New answers tagged

1

I was able to complete my installation successfully by re-running the hook in a second terminal. juju resolved --retry keystone/0


0

In my case I have virtual machines and I can not power on any VM, I receive a Failed Commisioned message. my post


6

Try this. I have used the "next" PPA of conjure-up because it has a lot of fixes, but it also uses Juju 2.0 beta which is a beta :) Your call on whether you do that or use stock conjure-up from the stable archives. Note that the main archive will be updated to get the new conjure-up anyway, once Juju 2.0 is released. If you are going to build your cloud on ...


0

The packages and dependencies needed for OpenStack are in the Canonical Cloud Archive, which is not the archie you have installed. See https://wiki.ubuntu.com/ServerTeam/CloudArchive for details. You can use those packages together with puppet or chef or your preferred configuraiton management system to build whatever kind of openstack you like. However, ...


0

Well after sometime finally one friend advised me to set up a NAT on maas through iptables rules which I did and that made the deployed nodes capable of accessing to the Internet through maas. Here is what I did exactly on the maas server: iptables --flush iptables --table nat --flush iptables --delete-chain iptables --table nat --delete-chain iptables --...


0

I found that the API still supports this, I managed to exclude/reserve subnet ip's with the cli tool maas my-maas ipaddresses reserve ip_address=192.168.0.4


0

EDIT: I was able to get to the rpi_3_config part, but the next command gave an error. EDIT #2: I am now able to compile u-boot successfully from the official u-boot git repo EDIT #3: So far it looks successful with it's booting. Can't test it right now, but I will try with MaaS later! While you might not be able to PXE boot a Raspberry Pi 3 without an SD ...


0

MAAS should be using Ubuntu 14.04 to commission. Can you try to update MAAS and see if that solves the problem?


0

Every time a new machine or unit is created it increments that count in the model. In order to reset the counts you would need a new model.


0

I can't answer why it's no longer there, but I can confirm it's not in the API anymore. That being said, 2.0 is not in final release, and I believe is beta5, working on beta6. An official answer would be nice if WOL is going to return or if it was removed for a reason. Previously, 1.9 had ether_wake (Wake-on-LAN) https://maas.ubuntu.com/docs1.9/api.html#...


1

After reproduce this problem myself using JuJu/MAAS, I found out MAAS deployment install quite a few packages: Installing package: curl Installing package: cpu-checker Installing package: bridge-utils Installing package: rsyslog-gnutls Installing package: cloud-utils Installing package: cloud-image-utils Installing package: tmux Some of the packages ...


0

I suppose you could do either. However, depending on the size of your intended cloud, running MAAS separately from where you run openstack-install might make sense. Most deployments I would expect you to run openstack-install on your MAAS server, infact, the instructions intend on you doing exactly that. The first step of the installer is to install ...


3

While you may not be able to PXE boot via MAAS on the RPi3, juju supports manual provisioning including the bootstrap node - https://jujucharms.com/docs/1.24/config-manual With this you can install the systems with 16.04, then allow juju to provision them into the environment via SSH.


0

I could imagine a number of things going wrong but I suspect you may have defined an apt proxy in maas and perhaps formatted it like 'ip:port' instead of 'http://ip:port' Just a guess. That setting is in the MAAS UI, in the settings area.


1

Have you tried changing your Gateway IP on your MAAS set up? If the route to your outfacing NIC is not resolving, this could also be an IP forwarding issue. You might wanna check to see your firewall is not an issue here in case the traffic is blocked (though you can telnet to it).


1

1) just purged all juju packages and installed juju-core (1.25.5) 2) I set the timeout to a big number in juju environment.yaml 3) I deleted the node and recommissioned it again (MAAS vesion was 1.9.3) 4) I added no-proxy: 'list ip and your local network' in juju environment.yaml 5) I replaced spaces after "," in no-proxy definition in juju environment....


0

I figured I would answer this having more knowledge in CEPH then when I asked the question. Although, I can't say why the defaults are the defaults in Horizon. If they are generated or just a starting place for openstack. Landscape drive usage is pulled from ceph, which already includes replication. cluster eabab000-231c-4edb-855c-dd30b1bab56d ...


0

On the maas region controller, it needs to have a 10.20.x.x IP (the 10.20.x.x IP assigned on the maas) and be telling its VMs its helping boot to reach it there. This IP configured in /etc/maas/region*.conf. It likely is set to 192.168.122.159 right now. Hopefully there is a 10.20.x.x IP on the maasrc instance and this setting needs to be updated. By ...


0

berryboot will let you do network booting but its not real PXE http://www.berryterminal.com/doku.php/berryboot


0

Well first I must say all the credits must go to our friend "bc2946088" who gave the vital clue, thank you again dude. As he said I ran the sudo dpkg-reconfigure maas-region-controller command and changed the auto detected ip to 10.1.1.100. I also on the local PC from which I ssh to the maas, ran the route -p ADD 10.1.1.0 MASK 255.255.255.0 172.16.63.100 ...


0

As a hack, this is what I usually do: sudo dpkg-reconfigure maas-region-controller Do NOT change anything. Simply press ENTER. It would stop some of the processes you may need restarted. After the ENTER, it would start them back up. +1 for the answer from lanous.



Top 50 recent answers are included