| koaps | roaksoax: ya, i tried that command and nothing happened, you would think that would be right | 01:07 |
|---|---|---|
| mup | Bug #1602093 changed: [2.0r1] maas webui error updating subnet mtu when dhcp is configured <MAAS:Expired> <https://launchpad.net/bugs/1602093> | 04:36 |
| mup | Bug #1602093 opened: [2.0r1] maas webui error updating subnet mtu when dhcp is configured <MAAS:Expired> <https://launchpad.net/bugs/1602093> | 04:42 |
| mup | Bug #1602093 changed: [2.0r1] maas webui error updating subnet mtu when dhcp is configured <MAAS:Expired> <https://launchpad.net/bugs/1602093> | 04:45 |
| === frankban|afk is now known as frankban | ||
| neith | kiko: I understand I must not provide 2 default gw for maas server as well for enlisted nodes. But should I add a default gw to the public subnet? not the subnet on which nodes pxe boot | 14:11 |
| rock_ | Hi. I want to use KVM machines for maas setup using http://maas.io/docs/installconfig-add-nodes. So I created one kvm machine for maas-admin node. From that machine I ran $sudo virsh list --all. It was returning empty. Please provide me solution for this. | 14:25 |
| roaksoax | neith: you can change the interface which will configure the default gateway | 14:42 |
| roaksoax | neith: look into "maas <user> interface set-default-gateway" | 14:43 |
| roaksoax | rock_: i think you have the wrong conception of what KVM/virsh do | 14:43 |
| roaksoax | rock_: if you do: sudo virsh list --all -> that will list of KVM's VM's running on the system where you are running that command | 14:44 |
| roaksoax | rock_: if that's a VM, it is most likely you don't have nested VM's on that machine | 14:44 |
| roaksoax | rock_: so you need to log into a external KVM/libvirt, to be able to list running VM's | 14:44 |
| rock_ | roaksoax: Thank you. I have a question with KVM machines maas setup. | 14:46 |
| roaksoax | rock_: so, are you running MAAS as a VM right ? | 14:48 |
| roaksoax | rock_: so say, MAAS is in subnet 192.168.122.0/24, and your KVM host is 192.168.122.1/24 right ? | 14:49 |
| rock_ | roaksoax: Yes. | 14:49 |
| roaksoax | rock_: so you'd need to do something like: virsh -c qemu+ssh://192.168.122.1/system list --all | 14:50 |
| roaksoax | rock_: so you'd need to do something like: virsh -c qemu+ssh://<user>@192.168.122.1/system list --all | 14:50 |
| rock_ | roaksoax: I created one kvm machine for MAAS installation It has taken 192.168.122.98 [DHCP] | 14:51 |
| roaksoax | rock_: right, so you need to: | 14:52 |
| roaksoax | 1. configure you MAAS VM with static IP address | 14:52 |
| roaksoax | 2. disable libvirt DHCP, or it wont let MAAS provide DHC{P | 14:52 |
| rock_ | roaksoax: So can I configure static IP for MAAS VM using 192.168.122.0/24 range IP. | 14:54 |
| rock_ | roaksoax: For remaining 4 maas-slave nodes, No need to install OS. Enabling pxe boot is sufficient? Then MAAS will do auto discovery ? | 14:56 |
| rock_ | roaksoax: (or) for adding nodes manually, I need power ID and adress of slave nodes right. Power ID means name of slave-node. So How can I get power address? | 14:58 |
| roaksoax | rock_: so do 1 and 2 above, then name the VM's maas will control somthing different from MAAS< for example: | 14:59 |
| roaksoax | maas-server, node01,node02, etc (in livirt) | 14:59 |
| roaksoax | rock_: 3. name your MAAS VM vs your node's differently, so they have a different prefix | 15:00 |
| roaksoax | rock_: 4. Go to MAAS, click on 'Add Chassis', and input for power address: qemu+ssh://<user>@192.168.122.1/system power password: your psasword, prefix filter - a prefix to import all of those machines that start with that prefix (that does NOT match the MAAS server) | 15:01 |
| roaksoax | rock_: 5. configur DHCP in MAAS | 15:01 |
| roaksoax | rock_: 6. Download images in MAAS | 15:01 |
| rock_ | roaksoax: OK. Thank you. So I have give the same power address :qemu+ssh://<user>@192.168.122.1/system power for all slave nodes right. In MAAS node am I need to create a bridge (or) router kind of thing. | 15:04 |
