Re: [foxboro] ICCAPI error

Quoting Duc M Do <duc@xxxxxxxxxx>:

> At 09:32 AM 8/29/06 +1200, Kevin FitzGerrell wrote:
> 
>  >
>  >Duc,
>  >
>  >ICCAPI says this is a "class= 1" error which means IPC error.
>  >
>  >IPC error= -29 is "abortion of connect request" (see E_ABORT_CON in
> the IPC
>  >manual). The messages says c_chkio error, but I would guess that the
> error
>  >originated with a c_send or c_receive call.
> 
> Kevin,
> 
> Thanks for the elucidation (though you lost me after the first
> sentence).
>
> I don't understand the error message (okay, hold the laughter!) since 
> we can load this compound into an off-line CP without any problem, yet 
> cannot load it into an FSim simulator. I've sent it to my technical 
> contact at SimSci-Esscor and we'll wait to see what she has to say.
> 
> My original question still remains, and that is where I can find a list
> 
> of error messages for ICCAPI (other than the terse listing in iccapi.h,
> 
> thanks, David!) that also has some explanation for the errors. They are
> 
> not in the ICCAPI document, obviously.
> 
> Regards,
> 
> Duc

Duc,

Sorry about that.

As far as I know there is no good coverage of possible error messages in the I/A
system in Foxboro's documentation.  The ICC manual does a fairly good job of
trying to identify errors you are likely to see in the ICC though.

I can tell you how I think the message breaks down though (anyone who knows
better please feel free to step in with corrections - this is based on my
limited experience digging into other error messages):

FAIL     5 load 23FS01!C0704: Mon Aug 28 08:37:32 2006
iccdrvr.tsk log entry showing command and FAIL status of that command and the
time it occured.

-15 
I'm having a little trouble with this "-15" - probably just means ICC error, see
/opt/fox/ciocfg/api/examples/iccdt.h

ICCload error:
iccdrvr.tsk says it the error return code was from an ICCload call.  

class= 1
OK, this tells us that the error was an IPC error (different error classes are
listed in /usr/include/fox/iccapi.h).  So, at this point we know the ICCload
failed, and the failure has to do with the IPC system.  The 

error= -29
This is the specific IPC error, E_ABORT_CON -29 "abortion of connect request". 
The IPC manual covers what return values you should expect from different types
of calls.

text= icipcmsg: c_chkio error
icipcmsg is an undocumented internal function in iccdrvr (but based on the name
I'd guess it's used for passing ipc messages back to icc), and c_chkio is a
documented iccapi call.  I would have expected an E_ABORT_CON response to a
c_send or c_receive rather than a c_chkio though.

After all that, it's still hard to tell what caused the problem.  When you use
icc (or iccdrvr) to interact with a CP, an IPC connection is opened between your
workstation and the CP.  It appears that this connection is failing.

Regards,

Kevin













 
 
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
 
foxboro mailing list:             http://www.freelists.org/list/foxboro
to subscribe:         mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe:      mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave

Other related posts: