[Fusioninventory-user] Unable to find agent to run this job

Carlos Gunners goondaba at gmail.com
Fri Jan 18 13:37:55 UTC 2013


Update ..

I have found where the logs go for both glpi and fusioninventory, and I see
no issues of relevance being displayed in the logs ..

[Fri Jan 18 12:26:27 2013][debug] FusionInventory unified agent for UNIX,
Linux and MacOSX (2.2.7)
[Fri Jan 18 12:26:27 2013][debug] Configuration directory:
/usr/local/etc/fusioninventory
[Fri Jan 18 12:26:27 2013][debug] Data directory:
/usr/local/share/fusioninventory
[Fri Jan 18 12:26:27 2013][debug] Storage directory:
/usr/local/var/fusioninventory
[Fri Jan 18 12:26:27 2013][debug] Lib directory:
/usr/local/share/fusioninventory/lib

But the situation remains that:
-- the agent refuses to listen to the port defined (ie, 62354)  but is
running in daemon mode
-- now, from within the glpi plugin, the system consistently claims that I
have not configured an agent that is usable in netdiscovery or netinventory

I should also point out that I switched from a previous debian distro
install for the agent, into a cpan installation ..
Consequently, my  storage directory changed from /var/lib/fusioninventory
to /usr/local/var/fusioninventory .. But I cannot establish how this could
cause a break in communications between the agent and the server (which are
on the same host) ..

So, it does seem as if there is a clear break in communications, but I am
not exactly sure why ..

Possible options:
-- switch agent back to cron mode, and see if that helps
-- keep googling to see if one can ascertain exactly what config settings
are required to ensure http daemon is launched by the agent to listen on
specified port
-- follow code to see how the http element is invoked and why it is failing

IS THIS MAILING LIST IN USE AT ALL ?


On 18 January 2013 11:06, Carlos Gunners <goondaba at gmail.com> wrote:

>
> Hi,
>
> I have an agent setup to run from the same host where the glpi and the
> fusioninventory-plugin are also running from ..
>
> I have finally gotten the network discover and inventory modules
> functional on the host, and have run them successfully from the command
> line.
>
> However, when I try to initiate the action from the GUI (whether using a
> wizard) or as a new task, the server is unable to contact the agent .. the
> logs show that error message.
>
> On agents management page, I have enabled the host to run everything
> except 'wake on lan'  .. But on /etc/fusioninventory/agents.cfg, I do not
> have any settings that suggest the network inventory or discovery tasks are
> enabled (well, because I am not sure what settings I need, and also because
> I did not initially think it was essential in a server push situation)
>
> The situation remains the same whether I run the agent as a daemon or via
> cron ..
>
> Please, what do I have to do to get this to work ?  what am I missing?  ..
> Are there any logs that will give me some indications on what is going
> wrong ?
>
> Thanks in advance for any suggestions ..
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/fusioninventory-user/attachments/20130118/fadc64f3/attachment.html>


More information about the Fusioninventory-user mailing list