| roaksoax | rock_: no, since they are all vm's, theu will all access the internet via the KVM host machine | 15:05 |
| roaksoax | rock_: via 192.168.122.1 | 15:05 |
| roaksoax | rock_: if you were running maas in the KVM host machine (uinstead of inside a VM), they you may need to create a bridge | 15:05 |
| rock_ | roaksoax: Ok. Thank you. | 15:07 |
| rock_ | SO in maas VM i have to take 192.168.122.1 as gateway, 192.168.122.0/24 as network , and 255.255.255.0 as netmask and address as 192.168.122.x right. | 15:10 |
| rock_ | roaksoax: And have to make the given IP as static. | 15:12 |
| rock_ | roaksoax: Thank you for your patience in providing the detailed answers.. | 15:12 |
| neith | roaksoax: How can I remove the default gateway?? | 15:53 |
| PCdude | kiko: hé again, I had alot of work in the last couple of days, but gonna try and troubleshoot again :) . My question is if u know how to trim down the time that the installer fails. Right now its 3000 seconds which is enormous for troubleshooting | 16:48 |
| roaksoax | PCdude: define trim down the installer fails ? | 16:53 |
| roaksoax | neith: don't put a gateway int he subnet, but why would you want to do that ? machines wouldn't be able to install packages and such without it | 16:54 |
| PCdude | roaksoax: good point, I am trying to install openstack with "openstack-installer". I am using the autopilot option in there. It fails and I have had some good help a couple of days ago from different people on IRC. I am wanna try to troubleshoot some more, but the installer fail takes 3000 seconds. There should be a file somewhere on the disk that | 16:55 |
| PCdude | defines that number. I wanna change that so I can troubleshoot faster and see what changes in the error log. | 16:55 |
| PCdude | roaksoax: does that make sense? | 16:55 |
| === frankban is now known as frankban|afk | ||
| roaksoax | PCdude: it does, the 3000 seconds though seems to be a timeout someone else holds that it is not MAAS | 17:09 |
| roaksoax | PCdude: MAAS has a 40 minute deployment timeout, where if the machine has not been deployed in 40 minutes, it gets marked as failed deployment | 17:09 |
| roaksoax | PCdude: and that is non-configurable in MAAs | 17:09 |
| roaksoax | PCdude: that said, when deploying opestanck you may want to identify two things: | 17:10 |
| roaksoax | 1. MAAS 2. Juju. | 17:10 |
| roaksoax | 1. does MAAS mark the amchine deployed ? IF it does, then JUju (or something else) cases the machjine to have failed in Juju) | 17:10 |
| PCdude | roaksoax: it is know where the issue is. JUJU is staring an container on a node, but either is communicating wrong or LXC is giving no IP to the container. Which makes that the container cannot be reached. The problem is that I want to try some things out, but yeah that is gonna take me all day. The reason I asked kiko was simply coz I spoke about | 17:12 |
| PCdude | this a reasonable amount. So maybe he knew, but I am aware of the fact that this is not MAAS related. | 17:12 |
| sgoller | Hey all, I'm trying to deploy maas-2.0 and I'm running into problems. I have a region controller and a rack controller on separate boxes. The rack controller is registered, but when I try to provide dhcp for its subnet, the region controller has an empty list for the rack controller. | 17:49 |
| sgoller | Any ideas? | 17:49 |
| kiko | sgoller, seems like a bug. 2.0 final? | 17:55 |
| kiko | sgoller, basically you registered the cluster controller with the region and it doesn't show up? | 17:56 |
| kiko | sgoller, or perhaps are you missing the registration step? | 17:56 |
| sgoller | the steps I took were: | 17:58 |
| sgoller | installed region controller | 17:58 |
| sgoller | made sure it could be logged into | 17:58 |
| sgoller | installed the rack controller | 17:58 |
| sgoller | registered it with the region controller | 17:58 |
| kiko | and the registration succeeded? | 17:58 |
| sgoller | yes | 17:58 |
| sgoller | the subnet for the rack controller is present | 17:59 |
| kiko | roaksoax, ^^ what do you think? | 17:59 |
| sgoller | I created a separate fabric for the subnet, moved the rack controller's subnet over | 17:59 |
