[Date Prev][Date Next] [Chronological] [Thread] [Top]

ldapadd errors

[root@test mdb]# slapd -V
@(#) $OpenLDAP: slapd 2.4.39 (Feb  4 2014 09:07:17) $

mockbuild@buildvm-25.phx2.fedoraproject.org:/builddir/build/BUILD/openldap-2.4.39/openldap-2.4.39/servers/slapd

i am trying to add ipNetwork entries to my tree, and at first i was
getting "index generation failed" errors after several thousand entries
were added.  i removed the indexing of the cn attribute and restarted
slapd and reran the loading of the entries.  the existing objects were
moved past, and some new objects were added.  then some new errors came
out.

adding new entry "cn=64.89.32.0,c=US,ou=GeoLocation,dc=bpk2,dc=com"
ldap_add: Other (e.g., implementation specific) error (80)
	additional info: entry store failed

i am using lmdb.  i have checked filesystem space, and that is not an
issue.  the data.mdb is about 10 MB in size.

the job i am running is an ldapadd, using the rootdn ID.  this is only
because it is going against a test instance.  the job is started with:

ldapadd -h test.bpk2.com -D cn=Manager,dc=my-domain,dc=com -W -f
geolocation.ldif -S geo_errors.out -c

the data is a parsed version of the IpToCountry.csv geolocation info.
there are 137951 unique dn's to be added (with cn, ipnetworknumber,
ipnetmasknumber, l, and 2 objectclasses per dn), and i am part way
through the list.  the ldif file is about 25 MB in size.  all of the
parent objects exist, as that was a previous job to add them.  this is
just to create the networks below each of the parent objects.

where can i look for places to identify the issue?  has anyone run into
something like this before?  a quick google only shows threads for
BDB/HDB backends, but i am using LMDB.

thanks in advance,

brendan kearney