saltlake2 | Champs, I am really confused on how to add my VM nodes to the maas controller. Does anyone have a step by step guide on how to create VM using virsh and then provision it in maas controller ? | 02:40 |
---|---|---|
saltlake2 | MarkBaker ? | 02:41 |
saltlake | anyone up for helping me figure out why a VM does not leave the commisioning state in my maas | 05:53 |
p_l | hi, I found some information that MaaS supports installing CentOS in version 1.7.0, is there any documentation on how to set it up (or where can I find it?) | 09:30 |
kiko | p_l, blake_r would be the best person to help, but I do believe there is documentation | 10:40 |
=== lazyPower is now known as lp|Metrics | ||
=== liam_ is now known as Guest51703 | ||
=== roadmr is now known as roadmr_afk | ||
=== roadmr_afk is now known as roadmr | ||
=== roadmr is now known as roadmr_afk | ||
=== roadmr_afk is now known as roadmr | ||
=== lp|Metrics is now known as lazyPower | ||
saltlake | experts, am trying to run this command sudo maas-import-pxe-files. It appears a new node tftps to the maas server and looks for files such as pxelinux.0 .. | 19:39 |
saltlake | Howver that command needs a sources-file. Can somehelp help me by giving me their sources-file contents | 19:39 |
blake_r | saltlake: what version of MAAS? | 20:10 |
saltlake | blake_r: I have " Installed: 1.7.0~beta8+bzr3272-0ubuntu1.2" | 20:14 |
saltlake | blake_r: " | 20:21 |
saltlake | any ideas ? | 20:21 |
blake_r | saltlake: yes so your using a command from an older version | 20:22 |
blake_r | saltlake: just visit the WebUI and go to the images page | 20:22 |
blake_r | saltlake: you will be able to start the import there | 20:22 |
saltlake | I did download a ubutnu image from the ui as well | 20:22 |
blake_r | saltlake: then that is all you need | 20:22 |
saltlake | However when I start a client node .. I see on the client nodes console: tftp://192.168.122.100/pxelinux.0 ... No such file or directory | 20:23 |
blake_r | saltlake: did you configure the cluster interface? | 20:23 |
saltlake | Yes the cluster interface looks connected | 20:23 |
blake_r | saltlake: no did you go the cluster page and configure an interface to provide dhcp and dns? | 20:24 |
saltlake | Do I have to add the ndoe manually on the UI or does it show up automatically | 20:24 |
saltlake | blake_r: yes I did | 20:24 |
blake_r | saltlake: okay and 192.168.122.100 is the ip address of the maas server? | 20:24 |
saltlake | blake_r: yes that is so and I set the interface to manage DHCP and DNS and gave it a DHCP range from 192.168.122.101 -200 | 20:25 |
blake_r | saltlake: okay | 20:25 |
blake_r | saltlake: check that the import completed successfully | 20:25 |
blake_r | saltlake: see if there is any errors in /var/log/maas/pserv.log | 20:25 |
saltlake | I just saw an error when I started the client VM | 20:26 |
saltlake | blake_r: I started the client VM thus: virt-install --connect qemu:///system -n client1 -r 1024 --arch=x86_64 --pxe --network=network:default,mac=AA:AA:AA:AA:AA:01 --vnc --accelerate --disk=/var/lib/libvirt/images/client1,size=30 | 20:26 |
saltlake | blake_r: I see this: WARNING The guest's network configuration does not support PXE | 20:26 |
blake_r | saltlake: that is fine, its a lie | 20:26 |
saltlake | blake_r: :-( oh | 20:27 |
blake_r | saltlake: you need to check that pserv.log for me | 20:27 |
blake_r | saltlake: you can just paste the whole thing if you want | 20:27 |
blake_r | saltlake: pastebinit /var/log/maas/pserv.log | 20:27 |
saltlake | blake_r:http://pastebin.com/HgREdVPi | 20:28 |
blake_r | saltlake: that is only listening of internal interfaces | 20:29 |
blake_r | saltlake: it should show it connected to an external interface | 20:29 |
saltlake | is that not enough ? | 20:29 |
blake_r | saltlake: does the server have an eth0? | 20:29 |
saltlake | Hmm.. even if it has an image downloaded to it ? | 20:29 |
saltlake | blake_r: the maas server has eth0 which is the cluster interface network 192.168.122.0 | 20:30 |
saltlake | network | 20:30 |
saltlake | the maas server is a VM | 20:31 |
blake_r | saltlake: okay lets restart the clsuter service | 20:31 |
saltlake | service maas restart ? | 20:31 |
blake_r | saltlake: "sudo service maas-clusterd restart" | 20:31 |
blake_r | saltlake: also I would recommend upgrading to the latest version that we have in stable, | 20:31 |
blake_r | saltlake: sudo add-apt-repository ppa:maas-maintainers/stable | 20:32 |
blake_r | saltlake: sudo apt-get update && sudo apt-get dist-upgrade -y | 20:32 |
saltlake | blake_r: thanks for the tip.. will do .. do you recommend aborting this totally ? | 20:32 |
blake_r | saltlake: aborting what? | 20:32 |
saltlake | blake_r: aborting working with 1.7 and update | 20:33 |
blake_r | saltlake: its just going to upgrade to 1.7.0 stable | 20:33 |
blake_r | saltlake: the version you have is a beta | 20:33 |
blake_r | saltlake: just small fixes no large changes | 20:33 |
saltlake | blake_r: ok I want to wait.. this is the most progress I have made so want to make some more before I tear it down and restart | 20:33 |