| sgoller | then went to the untagged vlan for that fabric and selected "provide dhcp" | 18:00 |
| sgoller | the subnet shows up fine, but the list of rack controllers is empty. | 18:00 |
| sgoller | MAAS Version 2.0.0+bzr5189-0ubuntu1 (16.04.1) | 18:01 |
| sgoller | is there a newer version? I'm pulling from ppa:maas/stable | 18:01 |
| sgoller | i'm going to blow everything away and start from scratch | 18:05 |
| kiko | sgoller, that's the right version | 18:14 |
| kiko | sgoller, let me get somebody to help before you blow it away | 18:14 |
| sgoller | kiko, ok. | 18:15 |
| kiko | sgoller, does anything show up in your /var/log/maas or syslog? | 18:21 |
| kiko | sgoller, anything suggesting a problem? | 18:21 |
| sgoller | not seeing anything obvious. I'm going to go through provide dhcp again to see if something shows up | 18:22 |
| sgoller | Sep 12 11:21:23 maas-region sh[3771]: 2016-09-12 11:21:23 [HTTPChannel,20,127.0.0.1] Opening connection with IPv4Address(TCP, '127.0.0.1', 54928) Sep 12 11:21:23 maas-region sh[3771]: 2016-09-12 11:21:23 [-] Closing connection: <STATUSES=PROTOCOL_ERROR> ('Failed to authenticate user.') | 18:22 |
| sgoller | this shows up periodically | 18:22 |
| sgoller | on the region controller | 18:22 |
| sgoller | here's more | 18:23 |
| sgoller | Sep 12 11:22:37 maas-region sh[3771]: 2016-09-12 11:22:37 [-] 127.0.0.1 - - [12/Sep/2016:18:22:36 +0000] "GET /MAAS/rpc/ HTTP/1.1" 200 220 "-" "provisioningserver.rpc.clusterservice.ClusterClientService" Sep 12 11:22:41 maas-region sh[3771]: 2016-09-12 11:22:41 [HTTPChannel,23,127.0.0.1] Opening connection with IPv4Address(TCP, '127.0.0.1', 55068) Sep 12 11:22:41 maas-region sh[3771]: 2016-09-12 11:22:41 [-] Closing connection: <STATUSES | 18:23 |
| sgoller | WAT | 18:24 |
| sgoller | ok, now it's shown up | 18:24 |
| kiko | mpontillo, ^^ | 18:25 |
| kiko | sgoller, did you change anything? | 18:25 |
| sgoller | I had shutdown both the rack controller and the region controller | 18:25 |
| sgoller | because i was going to reinstall them | 18:25 |
| sgoller | but then on your suggestion I brought them back up | 18:25 |
| sgoller | ugh. | 18:25 |
| kiko | sgoller, I'm wondering what you changed beyond restarting things that made it function | 18:27 |
| koaps | sgoller: think I've seen something like that before when I played with 2.0 beta a while back | 18:27 |
| koaps | I had to restart the dhcp service | 18:27 |
| sgoller | koaps: hm. This was a fresh install, and no dhcp had been configured yet, so it wasn't running at all on the rack controller. It's running now, however, since I was able to provision that subnet. | 18:28 |
| koaps | I've been meaning to setup my dev env with 2.0, if I get a chance I'll do that an maybe file a bug if I get same behavior, or look for yours :) | 18:30 |
| kiko | sgoller, what did you do exactly to register the controller? | 18:30 |
| kiko | sgoller, I'm wondering if you changed anything else after that and before restarting | 18:31 |
| sgoller | i clicked on add rack controller, which outputs a series of commands | 18:34 |
| sgoller | like i said, i moved the subnet connected to the rack controller over to fabric-1. | 18:34 |
| sgoller | maybe something to do with that | 18:34 |
| sgoller | and i had moved it back but it still had a problem | 18:35 |
| sgoller | so i suspect there may be some issue there. | 18:35 |
| sgoller | the rack controller is having problems syncing images with the region controller, so I think i'm going to reinstall again anyway. | 18:37 |
| sgoller | kiko, thanks for your help. | 18:37 |
| kiko | sgoller, wait, don't reinstall :) | 19:05 |
| kiko | sgoller, what issues are happening? let's look at the logfiles? | 19:05 |
| kiko | sgoller, I think you hit an underlying bug I'd love to figure out | 19:06 |
| sgoller | sorry, I already reinstalled. :( | 19:17 |
| kiko | sgoller, okay, let's keep an eye out if the problem reoccurs | 19:19 |
| sgoller | kiko, for sure. I'm going to follow the same steps. | 19:20 |
