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

openLDAP not starting on redhat 8


I have a somewhat interesting/perplexing problem. I am unable to start openldap on a redhat 8 box after I copy a database to it that was created on a redhat 9 box. Here are the facts of the case so far:

1). The build process was almost identical on both machines. I used the same version of BDB (4.2.52) and openLDAP (2.1.25).

2). I captured the output of the BDB and OpenLDAP configuration scripts from both the rh8 and rh9 machines and diffed them. The only differences were in the BDB configuration outputs, and those differences were:

root@localhost:/tmp 1# diff config_bdb_rh9.log config_bdb_rh8.log
< checking if cc supports -fno-rtti -fno-exceptions... no
> checking if cc supports -fno-rtti -fno-exceptions... yes
< checking for mutexes... POSIX/pthreads/library
> checking for mutexes... x86/gcc-assembly

3). When OpenLDAP on the rh8 machine is started with an empty db, it starts fine.

4). Both machines have identical schema files, and their slapd.conf files are derived from templates, and thus are essentially the same.

5). After copying the data files from the rh9 machine to the rh8 machine, I try to start ldap on the rh8 machine with '/usr/local/libexec/slapd -d 256' and I get the following output:

[root@ws01 ldap]# /usr/local/libexec/slapd -d 256
daemon: socket() failed errno=97 (Address family not supported by protocol)
bdb_initialize: Sleepycat Software: Berkeley DB 4.2.52: (December 3, 2003)
bdb_db_init: Initializing BDB database
bdb(dc=vindicia,dc=com): Program version 4.2 doesn't match environment version
bdb_db_open: dbenv_open failed: Invalid argument (22)
backend_startup: bi_db_open(0) failed! (22)
bdb(dc=vindicia,dc=com): txn_checkpoint interface requires an environment configured for the transaction subsystem
bdb_db_destroy: txn_checkpoint failed: Invalid argument (22)
slapd stopped.
connections_destroy: nothing to destroy.

The error part starts after the "bdb_db_init: Initializing BDB database" line as the socket error on the top is encountered on all machines and appears to be non-fatal (and I believe has something to do with IPv6 addresses?). Now, being that both configurations are nearly identical, I am wondering if the differences in the BDB configurations may have something to do with this issue. Furthermore, I tried the same build on another rh8 machine, and diffed its config logs against the first rh8 machine. No differences were shown, and when I copied the rh9 data files across, I got the same error. This build process has been used on numerous rh9 machines so far and no problems were encountered when I copied the data files across from the master server. Any help would be greatly appreciated. Many thanks in advance.


Peter Traub