Ask Your Question
1

dnsmasq stop allocate host name to new hardware

asked 2015-05-06 10:28:31 -0500

PainInNetwork gravatar image

updated 2015-05-07 02:28:39 -0500

Hi,

I deployed an openstack cloud with Mirantis Fuel, it works fine. but when I move some hardware to deploy a new openstack cluster, the cobbler docker container allocate ip address to new hardwares, but doesn't allocate host names.

I can find the new hardware in Mirantis Fuel UI, add nodes page. the status of these hardware are "DISCOVERED".

normally when a new hardware is put into the environment, it will be added to the file /etc/dnsmasq.conf, with a line like below:

dhcp-host=net:x86_64,74:c9:9a:4f:ba:c8,node-60.domain,192.168.250.48

and under folder /var/log/remote, there should be a link from ip address to the hostname

lrwxrwxrwx 1 root root   29 May  5 10:38 192.168.250.48 -> node-60.domain

but now when the hardware "discovered" in Mirantis, I can not see any dhcp-host added to dnsmasq.conf file, and a folder with the allocated ip address will be created under /var/log/remote, instead of node-xx.domain.

drwxr-xr-x 3 root root   22 May  6 15:12 192.168.250.49

Is there any way to clean up these hardware discovery history in Fuel and force it to re-discover hardware and allocate host name again?

Thanks to omar-munoz! it's found that cobbler didn't update the discovered node.

[root@fuel ~]# fuel node
id | status   | name                     | cluster | ip             | mac               | roles                     | pending_roles | online | group_id
---|----------|--------------------------|---------|----------------|-------------------|---------------------------|---------------|--------|---------
66 | ready    | GEP5 Slot 1-5 (node-66)  | 16      | 192.168.250.51 | be:57:c6:cc:31:4b | compute                   |               | True   | 16
57 | ready    | GEP5 Slot 1-11 (node-57) | 16      | 192.168.250.44 | be:bc:bc:06:06:45 | compute                   |               | True   | 16
64 | ready    | GEP5 Slot 1-19 (node-64) | 16      | 192.168.250.52 | f6:7e:d3:39:6a:43 | compute                   |               | True   | 16
77 | discover | Untitled (ce:fe)         | None    | 192.168.250.50 | 00:13:5e:e9:ce:fe |                           |               | True   | None
73 | discover | Untitled (ce:8e)         | None    | 192.168.250.49 | 00:13:5e:e9:ce:8e |                           |               | True   | None
56 | ready    | GEP5 Slot 1-13 (node-56) | 16      | 192.168.250.42 | 2a:12:9f:f7:85:45 | compute                   |               | True   | 16
78 | discover | Untitled (18:7e)         | None    | 192.168.250.53 | 00:13:5e:ea:18:7e |                           |               | True   | None


[root@fuel ~]# cobbler system list
   default
   node-55
   node-56
   node-57
   node-58
   node-60
   node-64
   node-66
   node-67
edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
1

answered 2015-05-06 12:51:38 -0500

updated 2015-05-06 14:07:17 -0500

Can you try the following and once the node is removed, you will need to reboot the node so that it is discovered via PXE by Fuel.

Run these commands from Fuel Master

fuel node --delete-from-db --node node-60

then you need to remove this node from the cobbler with the following commands:

cobbler system remove --name node node-60
cobbler sync
edit flag offensive delete link more

Comments

Thank you for answer this question, omar-munoz!

With these 2 commands, the problem didn't fix, but I get more useful information. the cobbler node list is not updated after the node discovered. according to sequence diagram from mirantis(https://docs.mirantis.com/fuel-dev/develop/sequence.html),

PainInNetwork gravatar imagePainInNetwork ( 2015-05-07 03:53:30 -0500 )edit

seems the node already get bootstrap image, and get IP, DNS response from cobbler. I can ssh to it. the cobblerd service is running, and mcollective also started. from the cobbler container on Fuel master node, there's no update in cobbler.log for the newly added nodes

PainInNetwork gravatar imagePainInNetwork ( 2015-05-07 04:16:49 -0500 )edit

@omar-munoz , Also tried to restart the cobbler container, the node list in cobbler still not updated. and I can not find any useful information from cobbler.log. is there any way to analyze why cobbler node list is not updated?

PainInNetwork gravatar imagePainInNetwork ( 2015-05-08 04:13:04 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2015-05-06 10:28:31 -0500

Seen: 416 times

Last updated: May 07 '15