Hello openldap users!

I've got Openldap 2.4.23 that used as authentication and authorization server for about 40-50 servers. OS - FreeBSD 8.1.

It's not heavy loaded.

openldap# top -SP last pid: 45647; load averages: 0.15, 0.15, 0.07

up 81+22:29:21 15:18:57 99 processes: 3 running, 80 sleeping, 16 waiting CPU 0: 0.7% user, 0.0% nice, 0.0% system, 0.0% interrupt, 99.3% idle CPU 1: 0.4% user, 0.0% nice, 0.7% system, 0.0% interrupt, 98.9% idle Mem: 79M Active, 1402M Inact, 379M Wired, 84M Cache, 213M Buf, 31M Free Swap: 4060M Total, 8K Used, 4060M Free

PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND 11 root 2 171 ki31 0K 32K CPU0 0 3874.8 200.00% idle 4773 ldap 18 44 0 398M 53748K ucond 1 41.1H 0.00% slapd

But on my servers sometimes I see in logs something like

on FTP-server: Mar 25 21:55:32 someftp ftpd: nss_ldap: could not search LDAP server - Server is unavailable

Authentication works fine, no problems. But want to find out what can be wrong.

To understand this problem I installed ldap-stats utility and made it run:

/var/log/debug.log - it's half day openldap server usage log.

openldap# ldap-stats -c 1000 /var/log/debug.log

Report Generated on Tue Apr 5 15:16:47 2011 -------------------------------------------- Processed "/var/log/debug.log": Apr 5 00:00:00 - Apr 5 15:17:33

Operation totals ---------------- Total operations : 913845 Total connections : 101226 Total authentication failures : 2 Total binds : 99700 Total unbinds : 99181 Total searches : 714964 Total compares : 7 Total modifications : 0 Total modrdns : 0 Total additions : 0 Total deletions : 0 Unindexed attribute requests : 0 Operations per connection : 9.03

# Uses Filter ---------- ----------------------------------------------------------- 615504 (&(objectClass=posixAccount)(uid=mailer-daemon)) 90699 (&(objectClass=posixGroup)) 6833 (&(objectClass=posixAccount)(uid=root)) 2236 (&(objectClass=posixAccount)(uid=hiddenuser1)) 669 (&(objectClass=posixGroup)(memberUid=root)) 318 (&(objectClass=posixAccount)(uid=testacc)) 87 (&(objectClass=posixGroup)(memberUid=postfix)) 87 (&(objectClass=posixAccount)(uid=postfix)) 81 (objectClass=posixAccount) 68 (&(objectClass=posixAccount)(uid=debian-exim)) 68 (&(objectClass=posixGroup)(memberUid=Debian-exim)) 39 (&(objectClass=posixAccount)(uid=normaluser)) 34 (&(objectClass=posixAccount)(uidNumber=7333)) 30 (&(objectClass=posixGroup)(memberUid=hiddenuser1)) 29 (&(objectClass=posixGroup)(memberUid=chelovek)) 29 (&(objectClass=posixAccount)(uid=chelovek)) 27 (&(objectClass=posixAccount)(uid=user0)) 23 (&(objectClass=posixAccount)(uid=nobody)) 21 (&(objectClass=posixAccount)(uid=user1)) 18 (&(objectClass=posixAccount)(uid=user2)) 16 (&(objectClass=posixAccount)(uid=user3)) 15 (&(objectClass=posixAccount)(uid=user4)) 12 (&(objectClass=posixAccount)(uid=user5)) 11 (&(objectClass=posixAccount)(uidNumber=7330)) 10 (&(objectClass=posixAccount)(uid=user15)) 9 (&(objectClass=posixAccount)(uid=user16)) 8 (&(objectClass=posixAccount)(uidNumber=7333)) 6 (&(objectClass=posixAccount)(uid=user6)) 5 (&(objectClass=posixAccount)(uid=user7)) 5 (cn=defaults) 4 (&(objectClass=posixAccount)(uidNumber=7228)) 4 (&(objectClass=shadowAccount)(uid=user1)) 4 (&(objectClass=posixAccount)(uid=user9)) 4 (&(objectClass=posixAccount)(uid=user10)) 4 (&(objectClass=posixAccount)(uid=user11)) 3 (&(objectClass=posixAccount)(uid=user12)) 3 (&(objectClass=posixAccount)(uid=user13)) 3 (&(objectClass=posixAccount)(uid=user14)) ............... and MANY others that has 1 use in this stats. I think this many queries from mail relay server. * user1 and etc - just hidden real users.

What can I do to tune nss? Can you point me in a right direction? Do not know what to look at. If you need any additional information, logs and etc - I'll provide it.

Thanks in advance!