Hallo all,
in this week, the syncrepl in RefreshOnly mode has delete over 90% of the ldap entries at the consumer. It is the secont time in last month, it make this mistake. The other time it work ok. After I have remove the database and slapd restart, all entries are back. The consumer is my mailserver, so no user can receive mail in the time his entries are removed.
I work with debian, at the consumer with lenny and slapd 2.4.11-1 and at the Provider with entch and slapd 2.3.30-5+etch2 . I use gosa 2.5.16 to adminstrate ldap.
Here are the log entries of the consumer:
Apr 21 08:48:45 dcs slapd[7236]: do_syncrep2: rid=101 LDAP_RES_INTERMEDIATE - SYNC_ID_SET Apr 21 08:48:45 dcs slapd[7236]: do_syncrep2: cookie=csn=20090421064823Z#000003#00#000000,rid=101 Apr 21 08:48:45 dcs slapd[7236]: slap_queue_csn: queing 0xae7069c0 20090421064823Z#000003#00#000000 Apr 21 08:48:45 dcs slapd[7236]: slap_graduate_commit_csn: removing 0xae72dc78 20090421064823Z#000003#00#000000 Apr 21 08:48:45 dcs slapd[7236]: syncrepl_del_nonpresent: rid=101 be_delete uid=nutzera,ou=people,ou=arbgrp,ou=abt,ou=my,o=domain,c=de (0) : : slapd[7236]: syncrepl_entry: rid=101 LDAP_RES_SEARCH_ENTRY(LDAP_SYNC_ADD) Apr 21 08:48:45 dcs slapd[7236]: syncrepl_entry: rid=101 inserted UUID 33fdb48e-c28c-102d-9d8a-c5d1aca5fa09 Apr 21 08:48:45 dcs slapd[7236]: syncrepl_entry: rid=101 be_search (0) Apr 21 08:48:45 dcs slapd[7236]: syncrepl_entry: rid=101 cn=b9d1d7db1b3bc592e9b7b57acc40041e,ou=gosa,ou=configs,ou=systems,ou=tz,o=fal,c=de Apr 21 08:48:45 dcs slapd[7236]: syncrepl_entry: rid=101 be_add (0) Apr 21 08:48:45 dcs slapd[7236]: do_syncrep2: rid=101 LDAP_RES_SEARCH_RESULT Apr 21 08:48:45 dcs slapd[7236]: do_syncrep2: cookie=csn=20090421064845Z#000000#00#000000,rid=101 Apr 21 08:48:45 dcs slapd[7236]: slap_queue_csn: queing 0xae715640 20090421064845Z#000000#00#000000 Apr 21 08:48:45 dcs slapd[7236]: slap_graduate_commit_csn: removing 0xae702e40 20090421064845Z#000000#00#000000 Apr 21 08:58:45 dcs slapd[7236]: do_syncrep2: rid=101 LDAP_RES_INTERMEDIATE - SYNC_ID_SET Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 LDAP_RES_SEARCH_ENTRY(LDAP_SYNC_ADD) Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 inserted UUID 8b82805e-0867-102b-8021-e9137c6ded0b Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 be_search (0) Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 cn=mygrp,ou=groups,ou=my,o=domain,c=de Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 be_modify (0) Apr 21 08:58:45 dcs slapd[7236]: syncrepl_entry: rid=101 Apr 21 08:58:45 dcs slapd[7236]: do_syncrep2: rid=101 LDAP_RES_SEARCH_RESULT Apr 21 08:58:45 dcs slapd[7236]: do_syncrep2: cookie=csn=20090421065817Z#000003#00#000000,rid=101 Apr 21 08:58:45 dcs slapd[7236]: nonpresent_callback: rid=101 not UUID 8a5d17de-0867-102b-9fad-e9137c6ded0b, dn c=de Apr 21 08:58:45 dcs slapd[7236]: nonpresent_callback: rid=101 not UUID 8a5e5f86-0867-102b-9fb0-e9137c6ded0b, dn o=domain,c=de Apr 21 08:58:45 dcs slapd[7236]: nonpresent_callback: rid=101 not UUID 8a5ed024-0867-102b-9fb2-e9137c6ded0b, dn ou=my,o=domain,c=de : (90% of ldapr entries) pr 21 08:58:45 dcs slapd[7236]: slap_queue_csn: queing 0xae72c058 20090421065817Z#000003#00#000000 Apr 21 08:58:45 dcs slapd[7236]: slap_graduate_commit_csn: removing 0xae72cf50 20090421065817Z#000003#00#000000 Apr 21 08:58:45 dcs slapd[7236]: syncrepl_del_nonpresent: rid=101 be_delete cn=b9d1d7db1b3bc592e9b7b57acc40041e,ou=gosa,ou=configs,ou=systems,ou=my,o=domain,c=de (0) : (90% of entries) Apr 21 08:58:46 dcs slapd[7236]: syncrepl_del_nonpresent: rid=101 be_delete ou=my,o=domain,c=de (66) Apr 21 08:58:46 dcs slapd[7236]: syncrepl_del_nonpresent: rid=101 be_delete o=domain,c=de (66) Apr 21 08:58:46 dcs slapd[7236]: syncrepl_del_nonpresent: rid=101 be_delete c=de (66) Apr 21 08:58:46 dcs slapd[7236]: slap_queue_csn: queing 0xae72c058 20090421065817Z#000003#00#000000 Apr 21 08:58:46 dcs slapd[7236]: slap_graduate_commit_csn: removing 0xb4d03ff0 20090421065817Z#000003#00#000000
I now, that the logs no show errors, but why can the consumer say no entry found and delete. My colleague has delete 3 user, this change was made right and the al the other entriies are delete. How can I prevent this in the future?
The configurations: Provider: slapd 2.3.30-5+etch2 Config: .... moduleload back_bdb moduleload syncprov : : # Fuer syncreply index entryCSN,entryUUID eq : overlay syncprov syncprov-checkpoint 100 10 syncprov-sessionlog 100
Consumer: slapd 2.4.11-1 (lenny) Config: ..... index entryCSN,entryUUID eq : : syncrepl rid=101 provider=ldaps://ldap.my.local:636 type=refreshOnly searchbase="c=de" retry="60 10 300 3 600 +" scope=sub schemachecking=off bindmethod=simple binddn="cn=replicator,ou=my,o=domai,c=de" credentials=ReSumpti0n
regards Monika