blake_r | saltlake: it can upgrade without a restart | 20:34 |
saltlake | blake_r: Do i restart the VM | 20:34 |
blake_r | saltlake: it will restart the services for you | 20:34 |
blake_r | saltlake: no need to restart anything | 20:34 |
saltlake | blake_r: ok I srestarted clusterd. Now I see one image in the images section of the UI | 20:35 |
saltlake | blake_r: the client1 boot window does not progress I think that is given up, do I try to add the node manually ? | 20:35 |
saltlake | blake_r: By going on the UI->add node ? | 20:35 |
blake_r | saltlake: no you still need the client to boot | 20:35 |
saltlake | blake_r: so restart the client VM ? | 20:36 |
blake_r | saltlake: yes restart the client vm to see if it pxe boots | 20:36 |
saltlake | blake_r: It appears to be installing!! | 20:37 |
blake_r | saltlake: awesome | 20:37 |
blake_r | saltlake: it will not install yet, it will just enlist the node into maas | 20:38 |
saltlake | blake_r: Muah!! Thank you brother!! So after a cluster interface config. . I shoudl restart the service.. !! | 20:38 |
blake_r | saltlake: you will need to go to the newly added node in the maas ui and add the power parameters | 20:38 |
blake_r | saltlake: no you dont need to do that, it will restart the dhcp service for you on save | 20:38 |
blake_r | saltlake: it was just that for tftp service wasnt listing on an outbound interface, for some reason | 20:39 |
blake_r | saltlake: could have been that maas-clusterd came online before eth0 was ready | 20:39 |
saltlake | blake_r: the client rebooted by itself .. is that normal | 20:39 |
blake_r | saltlake: since this is a vm, maas cannot setup the power parameters for you, you need to add that information | 20:40 |
blake_r | saltlake: maas did not reboot the machine, virt-install did | 20:40 |
blake_r | saltlake: you now should power the machine off | 20:40 |
saltlake | blake_r: OK I will try that now and see if the gues will start back up | 20:40 |
blake_r | saltlake: add the power parameters on the maas ui and then commission the node | 20:40 |
blake_r | saltlake: once the node has commissioned you can then acquire and start the node | 20:40 |
blake_r | saltlake: starting the node will install Ubuntu for you | 20:41 |
blake_r | saltlake: you will then be able to ssh into your newly installed ubuntu machine | 20:41 |
blake_r | saltlake: here is some information about using maas to control virtual machine nodes | 20:41 |
blake_r | saltlake: http://maas.ubuntu.com/docs/nodes.html#virtual-machine-nodes | 20:42 |
saltlake | blake_r: The status changed to commissioning.. any ideas how long this takes ? I have been stuck at this for many days | 20:43 |
saltlake | I mean in all the diff install tries i nthe last week | 20:43 |
blake_r | saltlake: it should take at most 10 minutes | 20:43 |
blake_r | saltlake: if its stuck there then something went wrong | 20:43 |
saltlake | blake_r: But should I see something on the console of the client1 vm ? | 20:44 |
saltlake | I see it stuck on "No bootable devices!! | 20:44 |
blake_r | saltlake: is maas able to power the node on and off? | 20:44 |
saltlake | blake_r: It did a power node on for sure | 20:44 |
blake_r | saltlake: oh that is virt-install for you! it didnt make the first device pxe | 20:44 |
blake_r | saltlake: you need to change the first boot device to be pxe | 20:44 |
blake_r | saltlake: i bet its not even set to boot from pxe | 20:45 |
saltlake | blake_r: It is actually | 20:45 |
blake_r | saltlake: does it get an ip? | 20:45 |
saltlake | blake_r: Its got IP 192.168.122.101 but the console window of the client1 vm does not mode. That said.. I just saw the xml file fir the client1 Vm. | 20:46 |
saltlake | It has boot dev='hd' | 20:46 |
saltlake | It should have it pxe as u said right !! | 20:46 |
saltlake | :q! | 20:47 |
blake_r | yeah "--pxe" with virt-install only means for the first boot | 20:47 |
blake_r | so if you reboot it manually it will then commission | 20:47 |
saltlake | blake_r: got it.. let me kill the VM and edit the xml and restart the vm in virsh | 20:47 |
saltlake | blake_r: Can I PM u for a minute ? | 20:48 |
blake_r | saltlake: sure | 20:48 |
saltlake | blake_r: I added the boot order to include nwteork in the client1 virsh file. I restarted the node commissioning This time it does not an IP address and virt-viewer client1 shows "Waiting for guest domain to start" | 20:58 |
saltlake | blake_r: Thanks a bunch!! The node finally moved from commissioning to "ready" | 21:00 |
blake_r | saltlake: awesome glad to hear it! | 21:01 |
blake_r | saltlake: no problem anytime | 21:01 |
saltlake | blake_r: Thanks really!! You have no idea how frustrated I have been with the state of commissioning!!!!! | 21:01 |
saltlake | Thank you!! | 21:01 |
blake_r | saltlake: haha, your welcome! | 21:02 |
=== blake_r is now known as blake_r-travel | ||
=== zz_CyberJacob is now known as CyberJacob |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!