Kurt D. Zeilenga wrote:
I think we likely should have:
SLAP_SHADOW(bd) - is a shadow database SLAP_SYNC_SHADOW(bd) - is a syncrepl shadow database SLAP_SLURP_SHADOW(bd) - is a slurped shadow database
and: be_slurp_update(op) TRUE if SLAP_SLURP_SHADOW and is update DN be_sync_update(op) TRUE if SLAP_SYNC_SHADOW and is internal sync operation.
and (if useful): be_shadow_update(op) TRUE if be_slurp_update() or be_sync_update()
I[ve just committed a first round of fixes in this direction. All tests work as expected (significantly those related to replication, of course). I also noticed some possibly inconsistent testing for replication/shadowing, but I'm not sure, so I'll try to indicate the places and what are my suggested fixes, but I don't want to commit anything I'm unsure about.
Ando.
SysNet - via Dossi,8 27100 Pavia Tel: +390382573859 Fax: +390382476497