[jacorb-bugs] [Bug 925] Delay in serial (rxtx) communication

[bugzilla-daemon at jacorb.org]

http://www.jacorb.org/cgi-bin/bugzilla/show_bug.cgi?id=925





------- Comment #12 from jacorb at goots.org  2012-07-31 13:45 -------



> 
> Unfortunately, newest computers doesn't posses the serial port - you can use
> "usb-to-serial" converter cable but it costs - by us around 10 eur = 12.26 $.
> But in this case you need two converters - for each computer, and one
> "male-to-male" gender changer to connect the two converters.

As I have mentioned previously for specialised hardware support contact one of
the commercial companies at www.jacorb.org/support.html

> 
> Yes I have seen the code. I have no doubt in JacORB without gnu.io - I have
> done similar test in past. What I have seen interesting is that you create an
> array just once:
> and call it several times.

I did that for a couple of reasons - I wanted something operating quickly and I
wanted to only test the speed of marshalling not the time taken to construct
byte arrays/string.

> I plan to report another "bug", but, prior can you do some tests before - I

What about please?

> void and accept no parameters, namely delay1 and delay2. In delay1's body
> insert the code "Thread.sleep(5000)" and in delay2's body insert two void for
> loops "for (int l1= 0; i1 <=100000; l1++) for (int l2= 0; i2 <=100000; l2++);"

Using a slightly modified as per your instructions 2012/07/31 source code with
full debug logging - 

5047 ms
5013 ms
5012 ms
5021 ms
5017 ms
 === 
9 ms
5 ms
6 ms
5 ms
9 ms

i.e. as expected, the first batch is slowed down. The second operation does not
show much changes - in fact, the hotspot compiler might even those loops. Using
a Lenevo T420s with 8GB and Fedora 16 it has no affect on the system.

>  I don't know about
> Linux (which you use) but at me in Windows XP it blocks everything - does this
> has to do with virtual machine you mentioned in previous comment?

The virtual machine I referred to is your Java vm. Have you tried your tests on
another machine/operating system? How much memory do you have? Is your virtual
machine running using server class or client class mode?

>> Have you seen my previous questions and the simple test I have done with just
>> JacORB (showing no apparent delays) ?
> 
> I have done the tests with the sources you provided the delays are: 
> 0 ms
> 15 ms
> 0 ms
> 0 ms
> 16 ms

So does this mean for you this shows there is no apparent problem with JacORB?
[ for this case - of course there still may be issues as with any software ;-)
]


-- 
Configure bugmail: http://www.jacorb.org/cgi-bin/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.