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

bdb-backend problems

I had problems with ldbm last week, I wasn't able to write because my
reading machines blocked io.

I upgraded to bdb and for one day, all was perfect.
Now the following: Yesterday my ldap-server with bdb was unreachable (only
the daemon). I stopped it and started again (checked they were gone;), but
still 3 slapd processes used 33% of the cpu each, the ldap server doesn't
answer etc.
My hardware is a Celeron 333 with an 2940 SCSI and one HD.
I have about 50 machines asking, is there a chance to do this with that
I haven't started reading bdb tuning howtos, should that be the right thing?

What I can't understand is that I managed all these machines with an
openldap 2.0.23 for about 2 years without
any load problems.

Thanks for your great help