In the audit-log there is nothing what would explain this. But in
I see a lot of threads like:
The audit log itself (and search log) will generate IO themself :)
1621 be/4 dirsrv 0.00 B/s 3.95 K/s 0.00 % 0.46 % ns-slapd
/etc/dirsrv/slapd-ldap0 -i /var/run/dirsrv/slapd-ldap0.pid -w
Sadly this doesn't tell us much :(
I configured caching and have entrycachehits about 99 - but anyway
would have only impact to read-operations.
DBcache would help with writes.
What's conspicuous: One of the three server has a significant higher
write rate than the others. When I watch our munin stats the two
have with 40 to 60 write operations per second 10 times less. And
of the server suddenly reduces write-rates from more than 100 to
It would be good to see the ldbm config here I think. It would also be
good to know what kind of work load you have. In some configurations of
the network, things like load balancers could be sending more
read/write traffic to one server, rather than over all of them.
Another spot to check is your log buffer size, as it may be small
(causing the threads to all be writing small ammounts). The logging
subsystem is slated for a rework in the future to make it more
Hope that helps,
389-users mailing list -- 389-users(a)lists.fedoraproject.org
To unsubscribe send an email to
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html