| sgoller | kiko, ok, I'm in that exact same scenario now. The order of operations was: Install region controller. create account, login to make sure it was up. install rack controller software, use add rack controller in the ui to show the commands to run to register it. register rack controller via cli... | 20:13 |
| sgoller | kiko, then move the rack controller's subnet to a different fabric | 20:13 |
| sgoller | then go to the untagged vlan on the new fabric, and select "Provide DHCP". The following screen does not have a selectable rack controller. | 20:14 |
| sgoller | kiko, so this is reproducible | 20:14 |
| kiko | sgoller, same log entry? and if you restart rackd, does it work? | 20:14 |
| sgoller | so this is interesting. | 20:15 |
| sgoller | going to the node page for the rack controller | 20:15 |
| sgoller | it lists it as being the rack controller for things only connected to the region controller | 20:16 |
| sgoller | kiko, this seems to imply that maas really doesn't like a standalone region controller | 20:16 |
| sgoller | kiko, the region controller has associated the subnets that are attached to itself to the rack controller, which is on a completely separate network. | 20:17 |
| kiko | sgoller, I think it's actually that registration is failing | 20:17 |
| kiko | but I'm not sure why | 20:17 |
| kiko | which is why I'm keen on the logs | 20:17 |
| sgoller | what logs would you like? /var/log/syslog from the rack controller? | 20:18 |
| kiko | sgoller, /var/log/maas actually | 20:18 |
| sgoller | hm. you can't upload a file to pastebin? weak. | 20:21 |
| kiko | sgoller, you can pipe to pastebinit! | 20:22 |
| sgoller | kiko, got it, thanks! | 20:22 |
| roaksoax | sgoller: sudo pastebinit /var/log/maas/rackd.log , and that for maas.log and regiond.log | 20:38 |
| PCdude | kiko: https://www.reddit.com/r/homelab/comments/4p3k9j/trouble_getting_lxc_networking_up_containers_not/ | 20:40 |
| PCdude | kiko: no time tonight, but maybe the solution to my problem | 20:41 |
| mup | Bug #1565467 opened: twistd3 crashed with PermissionError in touch(): [Errno 13] Permission denied: '/etc/maas/rackd.conf' <amd64> <apport-crash> <xenial> <MAAS:Confirmed> <twisted (Ubuntu):Expired> <https://launchpad.net/bugs/1565467> | 20:42 |
| mup | Bug #1565467 changed: twistd3 crashed with PermissionError in touch(): [Errno 13] Permission denied: '/etc/maas/rackd.conf' <amd64> <apport-crash> <xenial> <MAAS:Confirmed> <twisted (Ubuntu):Expired> <https://launchpad.net/bugs/1565467> | 20:48 |
| mup | Bug #1565467 opened: twistd3 crashed with PermissionError in touch(): [Errno 13] Permission denied: '/etc/maas/rackd.conf' <amd64> <apport-crash> <xenial> <MAAS:Confirmed> <twisted (Ubuntu):Expired> <https://launchpad.net/bugs/1565467> | 20:51 |
| kv | just wonder if maas 2 works with juju 2 yet ? | 21:41 |
| kiko | kv, yes, though juju2 is not final yet -- still a beta | 21:43 |
| kiko | kv, I hear the next release is an RC now! | 21:43 |
| kv | so the combo maas 2, juju 2, and conjure-up should work? | 21:43 |
| kiko | yes | 21:43 |
| kiko | and stokachu and mmcc are here to ensure conjure-up does ;) | 21:44 |
| kv | cool... i am going to give it another shot | 21:44 |
| jwitko | Hey guys, I just installed the latest stable 2.0 release and although I have imported ubuntu 16.04 and 14.04 images I still can not add any bare metal machines (The save button is disabled). One weird thing I've noticed is that under my settings I have no default comissing OS or deployment OS (The lists are empty except for an item telling me there are none to choose from) | 22:10 |
| jwitko | Anyone ever seen this before? Logs don't seem too crazy, this is a fresh install of both OS and maas, installation performed via apt | 22:10 |
| kiko | jwitko, hmm, that's slightly odd | 22:22 |
| kiko | jwitko, have you added SSH keys? | 22:23 |
| kiko | jwitko, however it does look like you've run into.. a bug? | 22:23 |
| kiko | those lists AIUI should not be empty | 22:23 |
| kiko | have the images finished syncing completely? | 22:23 |
| jwitko | how can I tell ? | 22:23 |
