NAME<\/a> File::Sync \u2212 Perl access to fsync() and sync() function calls<\/p>\n use File::Sync qw(fsync sync); The fsync()<\/b> function takes a Perl file handle as its only argument, and passes its fileno()<\/b> to the C function fsync()<\/b>. It returns undef<\/i> on failure, or true<\/i> on success. fdatasync()<\/b> is identical in return value, but it calls C fdatasync()<\/b> instead of fsync()<\/b>, synchronizing only the data in the file, not the metadata.<\/p>\n The fsync_fd()<\/b> function is used internally by fsync()<\/b>; it takes a file descriptor as its only argument.<\/p>\n The sync()<\/b> function is identical to the C function sync()<\/b>.<\/p>\n This module does not<\/b> export any methods by default, but fsync()<\/b> is made available as a method of the FileHandle<\/i> class. