Calls to os.getcwd() can fail if issued from a cifs-mounted directory once any of the files or subdirectories have been changed remotely.  To recreate this: on Linux, mount a windows share using the mount.cifs command.  cd to the share and run python.   

Python 2.7.3 (default, Jul 24 2012, 10:05:38)
[GCC 4.7.0 20120507 (Red Hat 4.7.0-5)] on linux2
Type "help", "copyright", "credits" or "license for more information.
>>> import os
>>> print os.getcwd()
 /home/user/share
>>> exit()

That works.  Now, on the host machine, change a file in the current directory and save it.  The unix 'pwd' still works.  But now:
>>> import os
>>> print os.getcwd()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
OSError: [Errno 2] No such file or directory
>>>

Even though the unix 'pwd' and 'ls' commands still work, so I know my share is still accessible.

If I umount and then remount the share, the problem goes away until a file is changed remotely.

I'm running this on Fedora 17.

os.getcwd() just calls the libc getcwd(3), so Python's not the problem here.

it's likely an issue with the CIFS implementation (I guess you're using fuse?).

Could you post the output of:
$ strace python -c "import os; os.getcwd())"

upon failure?

> Even though the unix 'pwd' and 'ls' commands still work, so I know my share is still accessible.

'pwd' is likely your shell builtin, which doesn't call getcwd(). You could try with /bin/getcwd. As for 'ls', it doesn't walk the directory tree like getcwd() does.

I apologize for not posting the strace output - I didn't see the request until today.  

Libc getcwd() is indeed failing - when I run a c program which calls getcwd() and prints the output, the call to getcwd() fails - errno is ENOENT and the buffer is null.  

I will look elsewhere.  Thanks for getting me on the right track.