| jwitko | But they should have, its been a long time since I selected them to be imported | 22:24 |
| kiko | jwitko, there's UI feedback that it completed, so I'm guessing it did | 22:24 |
| jwitko | kiko, http://paste.ubuntu.com/23171284/ | 22:26 |
| jwitko | also, un-selecting the ubuntu releases that I've imported and hitting "apply changes" does not remove the images | 22:28 |
| kiko | jwitko, what shows up in /var/lib/maas? | 22:28 |
| jwitko | drwxr-xr-x 2 maas maas 4096 Sep 12 14:02 gnupg | 22:28 |
| jwitko | -rw------- 1 maas maas 6 Sep 12 15:21 maas_id | 22:28 |
| jwitko | -rw-r--r-- 1 maas maas 1689 Sep 12 15:21 maas-proxy.conf | 22:28 |
| jwitko | -rw-r----- 1 maas maas 32 Sep 12 14:02 secret | 22:28 |
| kiko | okay | 22:28 |
| kiko | jwitko, and anything suspicious /var/log/maas? | 22:28 |
| jwitko | maas.log:Sep 12 15:18:23 maas1 maas.bootresources: [WARNING] Unable to import boot images, no image descriptions avaliable. | 22:30 |
| roaksoax | jwitko: do you have a console log for the machine or /var/log/maas/rsyslog/maas-enlisting-node/* or /var/log/maas/rsyslog/<machine-name>/* | 22:31 |
| roaksoax | jwitko: i dont think you are running into issue | 22:31 |
| roaksoax | i think you are running into a bug with the kernel | 22:31 |
| jwitko | http://paste.ubuntu.com/23171298/ | 22:31 |
| jwitko | I'm running into a bug with the kernel... on the stable 2.0 release.. on a stock 16.04 fully updated install ? | 22:32 |
| kiko | regiond.log:2016-09-12 14:22:25 [-] Error on request (2428) controller.check_images: 4y3h7p | 22:33 |
| kiko | regiond.log: raise HandlerDoesNotExistError(pk) | 22:33 |
| kiko | regiond.log: maasserver.websockets.base.HandlerDoesNotExistError: 4y3h7p | 22:33 |
| jwitko | kiko, all errors, crits, warns, fails from the logs: http://paste.ubuntu.com/23171298/ | 22:33 |
| jwitko | oh you sa | 22:33 |
| jwitko | w | 22:33 |
| kiko | roaksoax? | 22:33 |
| roaksoax | jwitko: ah! yes, so no your bug is not what I was thinking about then | 22:34 |
| jwitko | have to let the dog out, brb in 5m | 22:35 |
| roaksoax | jwitko: can you pastebin all of rackd.log and regiond.log ? | 22:36 |
| jwitko | roaksoax, sure | 22:43 |
| jwitko | roaksoax, there is no rackd.log ? | 22:43 |
| jwitko | maas.log and regiond.log | 22:44 |
| mup | Bug #1622770 opened: [2.1, vanilla] Can't select CentOS images <ui> <ux> <vanilla> <MAAS:New> <https://launchpad.net/bugs/1622770> | 22:45 |
| jwitko | roaksoax, here is maas.log | 22:46 |
| jwitko | http://paste.ubuntu.com/23171341/ | 22:46 |
| jwitko | and here is regiond.log http://paste.ubuntu.com/23171343/ | 22:47 |
| roaksoax | jwitko: rackd.log ? | 22:55 |
| jwitko | roaksoax, I said above there is no rackd.log in /var/log/maas | 22:57 |
| roaksoax | jwitko: ok, so it seems you dont have a rack controller | 22:59 |
| roaksoax | jwitko: on this machine | 22:59 |
| roaksoax | jwitko: dpkg -l | grep maas | 22:59 |
| jwitko | lol are you serious | 23:00 |
| jwitko | this asshole | 23:00 |
| jwitko | (the guy who installed it) | 23:00 |
| jwitko | so a rack controller is the updated name for a 'cluster controller' ? | 23:00 |
| roaksoax | jwitko: yes, the internals are completely different but yes | 23:01 |
| jwitko | ty | 23:01 |
| newtomaas | Hi all - very new to maas and having trouble with pxe boots appearing not to do what they are meant to.. i see errors where it is trying to hit the 169.x.x.x subnet and many time outs and eventually it will land at an ubuntu prompt. running maas on physical hardware ubuntu 16 lts. also pxe boots are very unreliable. Machines trying to enlist are VMs. | 23:15 |
| kv | Unable to connect to: ws://xx.x.x.x/MAAS/ws | 23:25 |
| kv | any idea why I am getting that error kiko? | 23:25 |
| kv | a brand new setup...maas 2.0.0+bzr5189-0ubuntu1~16.04.1 | 23:27 |
| kv | i see what is going on..it's listening on 5240 by default | 23:30 |
| mup | Bug #1622786 opened: [2.0] Error "could not terminate the child" on node UI page above power parameters <oil> <MAAS:New> <https://launchpad.net/bugs/1622786> | 23:51 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!