Our Kubuntu computers are configured to automatically connect to their network drive via AutoFS. When we upgrade machines to Kubuntu 18.04 LTS this AutoFS connection no longer works. It is still working fine in Kubuntu 14.04 LTS and Kubuntu 16.04 LTS.
I have run autoFS (automount) in debug mode:
automount -f -d
.. and receive the following error:
mount error(126): **Required key not available** Refer to the mount.cifs(8) manual page (e.g. man mount.cifs)
I can use klist and see a kerberos key is present but still won't map (why?)
Now after some fooling around the kerberos key now works reliably only in debug mode (automount -f -d). It doesn't work at all when run as a service (service autofs start). Why does it only work in debug mode..?
Log : Error
Here is the error in daemon.log :
May 16 17:49:02 computer1234 cifs.upcall: get_cachename_from_process_env: pathname=/proc/1234/environ May 16 17:49:02 computer1234 cifs.upcall: get_existing_cc: default ccache is FILE:/tmp/krb5cc_12345678 May 16 17:49:02 computer1234 cifs.upcall: get_tgt_time: unable to get principal May 16 17:49:02 computer1234 cifs.upcall: krb5_get_init_creds_keytab: -1765328174 May 16 17:49:02 computer1234 cifs.upcall: Exit status 1
I had tried adding '-t' in /etc/request-key.d/cifs.spnego.conf and it still was not working
If you have similar problem you can try adding logging = "verbose" or logging = "debug" in /etc/autofs.conf to get more detail.
Yesterday we noticed that autofs is looking for the kerberos ticket file named in this format: /tmp/krb5cc_12345678
However the actual kerberos ticket file name has 7 more chars at the end and looks like this: /tmp/krb5cc_12345678_1A23B4