0

I am seeing some very strange results in a filesystem where I have one machine that creates a directory in a mounted NFS filesystem, does some work it in, then makes a remote call to another machine that then also does work in that mounted filesystem. Sometimes, it appears as though the first machine never created the directory or did any work, but the second machine has no way of starting its work without the remote call being made.

Is there a race condition with mounted drives, such that it's possible for the first computer to create the directory, do some work, then make the remote call, and the second machine hasn't synced to see the directory yet, creates it itself, then the original machine's work is gone?

Improve this question
2
  • 2
    Mounted with what protocol?
    – Paul
    Commented Dec 2, 2012 at 2:42
  • Not sure I know, honestly, so hopefully this answers your question? both machines are running Linux, and I type "mount machine:dir dir" and when I type "mount" alone afterward, it says "type nfs (rw,addr=ipaddress)" at the end. If there's something else I should do to check, let me know.
    – nathan
    Commented Dec 2, 2012 at 3:08

1 Answer 1

Reset to default
0

It sould work finen. But ensure the OS is the same (for better compatibility) and ensure to FLUSH the data in the first machine before the other one try to access the same data.

Refer to this link about how to flush in NFS:

force flush on linux NFS

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .