F12 -hang on boot - NetworkManger problem?

Peter Glassenbury (CSSE) P.Glassenbury at cosc.canterbury.ac.nz
Thu Oct 15 05:28:54 UTC 2009


Craig White wrote:
> On Thu, 2009-10-15 at 15:00 +1300, Peter Glassenbury (CSSE) wrote:
>> I have F12_x86_64 installed from alpha ISO and kept up to date
>> with a yum update. (running on a intel core2-duo)
>>
>> I am always getting a hang on boot at the stage of..
>>        Starting system message bus.
>>
>> It waits at this point for just over 10 minutes, then
>> continues through normally to runlevel 5.
>>
>> Errors in /var/log/messages say
>>
>> nss_ldap: failed to bind to LDAP server ......can't contact LDAP server
>>
>> It repeats  this many times while starting the dbus-daemon and eventually times out.
>> It repeats  the error through automountd and rpc.statd
>> and then finally NetworkManager runs again and starts the network
>>
>>
>> Where should I report this as a bug? and against what...
>> Network manager for not getting its act together earlier or
>> general init.d boot script order..??
>>
>> In an F10 install we had to remove NetworkManager to get a
>> working system.
>> Currently I'm testing F11 and F12alpha and F11 works fine.
>> (on the same hardware machines)
>>
>> Any help or pointers on what to report greatly received.
> ----
> add the following to /etc/ldap.conf
> 
> nss_initgroups_ignoreusers root, ldap, dbus, named
> timelimit 30
> bind_timelimit 30
> bind_policy soft
> 
> you might want to add other daemon 'users' from /etc/passwd the the
> 'ignoreusers' line too.

I just have the default ldap settings except for the ldap server config..
The timelimits default to 120
the bind_policy has been changed to soft as you said.
This is the default ignore users line....
nss_initgroups_ignoreusers 
root,ldap,named,avahi,haldaemon,dbus,radvd,tomcat,radiusd,news,mailman,nscd,gdm,polkituser

In /etc/nsswitch we set automount,aliases,services,netgroup,
protocols,group,shadow,passwd to "files ldap" and hosts to
"files dns" and the rest to "files"

The above change to bind_policy and shortening the timelimits
has no hang.. Gives the following ERROR (in red on startup)
but then continues on and appears to come up OK and in reasonable
time.
Starting NFS daemon: rpc.nfsd: unable to resolve ANYADDR: nfs to inet address: Name or service not known

So the workaround works -- is it a bug to be reported?
Pete
-- 
-------------------------------------------------------------------
Peter Glassenbury			Computer Science department
pete at cosc.canterbury.ac.nz		University of Canterbury
+64 3 3642987 ext 7762			New Zealand




More information about the test mailing list