Description of problem: The test in function is_multimount_entry from the patch autofs-4.1.4-non-replicated-ping.patch is wrong so multimount paths don't work. With /net/usr/share -ro,hard local-1,local-2:/export/site/d01/local/share in the log we have : automount[3422]: mount(nfs): host local-1,local-2: lookup failure Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: I think you have to replace return !strchr(what, ',') && (strchr(what, ':') == strrchr(what, ':')); by return strchr(what, ',') || (strchr(what, ':') != strrchr(what, ':'));
Yes, I caught this one too late for the FC 4 release. I've got it fixed in CVS, I'll queue a build. Thanks.
Can you try autofs-4.1.4-9 from rawhide? Thanks.
It's OK. Thanks.
autofs-4.1.4-9 does not work for me :-(. I have to change /dev/fd0's filesystem type in /etc/fstab to vfat to make it work.
Not sure if this should go here (multimount entries?), but I do mount home directories from ldap using wildcards. My current setup works well with FC3 and RHEL4. However it is broken in FC4 in both the stock autofs-4.1.4-5 or the autofs-4.1.4-9 package from rawhide. I took the latest release of autofs for FC3 (autofs-4.1.3-114.i386.rpm) and installed it in my FC4 machine and autofs works fine with the wildcards from ldap. If you need any kind of information to help troubleshoot the problem, I will be glad to provide logs, config files, etc. Diego
Please visit: http://people.redhat.com/jmoyer And provide all of the information requested under the "Filing bug reports" section. That will help me diagnose the problem. Thanks!
Created attachment 118216 [details] Tests run with stock autofs-4.1.4-5
We already know there are problems with 4.1.4-5. I'd be more interested in information gathered while running 4.1.4-9. Thanks!
Created attachment 118217 [details] Tests run with suggested rawhide package autofs-4.1.4-9
Created attachment 118218 [details] Debug log with rpm from FC3 which works correctly
Setting back to assigned.
Ian updated the package to more closely match the upstream implementation. Could you try again with the latest version?
Closed due to no feedback from reporter.