Begin forwarded message:

From: Paul Turgyan <pturgyan@umich.edu>
Date: August 30, 2007 2:43:19 PM GMT-04:00
To: openldap-software@OpenLDAP.org
Subject: Provide slapd Aborts on failed assertion


I have a consumer slapd in refreshandpersist mode and a provider slapd
that aborts on the first search operation issued by the consumer.
I've tracked it down to the cachesize option.  If I set the cachesize option to
any value,  the provider fails an assertion in pthread_mutex_lock.


# database settings
database        bdb
suffix          "dc=umich,dc=edu"
rootdn          "cn=root,dc=umich,dc=edu"
rootpw          *****

overlay         syncprov
#syncprov-checkpoint 100 10
#syncprov-sessionlog 100

directory       /var/slapd/tmp/db/db.save
cachesize       1000
idlcachesize    25000
checkpoint      1024 5
lastmod         on

When I move the overlay statements  later in the config file,  the provider 
stays up,  the the consumer sucks down the entire database, and everything
seem to work.
.
This is the new database section of my slapd.conf file for the provider.

database        bdb
suffix          "dc=umich,dc=edu"
rootdn          "cn=root,dc=umich,dc=edu"
rootpw          *****

directory       /var/slapd/tmp/db/db.save
cachesize       100000
idlcachesize    25000
checkpoint      1024 5
lastmod         on
#
overlay         syncprov