Hello again,

I was wondering if it would be possible - as a "quick fix" for this problem - to override the transition where the exception is happening with different code that for now will just catch the exception? I understand ECharts allows doing something like this but as I said I'm not an expert, so I'm wondering if someone might be able to advise, it would be very much appreciated.

Thanks,
-- Hauke D

Hi Hauke
I have reproduced the exception with SailFin v2-b31g, using a KitCAT test case. As you have described, the exception occurs when B2BUA sends BYE to UA1, before it has received ACK to the re-INVITE/200 transaction.

Tom and I will discuss where best to catch the exception and do some testing. Will get back to you.

I put in the following change to TransparentHandleRequestFSM.ech to catch any exception while relaying a request:

Index: machines/org/echarts/servlet/sip/machines/TransparentHandleRequestFSM.ech
===================================================================
--- machines/org/echarts/servlet/sip/machines/TransparentHandleRequestFSM.ech (revision 1573)
+++ machines/org/echarts/servlet/sip/machines/TransparentHandleRequestFSM.ech (working copy)
@@ -93,9 +93,13 @@
// Propagate request to sender
//
transition INIT - [ ! getIncomingRequest().getMethod().equals("ACK") ] / {
+ <* try { *>
sender! (outgoingRequest = sender.createRequest(getIncomingRequest()));
// save requests as attributes
setOutgoingRequest(outgoingRequest);
+ <* } catch (Exception e) {
+ putEvent(new ErrorEvent(e, "Error sending request"));
+ } *>
} -> REQUEST_RELAYED
else /
// Do not expect to receive ACK


This works. But unfortunately afterwards SailFin seems to be confused. When UA2 retransmits 200 OK to the re-INVITE, SailFin gives it to the app again (incorrectly), causing another exception in E4SS.

Give it a try - you may get better result depending on the behavior of your UA.

Hi Eric,

Thanks for the fix. I was trying to overload the transition in my machine (I was hoping this would be faster than re-building echarts with this fix applied) but this doesn't seem to work - is this at all possible or am I just doing something wrong?

I defined "transition IN_CALL.ACTIVE.TRANSPARENT.TRANSPARENT.INIT - ... -> REQUEST_RELAYED etc." but I'm getting the error "semantic error: in transition source state configuration segment IN_CALL.ACTIVE.TRANSPARENT.TRANSPARENT.INIT of state configuration [IN_CALL.ACTIVE.TRANSPARENT.TRANSPARENT.INIT] of (sub)machine : Illegal access of state TRANSPARENT declared in machine org.echarts.servlet.sip.machines.B2buaSafeFSM". (Note: IN_CALL is the name of the state in my machine that invokes B2buaSafeFSM.)

Also, I just encountered a problem where ECharts doesn't seem to be compatible with Python 2.6 (apparently the new default with Ubuntu 9.10):

EChartsSipServlet_DK_2.5.1-beta/lib/AbstractMachine.py:1846: DeprecationWarning: catching of string exceptions is deprecated
except AbstractMachineException, message:
exceptions must be classes or instances, not str

This doesn't happen with Python 2.5.

Thanks,
-- Hauke D

hi hauke

the reason you're getting seeing the semantic error is because the TRANSPARENT state isn't declared to be "public" in the B2buaSafeFSM which means you can't reference it unless your overriding machine is declared in the same package as the B2buaSafeFSM (org.echarts.servlet.sip.machines) - so it appears that the only way around this is to re-compile E4SS regardless of the approach you choose (yours or eric's)

since i know it's a bit of a hassle to recompile E4SS i built a new echarts-sipservlet.jar file for you with eric's change in it for you - you can grab it from here - change the name of this file to echarts-sipservlet.jar and replace the file with the same name in your EDK_HOME/lib dir - you will also need to make the change to the TransparentHandleRequestFSM.ech file in EDK_HOME/machines/org/echarts/servlet/sip/machines to match the changes that eric described in his post

with regards to python 2.6, i'm surprised - i thought we had tested against this release last year - anyhow i'll give it a try tomorrow to find out what's going on - i see that you're only getting a warning so it shouldn't be causing any problem for you other then generating a lot of annoying warning messages

greg

Hi Greg,

Thanks very much for the information, and sorry for my late reply, I've been busy with other things for the moment - I'll post again once I get around to working on this some more.

Regards,
-- Hauke D