Dear Tuxera,
can you clarify please *what* is

message ?
Cluster size, in fact, is exactly set to 64K. The data is fed through SAMBA, CPU load never goes for above 2% of a single core.

This causes problems with isochronous data transfers.

Correction: ordinary write ops load the system for very reasonable amounts of CPU (40% and more)

Hi,



This is the conclusion of a standard sanity check. I think this happens at initialization time while analyzing a transaction log file, which has never been written to (which means this device has never been written to by Windows), but I am not sure.

Unless you have some other error message, you need not worry.


This is not strange if you are feeding your computer from a high speed network. What is the throughput and what is the load on the ntfs-3g process ?

Regards

Jean-Pierre

But these are plenty of in /var/log/message
Please consider also that many NAS systems are built over SSD - a kind of media where the rewrite count per sector is limited


The NAS was not overloaded

I'am still unsure which actually caused stalls; most likely is a client