Re: Corrupt DB after Reboot

["matthew sporleder" <msporleder@gmail.com>]

On 9/17/06, Matthias Spork <hallo@matthiasspork.de> wrote:
> Hello,
>
> each time I reboot my Server, OpenLDAP goes corrupt. I had to slapadd a
> backup from an other server. This behavior is on all my SLES9 Server,
> which runs with same configuration and OpenLDAP-versions.
>
> Could you give me a hint, where I can more analyse my problem?
>
>
> I played with the DB_CONFIG. I changed the file an rebuilded the db
> (slapcat & slapadd). After first reboot it seemed to work. Am I on the
> right way?
>
> Old DB_CONFIG:
> set_cachesize 0 15000000 1
> set_lg_regionmax 262144
> set_lg_bsize 2097152
> set_flags DB_LOG_AUTOREMOVE
>
> New DB_CONFIG:
> set_cachesize   0       26214400        0
> set_lg_max      10485760
> set_lg_bsize    2097152
> set_lg_dir      /var/lib/ldap/transactionlog
> set_tmp_dir     /tmp
>
> Kind regards
> Matthias
>
>
> --------------------------------------------------------------
> srv:/var/lib/ldap # slapcat
> bdb_db_open: unclean shutdown detected; attempting recovery.
> bdb_db_open: Recovery skipped in read-only mode. Run manual recovery if
> errors are encountered.
>
> srv:/var/lib/ldap # db_recover -h /var/lib/ldap/
> srv:/var/lib/ldap # slapcat
> bdb_db_open: unclean shutdown detected; attempting recovery.
> bdb_db_open: Recovery skipped in read-only mode. Run manual recovery if
> errors are encountered.
> bdb_db_open: alock_recover failed
> bdb_db_close: alock_close failed
> backend_startup_one: bi_db_open failed! (-1)
> slap_startup failed
>
> Packages:
> --------------------------------------------------------------
> openldap2-back-perl-2.3.27-0.1.20060821
> openldap2-client-2.3.27-0.1.20060821
> openldap2-back-meta-2.3.27-0.1.20060821
> openldap2-2.3.27-0.1.20060821
> db-4.2.52-86.3
>
> slapd.conf
> --------------------------------------------------------------
> include         /etc/openldap/schema/core.schema
> include         /etc/openldap/schema/cosine.schema
> include         /etc/openldap/schema/misc.schema
> include         /etc/openldap/schema/nis.schema
> include         /etc/openldap/schema/inetorgperson.schema
> include         /etc/openldap/schema/openldap.schema
> include         /etc/openldap/schema/samba3.schema
>
> database        bdb
> suffix          "o=comp,c=COM"
> rootdn          "cn=admin,o=comp,c=COM"
> rootpw          {crypt}xxxxx
> password-hash   {CRYPT}
> directory       /var/lib/ldap
>
> TLSCipherSuite          HIGH:MEDIUM:+SSLv2:+SSLv3
> TLSCACertificateFile    /etc/ssl/certs/cacert.pem
> TLSCertificateFile      /etc/openldap/slapd_cert.pem
> TLSCertificateKeyFile   /etc/openldap/slapd_key.pem
>
> sizelimit       5000
> schemacheck     on
> lastmod         on
> loglevel        0
> pidfile         /var/run/slapd/slapd.pid
> argsfile        /var/run/slapd/slapd.args
>
> index   default         eq
> index   uid             eq,sub
> index   sn              eq,sub
> index   objectClass     eq
> index   sambaSID        eq,sub
> index   sambaPrimaryGroupSID eq
> index   sambaDomainName eq
> index   memberUid       eq
> index   gidNumber       eq
> ...

openldap 2.3 will automatically determine if you need to recover and
do it for you.
You are on the right track in that committing your logs more
frequently will mean syncing to disk will take less time when your
server issues a killall.  You could also try to shutdown openldap more
cleanly before rebooting (just wait for it to stop on its own).