First it was bad directory for exporting files. This is because if you
do something like this:
String dirPath = "target" + File.separator + "export-import";
String dir = "foo".replace("foo", dirPath);
then result on Windows is: "targetexport-import" and not
"target\export-import" as expected :-)
Solved with Matcher.quoteReplacement.
Also I didn't remove System properties related to export/import after
finish the test, so it was trying to import previously exported stuff in
other tests. Deleting users after test wasn't sufficient as still next
test was import them from previously exported files and hence was able
to saw them. Not sure why it isn't problem on my laptop as well (maybe
different test order?)
Finally on some windows systems, when you call "server.stop()", it takes
some time until undertow server is really stopped. This seems to be the
issue just on Jenkins windows machine (not on the remote machine in
Brno, which I used for remote testing). Still I added some code to
AbstractKeycloakRule.stopServer() to wait until server is really stopped.
On 16.7.2014 21:35, Bill Burke wrote:
Passes. Thanks. What was the problem?
On 7/16/2014 2:26 PM, Marek Posolda wrote:
> Finally it passes for me on Windows and also with CI, so I hope it pass
> for you as well:-)
> PR is https://github.com/keycloak/keycloak/pull/531
> On 16.7.2014 12:01, Marek Posolda wrote:
>> Looks like Windows specific issue again as CI is failing as well, but
>> just on Windows:-(
>> Will try to figure out windows machine, reproduce and fix it.
>> On 15.7.2014 21:51, Bill Burke wrote:
>>> When running the ExportImportTest with the build it causes other tests
>>> to fail. I spent an hour or two trying to figure it out, I couldn't.
>>> This is off of master *BEFORE* I did my big commit. Its not a caching
>>> issue as I ran tests without the cache. It seems that users are not
>>> being deleted. There's some other issues too even if I change the
>>> usernames used in the ExportImportTest.
> keycloak-dev mailing list