Description of problem:

When files/directories are continuously updated and deleted concurrently. At some point some of them get corrupted and accessing them gives EIO errors.

Version-Release number of selected component (if applicable): mainline


How reproducible:

Most of the times

Steps to Reproduce:
1. Create a distributed-disperse volume
2. Mount it on two clients
3. On one client run "while true; do mkdir a;done"
4. On the other client run "while true;do mv a b; mv b a;done"
5. Wait some minutes
6. Stop the jobs started on steps 3 and 4
7. Run "rm -rf *" from one client

Actual results:

The rm failed with EIO error

Expected results:

The rm shouldn't fail

Additional info: