[389-users] Unable to establish replication with STARTTLS

I have two hosts with 389-Directory/1.4.4.17 B2021.280.1354 on CentOS Stream release 8 (4.18.0-448.el8.x86_64)

On a.state.ak.us, there is one instance defined (call this instance #1)

On b.state.ak.us, there are two instances defined (call them #2 and #3)

Instances #1 and #3 have GlobalSign certificates installed. Instance #2 currently has a Let's Encrypt certificate installed. All instances also have root and intermediate certs in their databases for GlobalSign, which are marked with Trust Flags "CT,,".

I can define instance #2 as a supplier, and define a replication agreement which populates #3. This works with both LDAPS and STARTTLS.

If I, instead, try to define the same replication agreement on instance #1, it fails with:

slapi_ldap_bind - Error: could not send startTLS request: error -11 (Connect error)

NSMMReplicationPlugin - bind_and_check_pwp - agmt="cn=DS11-1to3" (b:389) - Replication bind with SIMPLE auth failed: LDAP error -11 (Connect error) (error:1416F086:SSL routines:tls_process_server_certificate:certificate verify failed (unable to get issuer certificate))

slapi_ldap_bind - Error: could not send startTLS request: error -11 (Connect error)

I am unable to figure out how instances #1 and #2 differ.

Instance #1 has long-established supplier-agreements (using both LDAPS and STARTTLS) with other instances of 389-Directory. So I know instance #1 can function correctly as a supplier. Instance #3 demonstrates it can be a consumer when supplied by instance #2. I can perform LDAPS and STARTTLS queries from a.state.ak.us to instance #3, so I know it is listening on the network and not blocked by a host-based firewall.

Any suggestions of where to look, or config-attributes to check, would be appreciated.

--   --  Do things because you should, not just because you can.     John Thurston    907-465-8591  John.Thurston@alaska.gov  Department of Administration  State of Alaska