Hi, all,
I set up a testing environment for the MMR(2 master server), it did work well when I tested it in a "test050" style, that was, send updated requests to one of the master server, waited for seconds, then send update requests to another master server. All updates could be properly replicated among both servers. However, when I tried to update entries in an interleave style to these servers, I mean, like, the first update request was sent to server A, and the second request was sent to server B, then A, then B, .... there was no waitting between the update requests, and all requests were differnt.
From my understanding of the MMR, I hoped all updated in server A would be
finally replicated to server B, and same to server B. But, the real test result was, I could see all updates in one of the server, but only partial updates in another server, the partial updates were exactly made by the requests directly comes to that sever, that is to say, the replication samed to be *monodirectional *in this situation.
My question, is this scenario a bug for current openLdap, or, it is just not a supported use case ?* *
I've been trying to find an answer in past mail threads, but I failed. Hope you guys can help me out from this confusion about MMR, if you need any further information, like configuration, I'll put them on, but I don't think that's a configuration problem, because the "test050" style testing works.
Thanks! Hu
Hy Hu,
I am pretty sure it is not supposed to be like that. I had a similar problem when testing MMR with more then 2 masters, which isn't your case.
Are you using 2.4.16 ? Could you please attach your configuration ?
Adrien Futschik
Adrien,
Thanks for you reply.
Yes, I'm using 2.4.16, and below my the configuration:
include /root/ldap1/schema/core.schema include /root/ldap1/schema/cosine.schema include /root/ldap1/schema/inetorgperson.schema
# Define global ACLs to disable default read access.
# Do not enable referrals until AFTER you have a working directory # service AND an understanding of referrals. #referral ldap://root.openldap.org
pidfile /usr/local/var/run/slapd2.pid argsfile /usr/local/var/run/slapd2.args
# Load dynamic backend modules: modulepath /usr/local/libexec/openldap
database bdb suffix "ou=tvg,dc=example,dc=com" rootdn "cn=admin,ou=tvg,dc=example,dc=com"
# Cleartext passwords, especially for the rootdn, should # be avoid. See slappasswd(8) and slapd.conf(5) for details. # Use of strong authentication encouraged. rootpw secret
# The database directory MUST exist prior to running slapd AND # should only be accessible by the slapd and slap tools. # Mode 700 recommended. directory /root/ldap1/db/
serverid 1 ldap://hu-dev-001:389 serverid 2 ldap://hu-dev-002:389
syncrepl rid=001 provider=ldap://hu-dev-001:389 binddn="cn=admin,ou=tvg,dc=example,dc=com" bindmethod=simple credentials=secret searchbase="ou=tvg,dc=example,dc=com" type=refreshAndPersist interval=00:00:00:10 retry="60 +" sizelimit=unlimited timeout=1
syncrepl rid=002 provider=ldap://hu-dev-002:389 binddn="cn=admin,ou=tvg,dc=example,dc=com" bindmethod=simple credentials=secret searchbase="ou=tvg,dc=example,dc=com" type=refreshAndPersist interval=00:00:00:10 retry="60 +" sizelimit=unlimited timeout=1
mirrormode true
overlay syncprov syncprov-checkpoint 10 1 syncprov-sessionlog 10000
# Indices to maintain index objectClass eq index entryCSN,entryUUID eq
On Mon, May 18, 2009 at 2:27 PM, Adrien Futschik < adrien.futschik@atosorigin.com> wrote:
Hy Hu,
I am pretty sure it is not supposed to be like that. I had a similar problem when testing MMR with more then 2 masters, which isn't your case.
Are you using 2.4.16 ? Could you please attach your configuration ?
Adrien Futschik
openldap-technical@openldap.org