perfsonar development work

[Andreas Hanemann <>]

My statement was not precise enough. The first local tests can be
carried out disregarding network-related effect, while the consequences
of network delay, etc can be determined in a second series of tests. The
latter should lead to a realistic assumption how many requests a service
can handle.

Best regards
Andreas

Fausto Vetter wrote:
> Hi,
>
> Maybe, Why not testing on local machine too, because LAN also may add
> undesired delay to the process.
>
> Fausto
>
> Andreas Hanemann wrote:
>
>> Michael Michalis wrote:
>>  
>>
>>>  
>>>   
>>>> -----Original Message-----
>>>> From: Andreas Hanemann 
>>>> [mailto:]
>>>> Sent: Thursday, May 10, 2007 2:39 PM
>>>> To: Michael Michalis
>>>> Cc: Nina Jeliazkova; Loukik Kudarimoti; Ilias Tsompanidis; perfsonar-
>>>> ;
>>>>  Athanassios C. Liakopoulos; Jochen Reinwand
>>>> Subject: Re: [pS-dev] Performance Testing
>>>>        
>>> Hi Andreas
>>>
>>>  
>>>   
>>>> Hi Michalis,
>>>>
>>>> I have used the commented version from Nina and added some comments on
>>>> my own. The testing method that you describe is very useful at
>>>> least as
>>>> a first step. We have to see whether we need more specific tests
>>>> later.
>>>> These could be needed to exactly determine the conditions of a service
>>>> performance problem (e.g. the differences between response time for LS
>>>> registration and deregistration requests reported during the perfSONAR
>>>> meeting by RNP).
>>>>
>>>> Best regards
>>>> Andreas
>>>>        
>>> Thanks for your feedback. I've also placed my comments inside the
>>> doc but I
>>> would like here to discuss a common comment that you and Nina made
>>> about my
>>> suggestion for using a local network in order to run the tests. My
>>> original
>>> thought was that we need to check our services performance alone.
>>> Using a
>>> WAN to run the tests would certainly import latencies that could
>>> affect our
>>> results because there is no way to remove network latency out of our
>>> measurements. So in a local network the results will only reflect the
>>> performance of the service and not the performance of the network.
>>>
>>> Now based on your feedback I'm guessing the reason that you would
>>> like to
>>> use a WAN for the tests is that different types of requests may take
>>> different amount of time to get through. If this is the case, then I
>>> suggest
>>> to first run the tests on local network, so we can have a clear view
>>> of the
>>> service performance and then conduct the tests on a WAN configuration
>>> testing the entire system's performance. I would be very interested
>>> to hear
>>> your opinion on the latter, and then I could produce an updated
>>> document.
>>>
>>>  
>>>   
>> This is a good proposal. My idea was not to say that tests on a LAN are
>> useless, but to suggest additional WAN-based tests.
>>
>> Andreas
>>  
>>
>>>    
>>>
>>> Michalis
>>>  
>>>   
>>>> Nina Jeliazkova wrote:
>>>>        
>>>>> Hello Michael,
>>>>>
>>>>> Please find my comments in track changes (attached).
>>>>>
>>>>> Best regards,
>>>>> Nina
>>>>>
>>>>> Michael Michalis 
>>>>> <>
>>>>>  wrote:
>>>>>
>>>>>
>>>>>            
>>>>>> Hi all,
>>>>>>
>>>>>>
>>>>>>
>>>>>> I've put together an initial document describing performance
>>>>>> testing. I
>>>>>> would greatly appreciate any comment or suggestions made on the
>>>>>>                
>>>> document.
>>>>        
>>>>>>
>>>>>> Best Regards,
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Michalis Michael
>>>>>>
>>>>>>
>>>>>>
>>>>>>                
>>>>>            
>>>> -- 
>>>> Andreas Hanemann, 
>>>> 
>>>> Boltzmannstrasse 1, 85748 Garching, Germany
>>>> Telefon:   +49 89 35831-8712
>>>> Fax:       +49 89 35831-9700
>>>>        
>>>  
>>>   
>>
>>  
>>
>

-- 
Andreas Hanemann, 

Boltzmannstrasse 1, 85748 Garching, Germany
Telefon:   +49 89 35831-8712
Fax:       +49 89 35831-9700