Since I just couldn't stand replication not working anymore I did some
more tests. After a while I remembered that the last thing I had been
playing with was de 'database monitor' settings for statistics.
This seems to be my problem.
Three scenario's tested (in alle cases the slave ldap server has
'database monitor' turned on).
1.
---
database monitor
replogfile /usr/local/var/lib/ldap/replog
replica host=slave.nl bindmethod=simple binddn=cn=admin,o=slave,c=nl
credentials=password tls=no
---
Result: Replication does nothing. Completely dead.
2.
---
replogfile /usr/local/var/lib/ldap/replog
replica host=slave.nl bindmethod=simple binddn=cn=admin,o=slave,c=nl
credentials=password tls=no
---
Result: Whee, all works again. But no statistics ;(
3.
---
replogfile /usr/local/var/lib/ldap/replog
replica host=slave.nl bindmethod=simple binddn=cn=admin,o=slave,c=nl
credentials=password tls=no
database monitor
---
Result: Eeeeeeeek !!!, slave database corruption after a couple of
add/delete queries at the master.
Is there something specific I should do when using the "database
monitor" feature?
bye,
Chris
Openldap-2.2.17 (in /usr/local), Fedora Core2, installed alongside
default FC2 openldap 2.1.29
----- Original Message ----- From: "Christiaan den Besten"
<chris@scorpion.nl>
To: <OpenLDAP-software@OpenLDAP.org>
Sent: Friday, October 22, 2004 2:33 AM
Subject: replog not created anymore ...
Hi all!
I have upgraded our Fedora Code 2 supplied version of ldap with the
one supplied at http://www.int-evry.fr/mci/user/procacci/SRPMS/FC2/.
Installed the thing in /usr/local for comfort and had no complaints
for days!
( I actually updated to this version because I had mayor db locks
when using the default FC2 supplied version of slap. I only just read
about DB_CONFIG a couple of days ago, so maybe these explain my bdb
lockups ..... )
Yesterday I noticed that my slave ldap server did not receive any
replication updates anymore :( .... after some investigation I am
stuck ...
@master server:
---
[root@master /usr/local/var/lib/ldap]# slaptest -d 64 -v 2>&1 | grep
"repl"
line 113 (replogfile /usr/local/var/lib/ldap/replog)
line 114 (replica ***)
[root@master /usr/local/var/lib/ldap]# ls -al
total 25168
drwx------ 2 ldap ldap 4096 Oct 22 02:01 .
drwxr-xr-x 3 root root 4096 Oct 22 01:37 ..
-rw------- 1 ldap ldap 466944 Oct 22 02:01 cn.bdb
-rw------- 1 ldap ldap 16384 Oct 22 01:37 __db.001
-rw------- 1 ldap ldap 278528 Oct 22 01:37 __db.002
-rw------- 1 ldap ldap 98304 Oct 22 01:37 __db.003
-rw------- 1 ldap ldap 450560 Oct 22 01:37 __db.004
-rw------- 1 ldap ldap 16384 Oct 22 01:37 __db.005
-rw-r--r-- 1 ldap ldap 411 Oct 22 02:01 DB_CONFIG
-rw------- 1 ldap ldap 344064 Oct 22 02:09 dn2id.bdb
-rw------- 1 ldap ldap 49152 Oct 22 02:01 gidNumber.bdb
-rw------- 1 ldap ldap 1310720 Oct 22 02:09 id2entry.bdb
-rw------- 1 ldap ldap 10485705 Oct 22 01:40 log.0000000001
-rw------- 1 ldap ldap 10475423 Oct 22 02:09 log.0000000002
-rw------- 1 ldap ldap 946176 Oct 22 02:01 mail.bdb
-rw------- 1 ldap ldap 167936 Oct 22 02:05 objectClass.bdb
-rw------- 1 ldap ldap 8192 Oct 22 01:37 ou.bdb
-rw------- 1 ldap ldap 253952 Oct 22 02:01 sn.bdb
-rw------- 1 ldap ldap 454656 Oct 22 02:01 uid.bdb
-rw------- 1 ldap ldap 49152 Oct 22 02:05 uidNumber.bdb
[root@master /usr/local/var/lib/ldap]#
---
It does not create any replog file anymore, and therefore slurpd does
not send any updates to our slave anymore ( i guess ?? ) :(
Any hints what to do to debug this ?
bye,
Chris