>>> "guilla...@cheramy.name" <guilla...@cheramy.name> schrieb am 02.04.2013 um 16:05 in Nachricht <515ae59b.8060...@cheramy.name>: > Le 20/03/2013 00:51, Andreas Kurz a écrit : > > On 2013-03-18 16:53, guilla...@cheramy.name wrote: > >> Hello, > >> > >> I'll try for a week to make a cluster with 2 servers for slapd HA. > >> ldap01 is the master slapd server, ldap02 is a replica server. All is ok > >> with Debian and /etc/ldap/slapd.conf configuration. > >> > >> So now I wants if ldap01 fail that ldap02 became master for continue to > >> add data in ldap database. And if ldap01 failback I wants it became > >> slave of ldap02 for resync data. > >> > >> So I'll try this configuration with pacemaker multi-state with this > >> configuration : > >> > >> # crm configure show > >> node ldap01 \ > >> attributes standby="off" > >> node ldap02 \ > >> attributes standby="off" > >> primitive LDAP-SERVER ocf:custom:slapd-ms \ > >> op monitor interval="30s" > >> primitive VIP-1 ocf:heartbeat:IPaddr2 \ > >> params ip="10.0.2.30" broadcast="10.0.2.255" nic="eth0" > >> cidr_netmask="24" iflabel="VIP1" \ > >> op monitor interval="30s" timeout="20s" > >> ms MS-LDAP-SERVER LDAP-SERVER \ > >> params config="/etc/ldap/slapd.conf" lsb_script="/etc/init.d/slapd" \ > >> meta clone-max="2" clone-node-max="1" master-max="1" > >> master-node-max="1" notify="false" target-role="Master" > >> colocation LDAP-WITH-IP inf: VIP-1 MS-LDAP-SERVER > >> order LDAP-AFTER-IP inf: VIP-1 MS-LDAP-SERVER > >> property $id="cib-bootstrap-options" \ > >> dc-version="1.1.7-ee0730e13d124c3d58f00016c3376a1de5323cff" \ > >> cluster-infrastructure="openais" \ > >> expected-quorum-votes="2" \ > >> stonith-enabled="false" \ > >> no-quorum-policy="ignore" \ > >> last-lrm-refresh="1363619724" > >> > >> I use a ocf slapd-ms script provided by this article : > >> > http://foaa.de/old-blog/2010/10/intro-to-pacemaker-part-2-advanced-topics/tra
> ckback/index.html#master-slave-primus-inter-pares > >> > > > > "interesting" setup and resource agent .... you should really go with > > the well tested slapd resource agent that comes with the resource-agent > > package, use slapd as a clone resource and setup slapd to run in > > mirror-mode. > > > > Regards, > > Andreas > > > > Hello, > > thanks for the slapd mirror idea it's works but I hava got one problem, > when one server stop I can continue to add information in the first one, > and when the second server failback It doesn't resync data automaticaly. > > I have to manualy done that : > > crn node standby ldap01-server > cd /var/lib/ldap > rm -rf __db.00* log.* *.bdb > crm node online ldap01-server Hi! I'm not sure, but I think those files are needed to recover the LDAP database after non-graceful termination. I have no experience with a dual-master LDAP configuration, but I think the node's clocks have to be in sync for it to work well (among other issues). Regards, Ulrich > > > Sincerely > -- > ''~`` > ( o o ) > +------------------.oooO--(_)--Oooo.---------------------+ > | Guillaume Chéramy - Guidtz | > | French open source system administrator | > | E-Mail : guilla...@cheramy.name | > | Website : http://blog.cheramy.name | > | | > | ( ) | > +---------------------\ (----( )-----------------------+ > \_) ) / > (_/ > _______________________________________________ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems