mdepriest wrote:
> When I force zenmodeler to run, only the 'uname' plugin is executed, which is 
> in contrast to a Linux device, which gets them all run, and gives me a clear 
> error when it can't run because the device is unreachable.
> 
> Moreover, there's no evidence that the 'uname' plugin actually updated 
> anything in the model even after running apparently successfully.
> 
> So I'm still at a loss to figure out how to bridge the concept gap. I have a 
> feeling that somewhere the results of the zenoss.cmd.uname command are being 
> used to determine how zenmodeler behaves, but I can't figure it out. 
> Otherwise, I can't make any sense of the inability of the modeler to at least 
> attempt to use the zenoss.cmd.linux.* plugins.


I've done some more digging and I think I am starting to get it. Part of the 
concept gap is I'm not a Python guy, and the difference between a .py and .pyc 
file doesn't immediately jump out. (You young whippersnappers and your new 
languages can get off my lawn - if it were left up to me we'd all be writing in 
VAX Fortran on VMS and scripting it with DCL - and we'd LIKE it!)

As part of my investigation I copied the 'linux' plugins to a new name and 
started hacking on them. There's a place in the plugin files where the base 
class method 'condition' can be overridden - this is the missing link I was 
seeking. I changed the reference from 'Linux' to the new uname value that 
zenoss.cmd.uname was uncovering, and then ran a zenmodel.

Of course this did nothing. Then I got to wondering about those .pyc files and 
finally realized they were the compiled version of the .py files I had been 
editing. Not knowing any better I just deleted all of them to try to force a 
dynamic recompile. (Hey, I told you I don't know Python, I was just guessing at 
this point.)

Of course it didn't work. Undaunted, I did a 'zenoss restart' and lo and 
behold, there were new .pyc files. I then remodeled my test device and the new 
modeling plugin was invoked. Huzzah!

I will put together a more coherent discussion on this topic in the near future 
after I build out a complete test case for an example.




-------------------- m2f --------------------

Read this topic online here:
http://forums.zenoss.com/viewtopic.php?p=28661#28661

-------------------- m2f --------------------



_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to