[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: 2.3.18 syncrepl problem



> This was "dn".
> Now it starts, but replication doesn't work.
> Looks like I didn't created databases correctly.
> I used slapadd -w -l lfid
> on both hosts.
> 
> Now I see in log only
> put_filter: "(objectClass=belkamAccount)"
> put_filter: simple
> put_simple_filter: "objectClass=belkamAccount"
> ldap_send_initial_request
> ldap_send_server_request
> ber_scanf fmt ({it) ber:
> ber_scanf fmt ({) ber:
> ber_flush: 221 bytes to sd 15
> =>do_syncrep2
> ldap_result ld 0x82e580 msgid -1
> ldap_chkResponseList ld 0x82e580 msgid -1 all 0
> ldap_chkResponseList returns ld 0x82e580 NULL

I dont think you should do slapadd -w on both hosts.  You should do slapadd -w on the master (so that contextCSN's are created) and then do a slapcat to get an ldif with the proper csn values.  Then slapadd that file to the slave WITHOUT the -w.  Or, if you are feeling adventurous, you can completely clear the slave's db and start it up to allow syncrepl to download the entire dir (doesn't work too well if you have multiple slaves trying to pull a non-trivial directory).

-- DK