1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
|
$Id$
This test involves a multithreaded client. The purpose of the test is
to verify that client-side and server-side RTCORBA processing works ok
when multiple threads are making invocations.
NOTE:
-----
Since output verification script relies on ORB debug messages, i.e.,
it checks for certain debug messages, it is BRITTLE. Whenever ORB
developers modify/remove certain debug messages in the ORB, the output
verification script may start reporting that there is an error, since
something it was looking for was not found. In such case,
process-output.pl must be updated to reflect debug output changes.
Description
-----------
Server creates an object with RTCORBA::CLIENT_PROPAGATED priority
model, and writes its IOR to the file.
Client spawns two threads of different priorities. Each thread sets
RTCORBA::ClientProtocolPolicy override to a single protocol on the
PolicyCurrent level. After setting the override, each thread performs
a number of invocations on the server object in a loop. The number of
invocations, each thread's priority and protocol are all specified
through command-line arguments.
This test consists of two parts. In the first, we run client and
server, which follow the description above. We also turn on the
debugging output for the server, and collect the entire test output
(client and server) into a data file. The second part of the test is
verifying the output of the test using process-output.pl perl script.
(The perl script is used since a large number of iterations results in
a sizeable output, which is impractical to verify manually).
run_test.pl does both, i.e., it runs client and server and then runs
the verification script on the output. See run_test.pl for
instructions on how to run both client/server and the verification
script.
|
