[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
SyncRepl is running but not syncing??
- To: <openldap-software@OpenLDAP.org>
- Subject: SyncRepl is running but not syncing??
- From: "Casper Angelo" <c.angelo@ofir.com>
- Date: Thu, 26 May 2005 13:46:49 +0200
- Content-class: urn:content-classes:message
- Thread-index: AcVh6JddaCS8a32ESe+Wy7eTbJOf3Q==
- Thread-topic: SyncRepl is running but not syncing??
I need two slaves to my master directory. I used to run replication
between them, but having 400.000+ entries and lots of changes constanly
there has been slips, where modifications simply gets lost somewhere in
the replication.
So I would like to use SyncRepl instead, to insure a perfect copy on my
slave ldaps.
The three machines are identical:
Debian
ReiserFS
OpenLDAP 2.2.23-5
Only difference is master running hdb ldapbackend, and the slaves
running ldbm.
I setup the following in the master:
===== MASTER ==========================
Sessionlog 100 500
=======================================
And the following in both slaves
===== SLAVES ==========================
syncrepl rid=102 (103 in the other slave)
provider=ldap://openldap01
type=refreshAndPersist
interval=00:00:10:00
searchbase="o=Ofir"
filter="(objectClass=*)"
scope=sub
schemachecking=off
updatedn= <dn>
bindmethod=simple
binddn= <dn>
credentials= <passwd>
=======================================
In both master and slaves, all traces of replication setup has been
removed.
When all three servers are started, the two slaves connect to the master
as they are supposed to:
=======================================
openldap01:~# lsof -i tcp:ldap
COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME
slapd 13315 root 6u IPv4 757901 TCP *:ldap (LISTEN)
slapd 13315 root 9u IPv4 758169 TCP
openldap01:ldap->openldap02:32806 (ESTABLISHED)
slapd 13315 root 10u IPv4 758632 TCP
openldap01:ldap->openldap03:32793 (ESTABLISHED)
=======================================
In the slaves directory the update cokies are created:
=======================================
openldap02:~# slapcat -k
[snip root entries]
dn: cn=syncrepl102,o=ofir (dn: cn=syncrepl103,o=ofir in the other
slave)
objectClass: top
objectClass: subentry
objectClass: syncConsumerSubentry
cn: syncrepl102
subtreeSpecification: {}
syncreplCookie: csn=20050526104319Z#000001#00#000000
entryUUID: 4cc448ae-6221-1029-8fa9-fee3d57bfab2
creatorsName: cn=admin,o=Ofir
createTimestamp: 20050526110148Z
entryCSN: 20050526110148Z#000001#00#000000
modifiersName: cn=admin,o=Ofir
modifyTimestamp: 20050526110148Z
=======================================
So all in all, everything seems to run like it's supposed to. And when
the slaves start, they recieve the root entries, thats been slapadd'ed
to the master, but a large group of 'ou' entries thats been ldapadd'ed
to the server is not synched. When I add new entries to the master ldap,
nothing happens on the slaves.
Running the slave ldap in debug mode, shows me no activity at all when I
add new entries to the master.
Any good ideas anyone?
Casper Angelo
OFiR SysAdmin