Bug #6377
closed
Hostname changed to localhost - DNS configuration problem
Description
Hi everyone,
I upgrade my rudder-server-root and my nodes last week from rudder 2.10 to rudder 3.0.1
Today, I have 2 nodes which have changed their hostname to 'localhost'.
So I have in my list 2 nodes with 'localhost' as hostname...
I have tried to do a rudder agent reinit on the 2 nodes but I have the same result in the section 'Accept new nodes', they come back with 'localhost'.
And if I try to accept them, I have no error but it is not working.
However when I launch rudder server debug, I see the node with the good hostname in the log.
Have any ideas ?
Thanks for your help :)
Updated by François ARMAND over 9 years ago
- Category changed from Web - Nodes & inventories to 26
- Priority changed from N/A to 3
- Target version set to 3.0.3
If the node's hostnames are correctly view with server debug, it may be a problem lying in the inventory sent by these node. Could you please attach one of the two inventories here so we can try to reproduce ?
If you can't for privacy reason, could you tell us the content of the field:<HOSTNAME> (under <REQUEST>) and <NAME> (in the <HARDWARE> sub-tree).
Thanks,
Updated by Clem Def over 9 years ago
Hi,
This the content of field of the ocs file from /var/rudder/inventories/received
Under Request : <HOSTNAME>axium</HOSTNAME>
Under Hardware : <NAME>Axium</NAME>
Thanks
Updated by François ARMAND over 9 years ago
Unfortunally, that seems ok :/
Could you send me one inventory privately if you can't/don't want to publish it ? francois.armand at normation.com ?
Updated by François ARMAND over 9 years ago
OK,
With further analysis, it appears that the <HOSTNAME> tag is present at two places in the inventory: one is under the <RUDDER> tag and is filled with the Rudder fusion inventory plugin. The other is at the end, and is added as a post processing, and that the two values are not the same.
The <HOSTNAME> tag under <RUDDER> is the most prioritary, and is defined to "localhost".
The command used for getting it localhost is: hostname --fqdn? Could you check the result of that command on the bad nodes ?
Thanks
Updated by Clem Def over 9 years ago
Oh, Nice!
Indeed, on the two nodes hostname --fqdn have as output localhost.
But, do you think that this problem is appeared with the upgrade ?
I did not change the hostname or other DNS between the version 2 and 3...
But after change the location of 'localhost' in my /etc/hosts, it's working !!
Example : "127.0.0.1 YOUR_NAME_INTO_RUDDER localhost"
Thanks a lot François! :-)
Updated by Benoît PECCATTE over 9 years ago
- Category changed from 26 to Web - Nodes & inventories
Updated by François ARMAND over 9 years ago
- Subject changed from Hostname changed to localhost to Hostname changed to localhost - DNS configuration problem
Well, no I have no idea why the upgrade triggered the problem.
What version did you updated from ?
Updated by François ARMAND over 9 years ago
- Description updated (diff)
- Reproduced set to No
Updated by Vincent MEMBRÉ over 9 years ago
- Target version changed from 3.0.3 to 3.0.4
Updated by Vincent MEMBRÉ over 9 years ago
- Target version changed from 3.0.4 to 3.0.5
Updated by François ARMAND over 9 years ago
- Status changed from New to Rejected
I'm going to close that one since we were not able to reproduce that one.
Clément (or anybody else, actually), if you see that happening again, please reopen the issue, if possible giving new information that could help reproduce.