[02:18] echo === Guest44232 is now known as zz_Guest44232 [11:01] Bug #1498851 opened: [1.9.0 - alpha 1] Can't save power config [11:01] Bug #1498854 opened: [1.9.0 - alpha 1] cannot select power details on node details [12:40] Bug #1498899 opened: [1.9.0 - alpha1] Can't delete node [13:16] Bug #1498899 changed: [1.9.0 - alpha1] Can't delete node [13:22] Bug #1498899 opened: [1.9.0 - alpha1] Can't delete node [13:28] Bug #1498899 changed: [1.9.0 - alpha1] Can't delete node [13:31] Bug #1498899 opened: [1.9.0 - alpha1] Can't delete node [13:34] Bug #1498899 changed: [1.9.0 - alpha1] Can't delete node [13:37] Bug #1498899 opened: [1.9.0 - alpha1] Can't delete node [13:40] Bug #1498899 changed: [1.9.0 - alpha1] Can't delete node === jfarschman is now known as MilesDenver === jfarschman is now known as MilesDenver === Sagar is now known as Guest45894 [14:24] hi ... I have a MAAS cluster running 1.8.0 version, want to install OS on a physical server having a bunch of hard drives. I want control to install OS on a specific HDD [14:24] currently, MAAS installs OS randomly [14:25] any suggestions? [14:26] #join [14:41] I don't see anyone active on the list [17:56] hows it going everyone? [17:57] I have a concern about our production storage infrastructure that is deployed via maas/juju .... [17:59] I cannot seem to wrap my head around the process for replacing failed disks in a storage node deployed via MAAS...... is there anyone here who might have an idea on a best practice for disk failure/recovery without having to re-commission a node? [18:02] bdx: at the moment we don't support that without re-commissioning, but we are discussing such feature [18:10] bdx: although, you could always update the serial and such directly through the api [18:28] roaksoax: Is that what you do on large scale production deployments when a disk fails? [18:29] updating the serial directly only fixes 1/3 of the issue [18:30] good call on that....but it is no solution...just a heads up [18:31] bdx: as I said, It is just a work around not a solution [18:31] bdx: as we don't support that without re-commissioning [18:34] roaksoax: gotcha, thanks [18:56] Bug #1499044 opened: [1.9.0 alpha2] Ready node has OS [19:11] Bug #1499047 opened: [1.9.0 alpha2] Node stuck in deployment [19:14] Bug #1499047 changed: [1.9.0 alpha2] Node stuck in deployment [19:19] Hi, so I have a MaaS controller in a VM. I am trying to get it to work with the other 2 VMs I have setup and prepared for Nodes. But PXE boot shows "No boot filename recieved". It should be noted. I have a host-only network specifically for the 2 nodes and the controller. There is a secondary host-only network for internet though that is also connected to all three VMs. [19:19] Help? [19:22] (the hypervisor is vmware fyi) [19:23] Bug #1499047 opened: [1.9.0 alpha2] Node stuck in deployment [19:29] Bug #1499047 changed: [1.9.0 alpha2] Node stuck in deployment [19:29] Bug #1499050 opened: [1.9.0 alpha2] can't access networks tab [19:32] Bug #1499050 changed: [1.9.0 alpha2] can't access networks tab [19:32] Bug #1499047 opened: [1.9.0 alpha2] Node stuck in deployment [19:35] Bug #1499047 changed: [1.9.0 alpha2] Node stuck in deployment [19:35] Bug #1499050 opened: [1.9.0 alpha2] can't access networks tab [19:38] Bug #1499050 changed: [1.9.0 alpha2] can't access networks tab [19:38] Bug #1499047 opened: [1.9.0 alpha2] Node stuck in deployment [19:41] Bug #1499047 changed: [1.9.0 alpha2] Node stuck in deployment [19:41] Bug #1499050 opened: [1.9.0 alpha2] can't access networks tab [19:57] JZTech101, did you run thru https://maas.ubuntu.com/docs/install.html [19:59] yes [19:59] I did [19:59] I installed it using the Multi-Server Option on the CD [20:00] The eth1 (Main host-only network is setup http://vgy.me/iHK8gK). I removed eth0 (the host-only internet network) from the cluster controller, but