On Feb 17, 2014, at 5:45 AM, Josef Cacek <jcacek(a)redhat.com&gt; wrote: Hi *, I'm writing just to let you know, I have received no comment wrt PicketLink STSClientPool test plan. I have finished the test development (git commit [1]) and some of the tests with PicketLink 2.1.10 fail (I was trying the upstream PL 2.1.10 with EAP 5.2.0). The issues will be blockers for EAP 5.3 from my PoV. [1] https://github.com/jboss-security-qe/picketlink-integration-tests/commit/...

Best regards, -- Josef ----- Original Message ----- > From: "Josef Cacek" <jcacek(a)redhat.com&gt; > To: "Anil Saldhana" <asaldhan(a)redhat.com&gt;, "Peter Skopek" <pskopek(a)redhat.com&gt; > Cc: "Jitka Kozana" <jkudrnac(a)redhat.com&gt;, security-dev(a)lists.jboss.org, "Mark Yarborough" <myarboro(a)redhat.com&gt; > Sent: Wednesday, February 12, 2014 1:48:13 PM > Subject: [security-dev] PicketLink 2.1.10 and 2.5.3.Beta2 issues > > Hi Anil & Peter, > > I have several issues with PicketLink 2.1.10. I'm preparing STS client pool > tests for EAP 5.3 and I would like to voice the following concerns here: > > 1) missing release numbers in the JIRA > (https://issues.jboss.org/browse/PLINK2) > 2) There is a regression (NPE) in JBWSTokenIssuingLoginModule - > https://issues.jboss.org/browse/PLINK2-127 > 2 test cases from the picketlink-integration-tests hit this. Did your run > of the testsuite before the release passed without any issues? > 3) missing documentation of the new STSClient pool feature. Even the JavaDoc > for methods is missing in STSClientPool and STSClientFactory classes. This > documentation will be needed by our documentation team. > 4) I would like to bring to your attention the following features of the > implementation of the STS client pool: > - STSClientFactory doesn't provide access to the pool used - user doesn't > know he should return the STSClient instance to a pool. Should this be > documented, so leaks can be avoided? > - scaling is not possible - singleton is used in class STSClientPool > - once the pool is created it is not possible to resize it during runtime > (or create another one with different size) > - I don't see a possibility how to clean-up the pool - e.g. when a thread > which uses a client from pool dies > > 5) All the points above are also valid for PicketLink 2.5.3.Beta2, which is > included in EAP 6.3.0.DR0 - > https://bugzilla.redhat.com/show_bug.cgi?id=1064331 > > Anil, Peter, can you please comment on these? > > I've created an EAP 5.3 test plan for STS pool here - > https://issues.jboss.org/browse/JBQA-8620 > Don't hesitate to comment there, if you see invalid entries or if you are > missing some more points. > > Thank you in advance, > > -- josef > > _______________________________________________ > security-dev mailing list > security-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/security-dev _______________________________________________ security-dev mailing list security-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/security-dev