[rudder-users] Unable to add the same node

[Nicolas Charles]

On 01/03/2013 09:19, Olivier Desport wrote:
>
>
>
>
> Le 01/03/2013 09:11, Nicolas Charles a écrit :
>> On 01/03/2013 08:59, Olivier Desport wrote:
>>>
>>>
>>>
>>>
>>> Le 28/02/2013 14:29, Jonathan Clarke a écrit :
>>>> Hello Olivier,
>>>>
>>>> On 28/02/13 12:26, Olivier Desport wrote:
>>>>> Hello,
>>>>>
>>>>> I'm working with Rudder 2.4. I've deleted a node in the web 
>>>>> interface because I had to install a new server with the same 
>>>>> name. When I try to add this node, it never appears in "Accept new 
>>>>> node" menu. Can you tell me the way to clean this node in Postgre, 
>>>>> LDAP and files ?
>>>>>
>>>>> The logs for this node on the Rudder server :
>>>>>
>>>>> Feb 28 12:21:47 celebrimbor cf-serverd[17935]:  Accepting 
>>>>> connection from "172.31.136.112"
>>>>> Feb 28 12:21:47 celebrimbor cf-serverd[17935]:  REFUSAL of request 
>>>>> from connecting host: (SYNCH 1362050507 STAT 
>>>>> /var/rudder/share/d2abce4d-630e-4d05-932f-8e2b87942392/rules/cfengine-community)
>>>>
>>>> For the node to reappear in "Accept new nodes", it needs to send a 
>>>> new inventory. However, by default this only happens at night 
>>>> (between 0:00 and 6:00).
>>>>
>>>> You can force the node to run an inventory *now*, by running:
>>>> /var/rudder/cfengine-community/bin/cf-agent -KI -Dforce_inventory
>>>> (see 
>>>> http://www.rudder-project.org/rudder-doc-2.4/rudder-doc.html#_inventory) 
>>>>
>>>>
>>>> Then, it should appear in the Accept New Nodes screen (after 5-10 
>>>> minutes).
>>>>
>>>> Let me know if this works for you!
>>>
>>> I've launched this command several times but there is still no 
>>> pending node.
>>
>> Ha. This is probably an unfortunate bug then. Could you :
>> - force the node to run an inventory with the 
>> /var/rudder/cfengine-community/bin/cf-agent -KI -Dforce_inventory command
>> - wait for it to be propagated within Rudder (it should take about 5 
>> minutes)
>> - On the policy server, dig into 
>> /var/log/rudder/webapp/2013-03-01.stderrout.log, and look for 
>> messages like the following
>>     "Can't parse the input report, aborting"
>>     "Error when trying to parse report"
>>     "Exception when processing report"
>> And copy the content of the text here, it will help us understand 
>> what's going on
>
> This type of messages doesn't appears about this node in 
> /var/log/rudder/webapp/2013_03_01.stderrout.log :
>
> [2013-03-01 09:00:12] INFO  net.liftweb.http.SessionMaster -  Session 
> 1nflzcr9e6hv6zchpvizyboox expired
> [2013-03-01 09:05:13] INFO  net.liftweb.http.SessionMaster -  Session 
> 1o6xlcj9i8voakydposa31m7q expired
> [2013-03-01 09:10:23] INFO  net.liftweb.http.SessionMaster -  Session 
> qkq84zos31rjgja2wcuje9h1 expired
> [2013-03-01 09:15:13] INFO  net.liftweb.http.SessionMaster -  Session 
> 1qcpzgo8oelnbnvhgs8nwybvo expired

So it seems the inventory doesn't make its way to the policy server.
Did you change the IP address of the node, so maybe it has now an 
address outside the accepted IP range ?
Could you check if the inventory is present on the policy server in 
/var/rudder/inventories/incoming, right after you run 
/var/rudder/cfengine-community/bin/cf-agent -KI -Dforce_inventory on the 
client ?

Nicolas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.rudder-project.org/pipermail/rudder-users/attachments/20130301/4df7be2e/attachment.html>