[et-mgmt-tools] Cimbiote methods

[Michael DeHaan]

Michael DeHaan wrote:
> Bart Baars wrote:
>>
>> Hi all,
>>
>> For a project at the dutch tax-office, we are planning to use CIM. Our
>> primairy goal is to be able to start/stop services that have a script in
>> /etc/init.d. Unfortunatly, we haven't been able to find a 
>> MOF/provider that
>> is able to do this for us. So we decided to write our own.
>>
>> We want to use Cimbiote for this. The example you wrote 
>> (example_plugin.py)
>> is pretty clear, except for the methods. I believe there are tho
>> methods, set and get_array_element, and they both require some args.
>>
>> I am trying to access the method using this command:
>>
>> # wbemcli cm
>> 'http://localhost/root/cimv2:ExamplePlugin.Filename="file-9.txt"'
>> get_array_element
>>
>> This returns "a general error occured, not covered by more specific 
>> code",
>> and I should check some logs. I have tried various options for the 
>> method
>> invocation, but I never get a valid response. Can you give me an 
>> example of
>> how to execute this method?
>>
>> Regards,
>>
>> Bart
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> et-mgmt-tools mailing list
>> et-mgmt-tools at redhat.com
>> https://www.redhat.com/mailman/listinfo/et-mgmt-tools
>
> Wow, a post about Cimbiote :)   I'm happy and sad at the same time.   
> You're basically the first person who has asked about it.    Cimbiote 
> is now largely unmaintained, and I should update the web page to 
> indicate we are looking for a new owner.    I would really like to see 
> someone who is well connected with CIM development help continue it.
>
> At the time when it was written we saw a lot of large company interest 
> in CIM and wanted to help out in making writing providers easier for 
> those that were not CIM experts -- writing CIM providers is way too 
> painful, and documentation is scarce.     Standardized interfaces, we 
> thought, were definitely a good idea, but the problem was these 
> standards were overcomplicated and that created a high barrier to 
> entry for open-source players (who generally preferred simpler and 
> more transparent technology).  Unfortunately, we were unable to drum 
> up community interest among pegasus-list and open source CIM project 
> leaders.   As we generally prefer secure XMLRPC ourselves (and are 
> also looking forward towards message bus systems like QPid -- 
> http://cwiki.apache.org/qpid/), there wasn't a lot of sense in 
> continuing to invest in it without community interest from a few CIM 
> experts that could help us out.   I should probably update the 
> Cimbiote page to better address it's current status. 
> As I recall, the method support did work fine ... though I forget how 
> to use wbemcli myself.   There were some strange issues with it and 
> the way it parsed arguments.   Most likely wbemcli is not reading the 
> argument list correctly, and we did encounter a few bugs where calling 
> some CIM functions though wbemcli didn't appear to work (using other 
> providers too, not just with Cimbiote).
>
> Maybe I can help you out though ... we used some of the sblim 
> providers for reference when working on Cimbiote, and 
> http://sblim.wiki.sourceforge.net/ might have a service provider.  I 
> believe it did, though I do not remember exactly what it supported.
> Hope that helps out some.   If you have any interest in contributing 
> to the project, you are definitely welcome to.
>
> Thanks,
>
> --Michael
>
>
>
>
>
>
>

I just remembered that we were using raw WBEM XML with "wbemexec" in the 
few instances that wbemcli wasn't working for us.   Unfortunately I 
don't have those xml test files lying around anymore.
If you want to get involved with something a bit newer, you might want 
to look at QPid/AMQP and developing your own way of sending commands to 
and from your nodes rather than CIM.  Message buses
and CIM are of course totally different, but I think there is a lower 
barrier to entry and a lot more flexibility there.   In particular, I'm 
not particularly fond of the way CIM handles events (indications)...

--Michael