Repost: Preventing DB corruptions or other bad surprises

[marc.bigler@day.com]

Hello,

I am using the latest stable version of OpenLDAP on Solaris 9 and want to
make sure that in case of a hardware crash of any kind that my OpenLDAP
Berkley database is fine. So I've read that I should use the checkpoint
parameter in my slapd.conf, and that's what I did, I've added the
following:

checkpoint 1024 30

That would mean that the database buffers will be flushed to disk every
1024 kBytes of written data AND every 30 minutes. Do you think that's
allright ? This LDAP server is quite small only a 1000 entries, mostly user
accounts and sendmail stuff and modifications or new entries are not
happening very offten.

Now is that enough to prevent database corruptions or such kind of things
to happen in case of a hardware failure or power outtage ?

Another question, I've read in the FAQ that it's now possible to run
slapcat for backups on a running OpenLDAP instance without any problems, is
that right ?

Many thanks !

Regards
Marc