Skip to end of metadata
Go to start of metadata

Agenda

This week's agenda:

  1. Lingering KFS/Rice Integration Issues

Notes

chris
chad
eric
matt
peter
jonathan
aaron

Eric:
The issues in 2.1.2 that are ranked in the agile view are the ones that we're looking at. they've been ranked
and some are blockers, etc.

Eric: Should we review all the items or has that already happened in other forums?

Peter: I just resolved an issue that's not on that list that's worth talking about. KULRICE-7605. Seems to be a build
or deploy issue. Is an issue with role qualifiers, memberships, namespace. Both opening the role or a person with a role
is throwing an error in can't find EOB in namespace BO. Basically I'm concerned about the amount of time put into this in
convincing myself that it's an issue in KFS STFG.

Jonathan : should the objects on KIM attributes be...

Eric: Seems we've been having a bit of EBO issues in Rice 2.0, we've had a bunch of JIRAs at the release and now this one. I wonder
if there's something in the EBO framework that we should look at.

Jonathan : It's probably been there for quite a while, but with 2.0 we're using a lot more EBOs with the services and probably
exposed these issues.

Eric : Looking at the JIRA i don't see a full stack trace. So it's resolved, but it's still an issue?

Peter : I'm not sure where to go at this point. It only works when I build with the KFS rice web war file.

Eric : Have you tried to compare the contents of the two war file

Peter : I'm assuming there are differences, but haven't done a compare yet. I still think there might be something higher up
in the stack.

Eric : Any other ideas of what we could look at from the collective? Namespace is a lookup which is on Rice trying to resolve the URL
on the KFS server. For attribute definition, it should be able to resolve the URL though. Is it just the fact that the mapping isn't
setup on the module service?

Jonathan : The mapping shouldn't be wrong and should work.

Eric : I think we (Peter and I) look at this one and see what we can figure out. We could add logging to the code in the case where
it's not working and see what's happening on the KFS side. And this is blocking the testing of the KIM screens...

Jonathan : Let me check the build directory and ensure that nothing is in there from old builds. I'm gonna double check class paths and
such to make sure there's no old code coming through.

Eric: OK, lets look at the other blockers that are out in the list then.

Chris : We're seeing issues with some roles that have multiple thousand folks in it. We would like to have an API to have role qualifier criteria as part of the
filter to determine what members to retrieve instead of all of them.

Eric : There is some code inside of principal has role that allows you to find the qualifiers for exact
match to do a direct database query instead of getting all. should be faster, but still has to go through delegations. Really awesome if your calls are remote. We
ran into this at IU with some derived roles and fixed some of these performance issues up. There were also default cache size updates that have been contributed back.
I would say, Chris, if you guys try to do your test with 2.1.1, you should see a bump in performance.

Chris : OK, hopefully that fixes some things up. The large membership roles, in our implementation has been rough.

Eric : There are some other performance issues that are being worked on, but probably wont be part of 2.1.1, but will be contributed back later after we've verified here at IU.

Chris: I'll also make a comment on KULRICE-7689 about the work we'll be doing to get 2.1.1 and test things out.

Attendees

  • Peter (Rice)
  • Jerry (Rice)
  • Matt (Rice)
  • Gayathri (KC)
  • Will (UMD/KS?)
  • Kristina (Iowa St/KPME)
  • Chad (Cornell)
  • Bryan (Cornell/KFS)
  • Erik M. (Rice)
key summary assignee fixVersion priority status created updated

com.atlassian.sal.api.net.ResponseException: javax.net.ssl.SSLHandshakeException: Received fatal alert: handshake_failure

View these issues in JIRA