Skip to Content.
Sympa Menu

grouper-users - RE: [grouper-users] Grouper LDAP loader - large result sets

Subject: Grouper Users - Open Discussion List

List archive

RE: [grouper-users] Grouper LDAP loader - large result sets


Chronological Thread 
  • From: Chris Hyzer <>
  • To: Jeff McCullough <>
  • Cc: Tom Zeller <>, "" <>, CalNet Administration <>
  • Subject: RE: [grouper-users] Grouper LDAP loader - large result sets
  • Date: Tue, 3 Apr 2012 04:43:09 +0000
  • Accept-language: en-US

TomZ feel free to chime in...

I did a loader job from our ldap which returned 30k subjects. I didn't have
the source configured, but I saw that many results come back from ldap.

Which type of job is this? I tried LDAP_SIMPLE

I set this in the log4j.properties:

log4j.logger.edu.internet2.middleware.grouper.ldap = DEBUG

I saw this in the logs:

2012-04-03 00:35:13,908: [DefaultQuartzScheduler_Worker-1] DEBUG
LdapSession$1.callback(274) - - Found 30650 results for serverId:
personLdap, searchDn: ou=whatever, filter: '(cn=someFilter:whatever)',
returning attribute: hasMember, some results: ArrayList size: 30650: [0]:
whatever

Do you see that or is it 500?

It is in the LdapSession class, doing this:

searchResultIterator = ldap.search(searchDn,
searchFilterObject, searchControls);

I would think the problem is in vt-ldap (or one of its configs), or the ldap
server setting. TomZ do you have some help here? :)

Thanks,
chris


-----Original Message-----
From: Jeff McCullough
[mailto:]

Sent: Monday, April 02, 2012 3:13 PM
To: Chris Hyzer
Cc: Tom Zeller;
;
CalNet Administration
Subject: Re: [grouper-users] Grouper LDAP loader - large result sets

I should add that the subject ids in the group are the first 500 (of the
total 76119) records returned using ldapsearch. -Jeff

On Apr 2, 2012, at 12:02 PM, Jeff McCullough wrote:

> Hi Chris,
>
> I'm using gsh. The possible result set is 76119, and the group created is
> of size 500.
>
> Jeff
>
> ----------------
> gsh script from your examples
> -----------------
> grouperSession = GrouperSession.startRootSession();
> group = new
> GroupSave(grouperSession).assignName("someStem:myLdapGroup").assignCreateParentStemsIfNotExist(true).save();
> attributeAssign =
> group.getAttributeDelegate().assignAttribute(LoaderLdapUtils.grouperLoaderLdapAttributeDefName()).getAttributeAssign();
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapTypeName(),
> "LDAP_SIMPLE");
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapFilterName(),
> "(&(objectclass=person)(uid=*))");
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapQuartzCronName(),
> "0 * * * * ?");
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapServerIdName(),
> "ucbLdap");
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapSubjectAttributeName(),
> "uid");
> attributeAssign.getAttributeValueDelegate().assignValue(LoaderLdapUtils.grouperLoaderLdapSubjectIdTypeName(),
> "subjectIdentifier");
> group = GroupFinder.findByName(grouperSession, "someStem:myLdapGroup");
> loaderRunOneJob(group);
>
>
> On Mar 31, 2012, at 8:06 PM, Chris Hyzer wrote:
>
>> What type of ldap loader job is it? How big is the resultset supposed to
>> be and how big is the group which it is not the right size?
>>
>> thanks,
>> chris
>> ________________________________________
>> From:
>>
>>
>> []
>> on behalf of Tom Zeller
>> []
>> Sent: Saturday, March 31, 2012 11:52 AM
>> To: Jeff McCullough
>> Cc:
>> ;
>> CalNet Administration
>> Subject: Re: [grouper-users] Grouper LDAP loader - large result sets
>>
>> Are you receiving size or time limit error messages ? What makes you think
>> jndi is trying to page ?
>>
>> I will have to look, but I thought the grouper ldap loader wrapped
>> vt-ldap, which itself wraps jndi, so there might be knobs we can tweak.
>>
>> TomZ
>>
>> On Mar 31, 2012, at 1:38 AM, Jeff McCullough
>> <>
>> wrote:
>>
>>>
>>> I'm trying out grouper's LDAP loader. I'm finding that when I'm doing a
>>> large result query that I'm only getting a nibble of the total possible
>>> results back. Our LDAP server, Oracle DSEE, doesn't support paging,
>>> though it is a fairly common LDAP server feature for dealing with large
>>> data result sets. It is looking like JDNI is trying to page despite both
>>> the bind I'm using and the server can return a very large data result set
>>> using ldapsearch. Does anyone have experience with this, and have
>>> suggestions about how to let JDNI/Grouper retrieve the entire result set
>>> in one pass?
>>>
>>> Thanks,
>>> Jeff
>




Archive powered by MHonArc 2.6.16.

Top of Page