Hi. We have a one master and 8 slaves. From time to time replication has broken and data has stopped to update on slaves. In logs i see next:
syncrepl_null_callback : error code 0x50 syncrepl_entry: rid=000 be_modify cn=union,ou=roles,dc=staff,dc=com (80) syncrepl_entry: rid=000 be_modify failed (80) do_syncrepl: rid=000 rc 80 retrying do_syncrep1: rid=000 starting refresh (sending cookie=rid=000,csn=20211213150433.004754Z#000000#000#000000) do_syncrep2: rid=000 LDAP_RES_INTERMEDIATE - SYNC_ID_SET syncrepl_message_to_entry: rid=000 DN: cn=union,ou=roles,dc=staff,dc=com, UUID: 67c22522-0749-1039-933a-fdc6b5a9b3b7 syncrepl_entry: rid=000 LDAP_RES_SEARCH_ENTRY(LDAP_SYNC_ADD) csn=(none) tid 4
"error code 0x50" means other ldap error. How to determine, what exactly wrong?
There is no dependence with slaves or versions: - master has 2.4.50 - slaves have different versions: 2.4.50, 2.4.55, 2.4.57.
All of slaves can breake. We use next configuration:
syncrepl rid=000 provider=ldaps://ldap-master.domain.com type=refreshAndPersist retry="5 5 300 +" searchbase="dc=staff,dc=com" attrs="*,+" bindmethod=simple binddn="cn=slapd-slave,ou=services,dc=staff,dc=com"
We can't try newer version, because use Oracle solaris and there are only 2.4.X versions.
Should we create a bug or it was fixed in newer versions?