certi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [certi-dev] HLA_TestsSuite usage / dtest problems 3


From: Gotthard,Petr
Subject: RE: [certi-dev] HLA_TestsSuite usage / dtest problems 3
Date: Fri, 27 Jun 2008 09:17:44 +0200

Eric,
I think I know where's the problem: the .py test executes e.g.
test_TRTCCallbacks. This command is executed from the $HOME directory.
First I had the "command not found" error there. After I've inserted a
path to HLA_TestsSuite into my $PATH in .bashrc and I got a "RTI
exception: CouldNotOpenFED []." error.
Should I create another environment variable to tell the system where to
find the .fed files? (Isn't it getting too complicated to setup the test
environment?)

Petr

-----Original Message-----
From: address@hidden
[mailto:address@hidden On
Behalf Of Eric Noulard
Sent: Friday, June 27, 2008 8:51 AM
To: CERTI development discussions
Subject: Re: [certi-dev] HLA_TestsSuite usage / dtest problems 3

2008/6/26 Gotthard,Petr <address@hidden>:
> Eric,
> small success: this is what makes the ./dtest_create_destroy.py 
> running
>
> 1) the `ssh localhost pwd` must be working fine
> 2) the environment variables must be configured in .bashrc (I've 
> observed that the environment variables cannot be configured by a 
> manually executed script. Due to the ssh localhost issue, everything 
> must be in .bashrc or similar file.)

No you don't need to add envvar to your bashrc, you "only" need to setup
CERTI_HOME on the machine you run the script or use --certi_home
argument of the dtest scripts (try dtest_test_TRTCCallbacks.py --help)

The dtest scripts should be setting the needed environment using the
CERTI installation provided script.

If you open (for example) dtest_test_Sync you may read:

dtest.ReusableSequences.addConditionalRunShellScript(rtig,c_shell_cmd="s
ource
"+certi_home+"/share/scripts/myCERTI_env.csh "+rtig_param['host'],
                               bourne_shell_cmd="source
"+certi_home+"/share/scripts/myCERTI_env.sh "+rtig_param['host'])

which tries to guess you shell kind (C or bourne one) and launch
appropriate scripts in order to setup environment.

> 3) and you MUST NOT have a welcome message in your .bashrc ;-) A 
> welcome message confuses the SFTP used by dtest and the whole test
hangs up.

Ok I did not encounter that one,
this looks like a paramiko bug (the ssh python implemation used by
dtest), already encountered by others:
https://bugs.launchpad.net/paramiko/+bug/239709

I did add a comment.

> However, `make test` still fails for all 3 cases and so fail the other

> 2 ./dtest_xxx.py test cases (just create_destroy works). Attached is 
> the text output I got.

When running a dtest script like dtest_test_TRTCCallbacks.py the script
produce .in, .err, .out files.

.out contains the output monitored by dtest in order to perform the
dtest and .in contains data sent by dtest to the monitored program.
(the .err file is Currently void but it may change in the future)

Would you please send here the files:
test_TRTCCallbacks_First.out
test_TRTCCallbacks_First.in

corresponding to dtest_test_TRTCCallbacks.py

--
Erk


-- 
CERTI-Devel mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/certi-devel




reply via email to

[Prev in Thread] Current Thread [Next in Thread]