that has made no difference. [20:00] pserv.log shows : [20:00] 2015-09-23 15:31:03-0400 [Uninitialized] Event-loop ondria:pid=2000 (127.0.0.1:48076): Connection was refused by other side: 111: Connection refused. [20:00] 2015-09-23 15:31:03-0400 [Uninitialized] Event-loop ondria:pid=2000 (1.1.1.39:48076): Connection was refused by other side: 111: Connection refused. [20:00] and MaaS.log shows : [20:00] Sep 23 15:28:08 ondria maas.lease_upload_service: [INFO] Uploading 0 DHCP leases to region controller. [20:00] Sep 23 15:28:08 ondria maas.bootsources: [INFO] Updated boot sources cache. [20:00] Sep 23 15:28:08 ondria maas.import-images: [INFO] Started importing boot images. [20:00] Sep 23 15:28:08 ondria maas.bootresources: [INFO] Importing images from source: http://maas.ubuntu.com/images/ephemeral-v2/releases/ [20:01] Sep 23 15:28:08 ondria maas.boot_image_download_service: [ERROR] Failed to download images: Unable to open http://1.1.1.39/images-stream/streams/v1/index.json. mirrors=[] [20:01] Sep 23 15:28:09 ondria maas.bootresources: [INFO] Finished importing of boot images from 1 source(s). [20:01] Sep 23 15:28:09 ondria maas.import-images: [INFO] Started importing boot images. [20:01] Sep 23 15:28:09 ondria maas.import-images: [INFO] Finished importing boot images, the region does not have any new images. [20:01] Sep 23 15:30:47 ondria maas.calls: [INFO] Starting task 'write_full_dns_config' with args: () {'zones': [, ], 'callback': provisioningserver.tasks.rndc_command([u'reload'], False), 'upstream_dns': [u'1.1.1.1'], 'trusted_networks': u'1.1.1.0/24; 1.1.2.0/24;'} [20:01] Sep 23 15:30:47 ondria maas.calls: [INFO] Starting task 'rndc_command' with args: ([u'reload'], False) {} [20:01] Sep 23 15:30:47 ondria maas.calls: [INFO] Finished task 'rndc_command' with args: ([u'reload'], False) {} [20:01] Sep 23 15:30:47 ondria maas.calls: [INFO] Finished task 'write_full_dns_config' with args: () {'zones': [, ], 'callback': provisioningserver.tasks.rndc_command([u'reload'], False), 'upstream_dns': [u'1.1.1.1'], 'trusted_networks': u'1.1.1.0/24; 1.1.2.0/24;'} [20:01] Sep 23 15:30:48 ondria maas: [WARNING] Failed to create Network when adding/editing cluster interface maas-eth1 with error [{'ip': [u'Network with this Ip already exists.'], 'name': [u'Network with this Name already exists.']}]. This is OK if it already exists. [20:14] Bug #1499062 opened: Internal Server Error if 'Sync URL' Setting has a preseeding whitespace [20:24] JZTech101, click on clusters tab in maas gui, mine is called cluster master [20:25] mwenning, yes. Mine is called cluster master as well [20:25] I've got 2 interfaces, under the manage column, one says dhcp and dns [20:25] the other one is unmanaged [20:26] I only have one managed dhcp and dns. The other one that is managed by my router, I removed [20:26] Should I uh add it back? [20:26] removing it nor adding it has made a difference [20:26] not yet ;-) [20:27] ok [20:27] interface is eth0? [20:27] no eth1 [20:27] eth0 is the one managed by router [20:27] ok. what is eth1 set up as in virt-manager [20:28] uh [20:28] what [20:28] Not using virt-manager? [20:28] oh. I use vmware. Its a host-only network with no DHCP [20:29] oh ok. [20:29] oh and I should mention, the controller is a VM on its own as well its Nodes. But that shouldn't make a difference afaik [20:30] (Controller + Nodes are side-by-side VMs) [20:30] JZTech101, that's what I've got here using kvm/virt-manager [20:30] ok [20:32] Maas-server is set up as a static address, 172.16.0.1/22 [20:32] Nodes should be set up to pxe boot from the same network of course [20:33] maas-server should access the outside world on the other network [20:33] uh. Static IP Address? I just uh thought that MaaS controller would handle configuring this new network adapter [20:33] :| [20:34] what's the network addr for eth1 on Interfaces [20:35] Interfaces section of Cluster Controller [20:35] 1.1.2.0/24 [20:35] http://vgy.me/iHK8gK That's a screenshot of the interface config [20:39] And you are accessing the maas server gui from 1.1.2.0 ? [20:40] uh [20:40] no actually [20:40] I'm accessing it through eth0 [20:42] mwenning, I have not gotten an IP from MaaS so I can't access that particular subnet. Actually the VM itself has no IP for that subnet [20:42] o.O [20:42] how to make MaaS properly assign IPs? [20:43] ok, what I did was edit /etc/network/interfaces and add a static address for the internal maas network [20:43] in youre case eth0 [20:43] oops eth1 [20:44] mwenning, can you give me an example config so I don't mess it up? [20:44] sure stby [20:45] Bug #1499050 changed: [1.9.0 alpha2] can't access networks tab [20:45] Also, I changed the static IP field to 1.1.2.1 [20:45] And left the router field the same (1.1.2.1) [20:47] auto eth0 [20:47] iface eth0 inet static [20:47] address 172.16.0.1 [20:47] broadcast 172.16.3.255 [20:47] netmask 255.255.252.0 [20:47] JZTech101, yeah that's the way mine's set up [20:47] ok [20:49] Try that. did you find the maas doc on setting the nodes to poweronoff using virsh? [20:49] o.o [20:49] why is the netmask 252.0? [20:49] I always set it 255.0 [20:51] maas 1.8 uses 2 sets of dhcp ranges, one for dynamic and another for static [20:51] mine setup assigns 255 to each range [20:52] uh [20:52] wat? o.O [20:52] * JZTech101 is confused [20:55] * mwenning tries to take a screenshot of his maas server [20:55] mwenning, ok. So eth1 now has an IP on the VM. But the DHCP Server is still not functioning as my host has yet to have an IP from MaaS [20:57] Bug #1499050 opened: [1.9.0 alpha2] can't access networks tab [20:57] JZTech101, try that http://vgy.me/DrEMup [21:00] Bug #1499050 changed: [1.9.0 alpha2] can't access networks tab [21:01] uh mwenning I see your main IP is 0.1. Well.. the thing is, the router hands out IPs in the 1.1.1.x range. So would using 252.0 conflict? [21:04] JZTech101, in your case probably. It's not important, this setup is for controlling big mass clusters, where we have 100's of nodes [21:04] If you bring up another node on the internal maas network, can you ping the maas server? === zz_Guest44232 is now known as CyberJacob [21:07] mwenning, well its not a node. Its just another device. And no. it is not getting an IP from the MaaS server :( [21:07] so it cannot ping the MaaS server [21:08] Imma brb for dinner [21:08] JZTech101, ok [21:12] JZTech101, I'm gonna have to bail soon too. Try this doc: http://certification-static.canonical.com/docs/MAAS_Advanced_NUC_Installation_And_Configuration.pdf [21:13] That's the way I set up mine. It assumes real hw but it worked fine from the MAAS_server VM [21:16] When you get that working you will have to set the nodes up to boot from virsh, see https://maas.ubuntu.com/docs/nodes.html [21:54] mwenning, still around? [21:55] JZTech101, yup, can you reach that document? [21:56] mwenning, I can indeed. And the setup script, I'm not entirely sure how to get it seeing as I did a MAAS install direct from the Server CD [21:56] rather than on an existing install [21:57] but that might be what is lacking with the DHCP Server not working? [21:58] :O [21:58] mwenning, hmm. Ok, so its just the host being stupid. I can't blame it. But Netboot seems to be working now [21:58] :D [21:58] hm [21:58] mass-enlisting node eh? [21:58] :o [21:59] it's working now? [21:59] rephrasing, it's pxebooting now? [22:00] it is pxebooting now :D [22:00] |\o/ [22:01] Nodes go thru 3 phases, enlisting, commissioning, and deploying [22:01] enlist they pxeboot and make themselves known to MAAS [22:01] commission they will pxe boot again and collect system info [22:02] deploy they will pxeboot, load an image on the disk, and reboot to the disk. [22:03] hm [22:03] how do I tell it to deploy the image? [22:03] after enlist you will have to set up the virsh thing above, since maas assumes that it can control power to the node for commission and deploy [22:03] I only have one Ubuntu LTS [22:03] lol [22:04] on the node page on the maas gui, the enlisting node should show up as "New" [22:04] uh [22:04] then you click in there and edit the power settings [22:04] how do I setup virsh if the controller itself is a VM alongside the other VMs? [22:05] and how do I do that in VMWare [22:10] hmm. I googled 'maas virsh vmware' and 'maas vmware power type' and got some hits :-O [22:11] Also you can be the power controller - hit commission and then manually turn on the vm [22:11] Yeah I know [22:11] thats actually what I'm doing right now [22:11] lol [22:11] ok well that will work its just a PITA [22:12] * JZTech101 shrugs [22:13] the problem is, afaik vmware workstation does not support Wake-On-Lan [22:13] and none of the others will work because it is a VM alongside VMs [22:13] The host is not the controller [22:13] this is fun [22:13] it is deploying [22:13] :D [22:14] Thanks for your help mwenning :D [22:14] JZTech101, np. thanks for your patience :-) [22:26] hm [22:27] mwenning, is there anything special I need to do to get the ssh-keys working? Basically I added the wrong key After deploying the node. So I have redeploy the node or something? [22:27] (I've already changed it to the correct SSH Key [22:29] This is for power control? [22:29] uh no. [22:30] I'm just trying ot access this new node [22:30] lol [22:30] * mwenning is still thinking virsh [22:31] you will have to re-deploy so the new key gets set up on the deployed node after the install. [22:31] I would expect so. I think the key is passed during deployment so you can ssh into the node without a password [22:31] what catbus1 said ;-) [22:31] :( [22:31] ok [22:31] was hoping it would automatically update [22:32] that's the cool thing about vm's it doesn't take 20 minutes like real hw [22:33] Also [22:33] Why am I getting 14.04.1? [22:34] 14.04.3 has long been out [22:34] :< [22:34] enlist or commission? [22:34] commission? [22:35] the pxeboot images are still 14.04.1, once you get to the real image on the hd (in deploy), you should see 14.04.3 [22:35] ah [22:40] o.O [22:41] hey mwenning what am I doing wrong? I put the public SSH Key in there. (the .pub file). And I still can't connect. I'm using owner as the username. Should I use something else? [22:42] hm. Apparently I should [22:42] ssh ubuntu@mynode [22:42] the user is ubuntu [22:44] well its deployed and it says "Welcome to Ubuntu 14.04.1 LTS (GNU/Linux 3.13.0-35-generic x86_64) [22:44] " === CyberJacob is now known as zz_CyberJacob [22:46] I had a fresh maas install and configuration on Monday, and the imported image is 14.04.3 on my maas. [22:46] JZTech101: what your maas version? [22:46] not that I think it's related, but would be good to know. [22:47] uh [22:47] where do I get the version? [22:47] apt-cache policy maas [22:47] if you are using maas 1.8, you should see maas version on the bottom left of the maas webpage [22:48] Installed: 1.7.6+bzr3376-0ubuntu2~14.04.1 [22:48] o.O [22:49] I installed from a Server CD I downloaded today [22:49] ;_; [22:50] 1.7.6 is good [22:51] wait, 14.04.1 but with newer kernel 3.13, your node is using 14.04.1 user space and 3.13 kernel, the same kernel version as in 14.04.3. [22:56] JZTech101: I used MAAS to deploy a node and it comes up the same as you are seeing. [22:56] JZTech101: the other node 14.04.3 I have, I used juju to deploy, so I guess after the install, it might do apt-get dist-upgrade. [23:01] sorry, I was wrong about the kernel version in 14.04.3, it is 3.19. It is 3.13 in 14.04.1. [23:01] JZTech101: so I did apt-get dist-upgrade on the MAAS deployed node, and now it comes up 14.04.3 with 3.13.0-35 kernel. [23:17] ah ok