Yep. But are we ok at how we manage this from the poller point of view?
I don't know if we need a new check interface if we manage the PollerModule
as a standalone worker. I think this last one sould :
* have a main() like current (fork) workers, so get from a Queue Checks, and
then run them.
* h
> unless someone has begun the work on this : I'm all open to
> do this as I manage quite very well that part of the code ..
i didn't have the time, so it'll be fine if you could care for it.
Gerhard
--
Special Offer--
2011/2/1 Gerhard Lausser
>
> > We should so export nearly all Class from shinken-arbiter to
> > a file in the shinken lib and then just import them and call the main.
> exactly. So that the bin/-files only contain a minimum of code (cmdline
> arguments parsing) and all the rest is in libraries in
2011/2/1 Grégory Starck
> [...]
>
>
>> (you dev under Windows isn't it?)
>>
>
> (douhhh ; not at all :p but not a lot better than windows too: ubu*)
>
> Ok, so we are two :D
>
>
> but it's solved, and another very harsh : with the Hudson daemon it seems
>> that there problem with the os.getlogi
2011/2/1 nap
>
>
> 2011/2/1 Grégory Starck
>
>> [...]
>>
>
>> [...]
>>
>
>> So do not hesitate to review some of these changes and spit on it ;)
>>>
>> I've seen some problems with the __import__ of the shinken-arbiter
>
> (you dev under Windows isn't it?)
>
(douhhh ; not at all :p but n
> We should so export nearly all Class from shinken-arbiter to
> a file in the shinken lib and then just import them and call the main.
exactly. So that the bin/-files only contain a minimum of code (cmdline
arguments parsing) and all the rest is in libraries in the shinken subdir.
Gerhard
> It's my fault here :)
Ah, ok *spit* :-))
> It was for making the __import__('shinke-arbiter') work in
> the test. I thought under Windows it will duplicate the file,
> but no, it just failed :(
Although there are some kind of symlinks in Vista and Win7, they require
admin privileges.
And
On Tue, Feb 1, 2011 at 3:05 PM, Gerhard Lausser
wrote:
> I forgot...
> Would it be an option to put the Arbiter, IArbiters classes in
> shinken-arbiter.py and just the __main__ part in shinken-arbiter?
>
So remove a lot of code for the shinken-arbiter (no .py) name and ask for
the rest of the code
On Tue, Feb 1, 2011 at 3:01 PM, Gerhard Lausser
wrote:
> Hi,
>
>
> > don't like with nagios perfdata checks syntax is that the
> > plugin checks have to return the min/max perfdata value
> > thresholds in the output.. I find it's not needed and
> they don't have to. But if they do, it's great hel
5:02
> An: shinken-devel@lists.sourceforge.net
> Betreff: Re: [Shinken-devel] Design for python-based checks
>
> Hi,
>
>
> > don't like with nagios perfdata checks syntax is that the plugin
> > checks have to return the min/max perfdata value thresholds in the
> >
Hi,
> don't like with nagios perfdata checks syntax is that the
> plugin checks have to return the min/max perfdata value
> thresholds in the output.. I find it's not needed and
they don't have to. But if they do, it's great help for tools like pnp to
scale the y-axis properly.
> missed s
2011/2/1 Grégory Starck
> [...]
>
> Here I add a point to Jean : having checks/plugins in native python is a
> question, having perfdata results returned from theses native calls also in
> such a native python format/object is another question.. So this specific
> question can be left on the si
2011/2/1 nap
> Hi,
>
> Thanks for this thread, it seems interesting :)
>
it's always interesting when brains/ideas are competing each other ;)
On Mon, Jan 31, 2011 at 11:42 PM, Hartmut Goebel > wrote:
>
>> Am 31.01.2011 23:20, schrieb Gerhard Lausser:
>> > a single format for both python-base
Hi,
Thanks for this thread, it seems interesting :)
On Mon, Jan 31, 2011 at 11:42 PM, Hartmut Goebel wrote:
> Am 31.01.2011 23:20, schrieb Gerhard Lausser:
> > a single format for both python-based checks and traditional plugins
> > requires converting the output and exitcode of the latter. Thi
> > What about min/max/thresholds?
> I thought, simple lists of values are enough. Am I missing something?
http://nagiosplug.sourceforge.net/developer-guidelines.html#AEN201
Performance data usually are more complex than "label=value"
Gerhard
__ Hinweis von ESET NOD32 Antivirus, Signa
Am 31.01.2011 23:20, schrieb Gerhard Lausser:
> a single format for both python-based checks and traditional plugins
> requires converting the output and exitcode of the latter. This probably
> adds a lot more cycles.
While you are right here at the moment, this may change over time.
The current i
Am 31.01.2011 23:07, schrieb Aurélien Degrémont:
> Regarding check, verifications, data handling, etc...
> If you've got an object it will be easier to add such manipulation
> directly in the CheckResult class.
You are absolutely right here. I normay do this kind of stuff. I just
changes the draft
Hi
> Defining a single format of passing results simplifies handling
> and saves some cycle.
a single format for both python-based checks and traditional plugins
requires converting the output and exitcode of the latter. This probably
adds a lot more cycles.
>
> Question: should mess
Hello
Some comments on the design below:
> * checks have to return a 4-tuple:
By experience when doing similar developments in Python, create an
object here.
It will simplify the data handling and could be easily extended later.
That's the big point.
If you go for a tuple-like interface, you wi
Hi,
well, I did not had enough time at the weekend to write a base-class
for Python-bases checks, but I has some time to think about the
design.
Here are my thoughts, comments are welcome.
* checks have to return a 4-tuple:
- return code (exit status): integer
- message (status output): uni
20 matches
Mail list logo
