Hello

I have Netapp NFS drive and 6 RHEL linux servers currently access same share.

My problems is related with multiple file delete operation from NFS drive
I could create files in NFS mount with specific users. There no problem in this step.
I have housekeeping scripts for example deleting daily files. Number of daily files about 30000-40000. (total size about 3-4 TB)

My problem starts here. When I delete this files (cron script) in one command like "rm -rf *201612080*" or "rm *201612080*" commands sometimes could not delete any file or sometimes delete all files. When the command fails there is no error, command thinks, OS executed request without no problem. So this problem needs some additional checkings day by day, after cron script.

Another strange issue is. works everytime but following small script works sometime

So why cannot delete multiple files sometime ?

Thanks

With that many files your most likely getting an argument list to long error. The wild card actually expands so rm *xxxx* becomes rm file1 file2 file3 ... file400000. A better way to delete that many files is to use find.

find /dir/to/files -type f -name "*201612080*" -exec rm -f {} \;

If during the deletion there are files created (or deleted) by other processes then it becomes a race condition. If several NFS clients are involved, the NFS caches make the problem worse.
The list generated for the for loop is another buffer.
For mitigation, you must buffer less, like the suggested find.
Make sure your deletion does not run on several NFS clients simultaneously.