Re: [jgroups-dev] Screenshot of tree mesh structure

Mike Jensen <[email protected]>
Newsgroups gmane.comp.java.javagroups.devel
Message-ID <[email protected]>
Sorry for the delay, I been using this weekend to do some hiking before 
it starts to get really cold.

We don't have a great test environment to do 100's of nodes.  At work I 
was just running many jgroups instances within the same process to test 
the structures.  I also did some performance based testing, but just 
verifying that the latency does not grow too much as the nodes grow (but 
i recognize the numbers themselves are not very valid, since they are 
from within the same process).

At home I have 20gigs of ram spread across 3 machines.  I tend to use 
that for more real physical node testing.  But even then I definitely 
can't reach 100 nodes.  I might be able to squeeze a bit more, but I 
usually only run about 20 to 25 nodes here at my house.

I will have to look at ns2, I have never tried it actually.  Do you have 
any recommendations on setup to do good performance based tests?  Here 
is how my simple test actually conducts itself:
* Every 10 seconds every node sends a broadcast to the network which 
includes it's current timestamp
* Once a node receives a broadcast from another node it responds, 
including the original timestamp, but marking the message as a return, 
and routing it to the original node through the mesh (which may be 
several hops).
* Once we receive the message back, we see how much time has passed, and 
who was the node who sent it back to us.  We then record the time it 
took for node x to get our message, and send it back to us.

It's a fairly high load situation due to the original message being a 
broadcast.   But results were decent.  I want to extend this to include 
much more information.  For one thing, we need to track hops and we are 
not doing that currently.  Let me know what suggestions you have.  I 
recognize that the above example is probably not very good.  I was just 
trying to do something simple and quick to get a _general_ idea.  I also 
have little experience doing this kind of testing.

Thanks for your feedback.

Bela Ban wrote:
> Looks good ! Do you guys have test environments with 100s of physical 
> nodes, or do you use network simulators to test, e.g. ns2 ?
>
> Mike Jensen wrote:
>> I always come on here and ask for questions, I thought I would 
>> lighten things up with some progress.
>>
>> I still have some outstanding issues (for example the one I emailed 
>> about earlier, that I still have not gotten a chance to look into).  
>> But things are looking good, and I am even starting to get the QA 
>> team here at work setup to start testing this protocol.
>>
>> I am working on building some test programs to test performance 
>> (although I am not 100% decided on the best way to get those 
>> metrics).  But the first step was getting setup to run with very 
>> large mesh's.  I have done tests with 100's of nodes so far, and seen 
>> good results....At least nothing seems obvious, but it is hard to 
>> verify 100% at those sizes.  I have included a screenshot of the 
>> structure of a 50 node mesh.  Personally I am quite pleased with 
>> these results, and they seem very reproducible.
>>
>> I can provide an updated source code release if anyone is interested 
>> in playing with some slightly more stable code.  But personally I 
>> would say give me another week.  In another week I hope this stuff to 
>> become pretty darn solid.  That is also my estimate as to when I 
>> imagine my companies testers will start poking at this stuff pretty 
>> hard.
>>
>> Let me know if anyone has any thoughts, suggestions, questions, etc.
>>
>> ------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------
>>
>> ------------------------------------------------------------------------------ 
>>
>> Start uncovering the many advantages of virtual appliances
>> and start using them to simplify application deployment and
>> accelerate your shift to cloud computing.
>> http://p.sf.net/sfu/novell-sfdev2dev
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Javagroups-development mailing list
>>
>>   
>

------------------------------------------------------------------------------
Start uncovering the many advantages of virtual appliances
and start using them to simplify application deployment and
accelerate your shift to cloud computing.
http://p.sf.net/sfu/novell-sfdev2dev
_______________________________________________
Javagroups-development mailing list
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.