grouper-dev - Draft Minutes: Grouper Call 6-Feb-2008
Subject: Grouper Developers Forum
List archive
- From: Steve Olshansky <>
- To: Grouper Dev <>
- Subject: Draft Minutes: Grouper Call 6-Feb-2008
- Date: Fri, 08 Feb 2008 13:24:05 -0700
Draft Minutes: Grouper Call 6-Feb-2008
*Attending*
Tom Barton, U. Chicago (chair)
Shilen Patel, Duke U.
Kathryn Huxtable, independent
Gary Brown, U. Bristol
RL “Bob” Morgan, U. Washington
Joy Veronneau, Cornell U.
James Cramton, Brown U.
Chris Hyzer, U. Pennsylvania
Bill Kasenchar, U. Pennsylvania
Dave Donnelly, Stanford U.
Mike Olive, Stanford U.
Tom Zeller, Indiana U.
Steve Olshansky, Internet2 (scribe)
[AI] (Chris and Shilen) will investigate Hibernate 3 and report back to the group (via the list) with their opinion about its suitability for Grouper. http://www.hibernate.org/
[AI] (All) Review JIRA issues and discuss via the list as inclined.
*Discussion*
*Update from Grouper-UI subgroup*
They have been working methods for incorporating tooltips and/or infodots as helpful prompts for users. Chris provided some JSP libraries, checked into SVN. We need to create the actual text which will appear in the tooltips, for various components of the UI.
For more information, see this topic in the wiki:
https://wiki.internet2.edu/confluence/display/GrouperWG/UI+Analysis
As Grouper develops, the UI becomes increasingly important. The overarching goal is to make Grouper as usable and intuitive as possible, especially for non-technical users. Soliciting broader input from the community will be helpful ongoing, as well as obtaining specific feedback from potential users as the UI develops.
Terminology may be something that changes, in order to make them more usable to the users. E.g. “extension” when a group is created, perhaps “system name” would be better. It was observed that some users have a hard time understanding how the system works, and how changes are reflected in the infrastructure.
Overall, there is consensus that improving the UI is obviously a good thing, and we should not be concerned about staying with terminology used to date.
*Grouper v1.2.1 Performance Update*
James reported that for nightly provisioning, followed by routine changes, Brown is seeing ~60% reduction in runtime compared to v1.2.0.
*Review functionality planned for v1.3.0*
https://bugs.internet2.edu/jira/browse/GRP
Current plan is for Grouper v1.3.0 to be released in late April. In addition to the v1.3.0 issues in JIRA, we also discussed some of the "unscheduled" JIRA issues in the context of whether they should be included in v1.3.0. Changes will be reflected in JIRA as appropriate.
GRP-7: Simplify UI for browsing
This had to do primarily with performance issues for certain users having a large number of group memberships.
GRP-53: Privilege checking in GrouperQuery
This had to do primarily with improving performance on privilege checking when doing large queries. This was also affecting Ldappc testing. Rewriting group search procedures should help this.
GRP-70: Sub-optimal index for grouper_composites
There is currently an index across 2 fields, we will look at indexing across each field individually. Shilen will work with ChrisH on this.
GRP-86: Improve (and make consistent) the checkout and build process of grouper, grouper-ui, and grouper-ws
Chris is willing to work on improving the build process. Maven was suggested as a build system worth looking into, although it would be a significant shift and thus now is likely not the time. We will move this to “unassigned” and revisit it in the future.
http://maven.apache.org/
The bigger picture is how Grouper fits into I2MI (Internet2 Middleware Initiative) code integration issues. Thus we need to ensure that our approach is consistent across the other projects (Signet and Shibboleth).
-- Currently Unassigned Issues, moving to v1.3.0 --
GRP-46: Support deleting group membership for subjects who are no longer in the directory
This is the unresolved subject deletion utility. TomZ claimed this.
GRP-47: Do not permit creation of multiple identical membership paths
This is a bug we have discussed in the past, and the fix ought to be in v1.3.0. It was observed that this one is very similar to:
GRP-79: HibernateMembershipDAO GrouperDAOException: query did not return a unique result: 2
We will combine or link these. Kathryn offered to work on this.
GRP-13: Poor integration with Ldappc: attribute-matching-queries generates full table scan
Kathryn offered to work on this.
GRP-9: Group attribute queries generate full table scans
GRP-13 depends on this, and Kathryn offered to look into this one as well.
GRP-14: Poor integration with Ldappc: Null DTO error message in Ldappc run causes run to fail
Kathryn offered to work on this.
GRP-80: Performance of HibernateMembershipDAO.exists()
Shilen noticed this in Duke’s upgrade from 1.2.0 to 1.2.1. Using the existing indexes, some queries took 300-400 ms, but adding an additional index took it down to 5 ms. This adds up when doing this operation many times in normal use. Shilen will review this and report back to the group, and it may roll up other JIRA issues.
The next call will be Wed 20-Feb Noon EST.
--
Steve Olshansky
Internet2 Middleware & Security Flywheel
http://middleware.internet2.edu/
http://security.internet2.edu/
- ldappc update, Kathryn Huxtable, 02/06/2008
- Re: ldappc update, Kathryn Huxtable, 02/06/2008
- Re: [grouper-dev] Re: ldappc update, Tom Zeller, 02/08/2008
- Draft Minutes: Grouper Call 6-Feb-2008, Steve Olshansky, 02/08/2008
- Re: [grouper-dev] Re: ldappc update, Tom Zeller, 02/08/2008
- Re: ldappc update, Kathryn Huxtable, 02/06/2008
Archive powered by MHonArc 2.6.16.