CODEWITHSUNDEEP
centosopenldap
爱国者
爱国者

Reputation: 4348

openldap fails to bind ldaps://127.0.0.1:636

Here is my testcase :

[[email protected] ~$]slapd -d 1 -h ldaps://127.0.0.1:636
@(#) $OpenLDAP: slapd 2.4.23 (Apr 29 2013 07:47:08) $
[email protected]:/builddir/build/BUILD/openldap-2.4.23/openldap-2.4.23/build-servers/servers/slapd
ldap_pvt_gethostbyname_a: host=centos-6.3, r=0
daemon_init: listen on ldaps://127.0.0.1:636
daemon_init: 1 listeners to open...
ldap_url_parse_ext(ldaps://127.0.0.1:636)
daemon: bind(7) failed errno=98 (Address already in use)
slap_open_listener: failed on ldaps://127.0.0.1:636
slapd stopped.
connections_destroy: nothing to destroy.

But if I change another port , such as 6361, it works.

My environment:

OS: centos 6.4 x86_64
OpenLDAP: 2.4.23 installed by yum

Any suggestion?

Upvotes: 0

Views: 8590

Answers (2)

Lipsum
Lipsum

Reputation: 546

Old post, but still ...

This error is also displayed when SELinux prevents slapd from starting. Personally I experienced this after manually copying data (/var/lib/ldap/) from another server, to this one. I had to restore the imported files to default SELinux security contexts:

restorecon -R /var/lib/ldap

And I see this doesn't apply to you, but this might also happen if you're attempting to bind slapd to a port out of the ordinary. Default on CentOS7, these are the allowed ports:

#semanage port -l | grep ldap
ldap_port_t                    tcp      389, 636, 3268, 7389
ldap_port_t                    udp      389, 636

Adding another one to the legal port range, could be done with semanage. (You might need to install the package policycoreutils-python.):

semanage port -a -t ldap_port_t -p tcp 10389

... if you wish to allow slapd to bind on TCP port 10389 in addition to the four listed above. After this, the previous result would look like:

# semanage port -l | grep ldap
ldap_port_t                    tcp      10389, 389, 636, 3268, 7389
ldap_port_t                    udp      389, 636

Upvotes: 1

free_easy
free_easy

Reputation: 5129

it seems that another service is already running on port 636:

daemon: bind(7) failed errno=98 (Address already in use)

you can try the following command to identify this service:

netstat -tulpn | grep ':636 ' | grep 'LISTEN'

Upvotes: 1

Related Questions