[Planetlab-users] Configuring a service for automatic startup

[David E. Eisenstat]

The authentication mechanism that Node Manager uses seems to be hosed :(. 
I'll bother the people who maintain it.

Sirius is broken at the moment as well.

-David

On Mon, 20 Aug 2007, Adrian Suter wrote:

> Thanks David - chkconfig could be found. But the python command did not work, 
> or am I wrong? I've run as root the command
> python -c "import xmlrpclib; 
> xmlrpclib.ServerProxy('http://localhost:812/').Start('epfl_nguyen');"
> and here is the output
> *snip*
> Traceback (most recent call last):
> File "<string>", line 1, in ?
> File "/usr/lib/python2.4/xmlrpclib.py", line 1096, in __call__
>   return self.__send(self.__name, args)
> File "/usr/lib/python2.4/xmlrpclib.py", line 1383, in __request
>   verbose=self.__verbose
> File "/usr/lib/python2.4/xmlrpclib.py", line 1147, in request
>   return self._parse_response(h.getfile(), sock)
> File "/usr/lib/python2.4/xmlrpclib.py", line 1286, in _parse_response
>   return u.close()
> File "/usr/lib/python2.4/xmlrpclib.py", line 744, in close
>   raise Fault(**self._stack[0])
> xmlrpclib.Fault: <Fault 108: 'Permission denied.'>
> *snap*
>
> Maybe I just have to wait for a real reboot. Or do you have any idea?
> - Adrian
>
>
> David E. Eisenstat schrieb:
>> On Mon, 20 Aug 2007, Adrian Suter wrote:
>> 
>>> Hello everybody
>> 
>> Hi Adrian,
>> 
>>> I'd like to implement a long-running service that should start 
>>> automatically after a reboot of the planet-lab node. I've read and reread 
>>> the subsection 4.3 of the user guide on 
>>> http://www.planet-lab.org/doc/guides/user but I get always the message
>>> 
>>> *snip* bash: chkconfig: command not found *snap*
>> 
>> chkconfig lives in /sbin, which is not on the default PATH. Try 
>> /sbin/chkconfig instead.
>> 
>>> I would like to start up a perl script. Has anyone experience in this kind 
>>> of stuff? Is it possible to run an infinite loop (i.e. while(1) {...} ) 
>>> such that the script never stops, or are there some admin-scripts that 
>>> abort infinite runing perl scripts?
>> 
>> Yes, it is possible to run forever. Slices are killed only when they are 
>> the biggest consumer of memory on a box that's almost out OR when they 
>> expire.
>> 
>>> Today I have edited the file "/etc/rc.d/rc.local" but as the node did not 
>>> reboot yet, I can't say if this is working or not.
>> 
>> You can ask Node Manager to start your slice as if the node were starting 
>> up. NM supplies an XML-RPC interface on port 812, and the call is
>> 
>> Start('your_slicename')
>> 
>> In Python this can be accomplished by
>> 
>> import xmlrpclib 
>> xmlrpclib.ServerProxy('http://localhost:812/').Start('your_slicename')
>> 
>> -David
>