slapd core dumping and locking up.

[Lew A <question@gwi.net>]

Hello,

We are running to pop servers with slapd running on both. We are having a
problem with slapd locking up and we have to force restarting it in order
to get to respond again. (I have enabled debugging and logging so we can
get some more information about this).

Another problem we're having is slapd core dumps sometimes. There is
nothing useful in the logs for this (debugging set to 256). Last night
slapd core dumped on one server and about 14 hours later it dumped on
another server. They both have only dumped 1 time before and it was at the
exact same time (odd that it happens so close to each other). Here are the
logs from one of the servers (pretty useless).

Jan 17 09:04:15 dingo slapd[97658]: daemon: conn=3543943 fd=65 connection
from IP=127.0.0.1:3596 (IP=127.0.0.1:389) accepted.
Jan 17 09:04:15 dingo slapd[97658]: conn=3543943 op=0 BIND
dn="UID=SHBOBALEK,OU=USERS,O=GWI.NET,DC=GWI,DC=NET" method=128
Jan 17 09:04:15 dingo slapd[97658]: conn=3543943 op=0 RESULT tag=97 err=0
text=
Jan 17 09:04:15 dingo slapd[97658]: conn=3543943 op=1 SRCH
base="uid=shbobalek,ou=Users,o=gwi.net,dc=gwi,dc=net" scope=0 filter="(objectClass=*)"
Jan 17 09:04:15 dingo slapd[97658]: conn=3543943 op=1 SEARCH RESULT
tag=101 err=0 text=
Jan 17 09:04:15 dingo slapd[97658]: daemon: conn=3543944 fd=66 connection
from IP=127.0.0.1:3597 (IP=127.0.0.1:389) accepted.
Jan 17 09:04:15 dingo slapd[97658]: conn=3543944 op=0 BIND
dn="UID=POPIG,OU=USERS,O=GWI.NET,DC=GWI,DC=NET" method=128
Jan 17 09:04:15 dingo slapd[97658]: conn=3543944 op=0 RESULT tag=97 err=0
text=
Jan 17 09:04:15 dingo slapd[97658]: conn=3543944 op=1 SRCH
base="uid=popig,ou=Users,o=gwi.net,dc=gwi,dc=net" scope=0 filter="(objectClass=*)"
Jan 17 09:04:15 dingo slapd[97658]: conn=3543944 op=1 SEARCH RESULT
tag=101 err=0 text=
Jan 17 09:04:15 dingo slapd[97658]: daemon: conn=3543945 fd=67 connection
from IP=127.0.0.1:3598 (IP=127.0.0.1:389) accepted.
Jan 17 09:04:24 dingo slapd[57855]: slapd starting

I debuged the core file and got this information:

dingo: {20} % gdb -q -core /var/qmail/supervise/slapd/slapd.core
Core was generated by `slapd'.
Program terminated with signal 6, Abort trap.
#0  0x1021cad8 in ?? ()
(gdb) quit
dingo: {21} %

I have no idea what this error code is. If anyone could help me that'd be
great. We have been having the lock-up issue alot more frequently but
haven't had the time to look into it until now.

We are running FreeBSD 4.5 p-23, OpenLDAP 2.0.25, Configured with
'--with-tls' option only. If you need anymore information that could help
please feel free. Any help you can give us would be great.

Thank you,
Lew A
GWI Operations

---------------------------------
  A tiger can smile
  A snake will say it loves you
  Lies make us evil
---------------------------------