[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

SASL with Kerberos



Hi,

I just tested SASL 2.1.18, change the host and service name to be the 
same name during the testing of the sample client and server, it 
actually added the new principles to the kerberos cache (running Heimdal 
Kerberos 5, the latest version as I downloaded today).
The klist shown the following new principles had been added to the 
kerberos cache:

root@fbsd [7:26pm] [...cyrus-sasl-2.1.18/sample]# klist
Credentials cache: FILE:/tmp/krb5cc_0
      Principal: sam@ROCK.COM

Issued           Expires          Principal               Jun  3 
17:17:53  Jun  3 23:57:53  krbtgt/ROCK.COM@ROCK.COM  Jun  3 17:18:53  
Jun  3 23:57:53  host/fbsd.rock.com@ROCK.COM
Jun  3 18:46:25  Jun  3 23:57:53  root/fbsd.rock.com@ROCK.COM
Jun  3 19:15:24  Jun  3 23:57:53  sam/fbsd.rock.com@ROCK.COM

The last three Principals were added during the test of sample client 
and server in Cyrul-sasl 2.1.18.
But but the test still returned error such as:
lt-sample-client: SASL Other: GSSAPI Error:  A token was invalid 
(Unknown error: 0)
lt-sample-client: Performing SASL negotiation: generic failure

What should I do to fix this problem? I m afraid this will bring in 
other problem when I further configure OpenLdap.

Thanks
sam


The Shell wrote:

> Hi,
>
> I finally got GSSAPI compiled with SASL, but error occured when 
> testing the sample client and server.
> The klist command of Heimdal Kerberos 5 shown the following priciples:
> root@fbsd [5:13pm] [...cyrus-sasl-2.1.18/sample]# klist
> Credentials cache: FILE:/tmp/krb5cc_0
>        Principal: sam@ROCK.COM
>   Issued           Expires          Principal               Jun  3 
> 17:17:53  Jun  3 23:57:53  krbtgt/ROCK.COM@ROCK.COM  Jun  3 17:18:53  
> Jun  3 23:57:53  host/fbsd.rock.com@ROCK.COM
> root@fbsd [5:31pm] [...cyrus-sasl-2.1.18/sample]#
>
> Message from the sample server::
> ./sample-server -s host -p ../plugins/.libs
> .......
> got 'GSSAPI'
> Sending response...
> S: 
> YGwGCSqGSIb3EgECAgIAb10wW6ADAgEFoQMCAQ+iTzBNoAMCARCiRgREAEQM3hY7ovvFlIeYJwJOZzxv+NwWaQnhoHi6007SbsVDMiJfeHZpYU/PHelUTE6CwS46H8N10ObrvAAwKDzXXb2nIh0= 
>
> Waiting for client reply...
> ^C
> root@fbsd [5:22pm] [...cyrus-sasl-2.1.18/sample]#
>
> Message from sample client:
> ./sample-client -s host -n fbsd.rock.com -u root -p ../plugins/.libs
> .....
> C:
> Waiting for server reply...
> S: 
> YGwGCSqGSIb3EgECAgIAb10wW6ADAgEFoQMCAQ+iTzBNoAMCARCiRgREAEQM3hY7ovvFlIeYJwJOZzxv+NwWaQnhoHi6007SbsVDMiJfeHZpYU/PHelUTE6CwS46H8N10ObrvAAwKDzXXb2nIh0= 
>
> recieved 110 byte message
> lt-sample-client: SASL Other: GSSAPI Error:  A token was invalid 
> (Unknown error: 0)
> lt-sample-client: Performing SASL negotiation: generic failure
> root@fbsd [5:21pm] [...cyrus-sasl-2.1.18/sample]#
>
> I m using the latest version of Cyrus-sasl, Heimdal Kerberos in 
> FreeBSD 5.2.1
> thanks
> sam
>
>
>


> eBSD4.9, the slave is openldap-2.1.22 on RH-7.3.
>
> So, it looks like the master is sending ldifs via slurpd to the slave, 
> and the slave is refusing to make the modifications, possibly due to a 
> hardcoded schema.
>
>
> The slurpd reject file looks like this:
>
> ERROR: entryCSN: no user modification allowed
> replica: ldap:0
> time: 1086269077.0
> dn: uid=myuser,ou=radius,dc=mydomain,dc=com
> changetype: modify
> replace: userPassword
> userPassword:: ********
> -
> replace: entryCSN
> entryCSN: 2004060313:24:37Z#0x0001#0#0000
> -
> replace: modifiersName
> modifiersName: uid=myadmin,dc=mydomain,dc=com
> -
> replace: modifyTimestamp
> modifyTimestamp: 20040603132437Z
>
>
> slurpd shows:
>
> Initializing session to ldap:0
> bind to ldap:0 as uid=myadmin,dc=mydomain,dc=com (simple)
> request 1 done
> replica ldap:0 - modify dn "uid=myuser,ou=radius,dc=mydomain,dc=com"
> request 2 done
> Error: ldap_modify_s failed modifying "entryCSN: no user modification 
> allowed": uid=myuser,ou=radius,dc=domain,dc=com
> Error: ldap operation failed, data written to 
> "/var/db/openldap-slurp/replica/ldap:0.rej"
>
>
>
> Have I missed something?  Is it obvious what's wrong?
>
> Thanks,
>
> Gavin
>
>