The documentation provides a succinct example, and one that looks like it's missing a hop/pair anyway. When I set up a script to access a remote server and pattern my NTLM communication similarly, I get the following output:
...
All clients connected, Test beginning.
Invalid code received.
Error accepting remote connection.
message: Run error detected, terminating clients...
message: Terminating all instances of wcclient...
Connecting to:
...
My scenario works just fine when I comment out the 3 NTLM centric lines. It fails when they are uncommented. The WCAT client is able to parse the scenario, but I've not actually dug into the parsing mechanism yet to see whether it's parsing correctly.
(I've already fixed the bug that prevents connection from multiple remote clients, but there may be more.)
Here's the pertinent part of the scenario:
request
{
url = "http://weatherforecast-d/";
statuscode = 401;
}
request
{
url = "http://weatherforecast-d/";
// problems begin here...
authentication = "ntlm";
username = "my username";
password = "my password";
statuscode = 200;
}
If anyone has a running, functional example of a WCAT NTLM script, I'd be much obliged. Thank you.
I've cleaned up the logging in wcat.wsf and done numerous comparisons. There's got to be a better way to do this, but I don't know it.
- "authentication" parses while other parm names do not
- "username" and "password" don't cause fatal errors on their own
- If I comment out the single parm line "authentication" the test runs
- If I uncomment the same line it errors fatally
- The script output is exactly the same up until the failure
- basic or ntlm result in the same failure, as does any garbage text
The problem seems to be hidden deep in the wcclient. I guess I'll have to start looking for another product.