple setups do the following: [21:29] - use a server with two interfaces for maas [21:29] from 10.0.2.254 imcp_seq=1 Destination Host Unreachable [21:29] - maas is configured to manage DHCP/DNS on one of those interfaces [21:30] - maas is configured to ignore the other interfaces [21:30] okay yeah I told maas to use eth1 [21:30] - you configure routing (and possibly NAT, though not in your case) to allow machines on the managed interface to get outside [21:31] right [21:31] and I setup dhcp using isc-dhcp-server before hand then provided the config to maas within the web gui [21:31] oh [21:31] you shouldn't set up dhcp manually [21:31] maas does that for you [21:31] well, unless you really want to [21:31] but that is not a trivial undertaking [21:31] Oh! I was wondering about that [21:31] you should also really be on 1.7 [21:31] REALLY [21:32] I need to split tongiht [21:32] I promise once I get a connection back I will upgrade immediately =) [21:32] heh [21:32] k thanks for the suggestions I will google routing [21:32] so I'm not sure what is wrong -- eth1 down should have [21:32] scratch that [21:32] I know what is wrong [21:32] your default route is wrong [21:33] how are you configuring your networks [21:33] network manager? [21:33] or /etc/network/interfaces? [21:33] all through terminal no gui [21:33] yeah /etc/network/interfaces [21:33] ok [21:34] so if you look at your setup [21:34] you have a default route set up to the address of your eth1 [21:34] that can't work [21:34] your default route needs to be the first hop external to you [21:34] which I would assume is 10.0.2.1 [21:34] if you shut down eth1 and fix your default route you will have internet back [21:35] and later [21:35] later thank you again! [21:35] you will need a second subnet which MAAS will manage, such as 10.0.3.0/24 [21:35] okay that's what I originally wanted to do [21:35] right [21:35] that should work [21:36] you don't even need to do anything to route if that's the config [21:36] as packets coming in from eth1 that are intended for outside will be told to go to your default route, which is reachable by eth0 [21:36] so what would 10.0.2.1 be if dhcp is turned off on the nat network and my dhcp server is currently 10.0.2.3 [21:36] well [21:37] if you are sure eth1 is down [21:37] can you ping 10.0.2.3? [21:37] should I turn dhcp back on for the nat network. Yes eth1 is down [21:37] nope host unreachable at the moment [21:38] dhcp is turned off where? outside the virtualbox? [21:38] I think testing this inside a VM is much harder fwiw [21:38] virtualbox has a virtual network that you can enable dhcp on and virtual box will distribute addresses like a router from the virtual adapter [21:39] that is what eth0 is connected to [21:39] sure [21:39] what is the gateway address supposed to be on that network? [21:40] if it's 10.0.2.3, when you should be reaching it from eth0 [21:40] but perhaps you confused virtualbox when you added a second interface with that same address [21:40] virtual box uses 10.0.2.1 as the gateway I believe [21:41] if it is turned on to handle it [21:41] can you ping 10.0.2.1? [21:41] yes [21:41] okay [21:41] then make that your default route [21:41] your gateway in /e/n/i speak [21:41] your internet will come back [21:42] okay [21:42] and it did! [21:43] sweet thank you! Takeaways: let maas configure dhcp, keep both nic on seperate subnets, use routing to configure default route and enable public internet on both nics [21:44] and upgrade my maas install [21:46] correct [21:46] uhh [21:46] no [21:47] don't enable public internet on both nics [21:47] well [21:47] yes [21:47] I think you can't basically [21:50] Sorry for dropping kiko, thanks again for your assistance. Huge help [21:50] correct [21:50] uhh [21:50] no [21:50] don't enable public internet on both nics [21:50] well [21:50] yes [21:51] I think you can't basically [21:51] did you get that CrummyAdmin? [21:51] maas should route [21:51] not virtualbox [21:52] okay I make up a name of an adapter in the maas web gui correct then put in the details for the subnet I would like it to create on eth1 the private subnet and let it do it's thing