Re: Are there threading issues with slapd?

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

You describe problems similar to those in ITS#1349 and ITS#1359.
There is a fix in OPENLDAP_REL_ENG_2 which hopefully addresses
these issues, please test.
        Kurt

At 08:32 AM 2001-10-08, Natarajan, Bartee wrote:
>Hi,
>
>We have a 'slapd' server running on windows 2000 with the ldbm backend
>database. We seem to be 
>encountering problems when we attempt to contact the server using more than
>one client. Everything 
>works great if we ensure that only one client communicates with 'slapd'. 
>
>For instance, if we attempt to modify the directory (using JNDI) with one
>client while 
>browsing with another,  OR
>if we attempt to modify two different directory trees (using JNDI) using 
>two clients, the JNDI client(s) abort with the following errors..
>
>Request: 6576cancelled
>
>Request: 19cancelled
>
>Since then, these two directories seem corrupted. Requests on this directory
>even with a single 
>client returns errors like the following:
>
>: [LDAP: error code 80 - cache add failed]
>LDAP: error code 80 - DN index generation failed
>
>I can't imagine that the 'slapd' server has multi-threading issues.. But I
>haven't dug deep
>to see what might be the cause of this.. I know I haven't given too much
>info. here, but I
>was wondering if there are multi-threading issues with OpenLDAP on windows
>2000. 
>
>I hope to hear otherwise and wish that this was a problem with our code or
>configuration!   :)
>
>Thanks in advance for any info..
>
>-bartee