> I am afraid the behaviour of your tail(1) is related to the > specification or feature of nfs, too. > Will you try unmounting nfs after rmdir in 'Mutt + /tmp+nfs', and see > what will happen? Although you may expect an error EBUSY, it may kill > your tail process too. Please try it. [EMAIL PROTECTED]:/mnt# touch A [EMAIL PROTECTED]:/mnt# tail -f A [1]+ Stopped tail -f A [EMAIL PROTECTED]:/mnt# bg [1]+ tail -f A & [EMAIL PROTECTED]:/mnt# cd .. [EMAIL PROTECTED]:/# umount /mnt [EMAIL PROTECTED]:/# umount /mnt [EMAIL PROTECTED]:/# pkill tail [1]+ Terminated tail -f A (wd: /mnt) (wd now: /) [EMAIL PROTECTED]:/# umount /mnt
Any way I wrote a piece of code to show up the problem. Please find it at the end. I called it "test_auf". AUFS NFS: estrella:~/testeaufs/root# ./test_aufs 0 -1 XFS NFS: estrella:/mnt# ./test_aufs 0 0 Bona #include <stdio.h> #include <sys/stat.h> #include <sys/types.h> int main(void) { FILE* file; int ret, fd; struct stat buf; mkdir("testdir", (mode_t) 255); file=fopen("testdir/testfile", "w+"); fd=fileno(file); ret=fstat(fd, &buf); fprintf(stderr, "%d\n", ret); /* now unlinks testtile */ unlink("testdir/testfile"); rmdir("testdir"); /* Now it returns erros with aufs NFS, but it works with ext2 xfs */ ret=fstat(fd, &buf); fprintf(stderr, "%d\n", ret); return 0; } ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/