> We have a weird legacy DIT which is going to be migrated to a new more > organised structure by an ongoing project. > > In the meantime, we are deploying provisioning tools which we would prefer > not > to modify at the time of the migration. > > So, back-relay seems to be the obvious solution (provision to the final > structure, rewrite with back-relay into the old structure). > > So, I have added a new database on one of our dev servers: > > > database relay > suffix "cn=webmail,cn=mail,ou=isp" > subordinate > relay cn=webmail,ou=mail,dc=isp massage > rootdn <rootdn> > > There is a bdb database that holds cn=mail,ou=isp (the legacy basedn). > > That works great, searches on cn=webmail,cn=mail,ou=isp returns an entry I > added under cn=webmail,ou=mail,dc=isp ... except that the first search on > the > cn=mail,ou=isp suffix kills the relay. Any other search which should find > anything under cn=webmail,cn=mail,ou=isp returns error 32. > > > I've also tried replacing the relay database with a configurations > using "database meta" and "database ldap" with rwm-suffixmassage, and they > do > the same thing. > > It seems I must be missing something here ...
I suggest you try to design a solution where you don't need the relay backend to be subordinate to anything. I suspect some strange interaction between slapo-glue and back-relay. If you can come up with an essential configuration + dataset that reproduces the problem, I suggest you file an ITS; tis could help back-relay stabilization. p. Ing. Pierangelo Masarati Responsabile Open Solution OpenLDAP Core Team SysNet s.n.c. Via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ------------------------------------------ Office: +39.02.23998309 Mobile: +39.333.4963172 Email: [EMAIL PROTECTED] ------------------------------------------
