[Deploying Sakai] [Building Sakai] Possible site member list issue in 2.9.1

Thu Mar 14 10:46:36 PDT 2013

Hi Sam,

I have created https://jira.sakaiproject.org/browse/SAK-23362 for this issue.

Thanks,
Anand

________________________________
 From: Sam Ottenhoff <ottenhoff at longsight.com>
To: Anand Mehta <anand.mehta at yahoo.com> 
Cc: John Bush <john.bush at rsmart.com>; Sakai Production <production at collab.sakaiproject.org>; Sakai Dev <sakai-dev at collab.sakaiproject.org> 
Sent: Thursday, March 14, 2013 10:25 AM
Subject: Re: [Deploying Sakai] [Building Sakai] Possible site member list issue in 2.9.1

Great.  Can you open a JIRA and share your change so we can get this back into Sakai trunk?

On Thu, Mar 14, 2013 at 11:17 AM, Anand Mehta <anand.mehta at yahoo.com> wrote:

Hi John and Sam,
>
>Thanks for your help. Setting ldapEid to lowercase in JLDAPDirectoryProvider fixed the issue.
>
>
>
> 
>Thanks,
>Anand
>
>
>________________________________
> From: John Bush <john.bush at rsmart.com>
>To: Sam Ottenhoff <ottenhoff at longsight.com> 
>Cc: Anand Mehta <anand.mehta at yahoo.com>; Sakai Production <production at collab.sakaiproject.org>; Sakai Dev <sakai-dev at collab.sakaiproject.org> 
>Sent: Monday, March 11, 2013 11:41 PM
>Subject: Re: [Deploying Sakai] [Building Sakai] Possible site member list issue in 2.9.1
> 
>
>ug actually sent that without finishing what I was trying to say.
>Basically I think the lookup is case sensitive,
>usersToSearchInLDAP.get(ldapEid).  So that actually doesn't return a
>UserEdit.  So when you go to do mapUserDataOntoUserEdit() you won't
>actually be populating the edit objects.  The reason you see what you
>do in the roster tool is because the only thing in that UserEdit
>object is the eid that first got put there in the getUsers() call, it
>never gets populated because of the lowercase eid keying mixup.  You
>really have to step through this to see what is happening, its pretty
>clear when you do that.
>
>So like I said our code is a bit different, but from just looking at
>it, I think maybe the same thing is going on here.
>
>On Mon, Mar 11, 2013 at 9:36 PM, John Bush <john.bush at rsmart.com> wrote:
>> This sounds exactly like an issue we ran into when we pulled this change in:
>> https://jira.sakaiproject.org/browse/SAK-21105
>>
>> Now our ldap code is a bits of the core stuff and bits or our own
>> stuff, but it might be relevant.  It was related to when eids are
>> lowercased. My change is here:
>> http://source.sakaiproject.org/viewsvn/msub/rsmart.com/providers/trunk/jldap/src/java/com/rsmart/ldap/JLDAPDirectoryProvider.java?root=svn&pathrev=117865&view=diff&r1=117865&r2=117864&diff_format=h
>>
>> if (forceLowercaseId) {
>> ldapEid = ldapEid.toLowerCase();
>> }
>> UserEdit ue = usersToSearchInLDAP.get(ldapEid);
>> UserEdit ue = usersToSearchInLDAP.get(ldapEid);
>>
>> In the core code its something like this, and I think it might suffer
>> from the same
 problem, but I haven't tested that:
>>
>> for (LdapUserData ldapUserData : ldapUsers) {
>>     String ldapEid = ldapUserData.getEid();
>>     UserEdit ue = usersToSearchInLDAP.get(ldapEid);
>>     mapUserDataOntoUserEdit(ldapUserData, ue);
>>     usersToSearchInLDAP.remove(ldapEid);
>> }
>>
>>
>> On Mon, Mar 11, 2013 at 5:50 PM, Sam Ottenhoff <ottenhoff at longsight.com> wrote:
>>> My instinct is no, it's not relevant.  Does this happen with all sites?  All
>>> users?  With a clean cache?  Only with a dirty cache?
>>>
>>>
>>>
>>> On Mon, Mar 11, 2013 at 6:11 PM, Anand Mehta <anand.mehta at yahoo.com> wrote:
>>>>
>>>> Hi Sam,
>>>>
>>>> The command line search for jsmith returns the sAMAccountName as JSmith
>>>> (mixed case). All eids in the sakai_user_id_map table are in lowercase. Not
>>>> sure if this relevant.
>>>>
>>>>
>>>> Thanks,
>>>> Anand
>>>> ________________________________
>>>> From: Sam Ottenhoff <ottenhoff at longsight.com>
>>>> To: Anand Mehta <anand.mehta at yahoo.com>
>>>> Cc: Sakai Dev <sakai-dev at collab.sakaiproject.org>; Sakai Production
>>>> <production at collab.sakaiproject.org>
>>>> Sent: Monday, March 11, 2013 3:56 PM
>>>>
>>>> Subject: Re: [Building Sakai] Possible site member list issue in 2.9.1
>>>>
>>>> Does your "jsmith" user indeed have an empty or null sAMAccountName?  What
>>>> happens when you search for the user via a commandline search?
>>>>
>>>>
>>>> On Mon, Mar 11, 2013 at 4:41 PM, Anand Mehta <anand.mehta at yahoo.com>
>>>> wrote:
>>>>
>>>> I just replaced the actual eid with jsmith, the first and last name with
>>>> John and Smith resprectively, and the institution name with school. The
>>>> filter does not contain any other
 username. Is there a way to ignore/skip
>>>> any user with a null sAMAccountName?
>>>>
>>>>
>>>> Thanks,
>>>> Anand
>>>> ________________________________
>>>> From: Sam Ottenhoff <ottenhoff at longsight.com>
>>>> To: Anand Mehta <anand.mehta at yahoo.com>
>>>> Cc: Sakai Dev <sakai-dev at collab.sakaiproject.org>; Sakai Production
>>>> <production at collab.sakaiproject.org>
>>>> Sent: Monday, March 11, 2013 3:17 PM
>>>> Subject: Re:
 [Building Sakai] Possible site member list issue in 2.9.1
>>>>
>>>> How heavily are you editing the logs your are posting?  Specifically, this
>>>> line:
>>>>
>>>> 2013-03-07 15:05:49,727 DEBUG ajp-apr-8009-exec-13
>>>> edu.amc.sakai.user.JLDAPDirectoryProvider - searchDirectory(): [filter =
>>>> (|(sAMAccountName=jsmith))][reusing conn = false]
>>>>
>>>> Is that really the filter?  Or does the filter actually contain a bunch of
>>>> usernames?  It seems like one of your users may be returning but with a null
>>>> sAMAccountName?
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> _______________________________________________
>>> production mailing list
>>> production at collab.sakaiproject.org
>>> http://collab.sakaiproject.org/mailman/listinfo/production
>>>
>>> TO UNSUBSCRIBE: send email to production-unsubscribe at collab.sakaiproject.org
>>> with a subject of "unsubscribe"
>>
>>
>>
>> --
>> John Bush
>> 602-490-0470
>
>
>
>--
>John Bush
>602-490-0470
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/production/attachments/20130314/b3db8654/attachment.html 