HA - SAMBA & CTDB Configuration Issue using LDAP as back end

SLES High Availability Extension
1

Using SLES 11 SP3 w/HA Extensions

I have setup a SAMBA 3 cluster (running on SLES-11sp3) with 2 nodes. CTDB is managing SAMBA, using a floating IP between the two hosts. Everything is running fine as far as the cluster fail-over, etc. However, the SAMBA cluster server(s) are supposed to be member servers, connecting to an existing PDC which is using LDAP as the password backend.

CTDB is inserting the following lines into the smb.conf file once it starts SAMBA on a given node:

[INDENT][CODE] # CTDB-RA: Begin auto-generated section (do not change below)
passdb backend = tdbsam
clustering = yes
idmap backend = tdb2
ctdbd socket = /var/lib/ctdb/ctdb.socket
fileid:algorithm = fsid

CTDB-RA: End auto-generated section (do not change above)[/CODE]

[/INDENT]

Unfortunately, even when I change the entries to
[INDENT]passdb backend = ldapsam:ldap://LDAP-Server idmap backend = ldap:ldap://LDAP-Server[/INDENT]

CTDB rewrites it to the tdbsam entries upon take-over/failover to that node.

I have searched, but can’t find any configuration option to override this, although the documentation says that you can use either tdbsam or ldap as the back-end for CTDB-managed SAMBA clusters.

I have tried the following:

[INDENT][LIST]
[]stop CTDB (and clusster services) on both nodes
[]stop SAMBA
[]edit /etc/samba/smb.conf to reflect the desired passdb backend entry.
[]synchronize both systems with the edited smb.conf file
[]copy the smb.conf file to /etc/ctdb/state/samba/smb.conf.cache
[]reboot both server & start all services
[*]
[/LIST][/INDENT]

Check files - still saying
[INDENT]passdb backend = tdbsam
idmap backend = tdb2
[/INDENT]

Is there a file (or set of files) that I’m missing to be edited? I know you say “it should just work” and my gut feeling says the same, particularly as the docs don’t have any specific mention if this, so it doesn’t indicate that there should be a problem.

Has anyone dealt with this issue before? Any pointers to a solution? This is the only piece missing from my SAMBA cluster setup to be able to go live with it.

George-G

2

On Wed, 04 Jun 2014 08:44:02 +0000, digisolv wrote:
[color=blue]

I have setup a SAMBA 3 cluster (running on SLES-11sp3) with 2 nodes.
CTDB is managing SAMBA, using a floating IP between the two hosts.[/color]

Is this CTDB the resource agent:

http://linux-ha.org/wiki/CTDB_(resource_agent)

or CTDB the SAMBA management thing:

https://ctdb.samba.org/

I’m guessing that it’s the latter.

[color=blue]

Has anyone dealt with this issue before? Any pointers to a solution?[/color]

This sounds to me like a bug in CTDB, especially if the observed behaviour
of the code doesn’t seem to match the documentation.

–

David Gersic dgersic_@_niu.edu
Knowledge Partner http://forums.netiq.com

Please post questions in the forums. No support provided via email.
If you find this post helpful, please click on the star below.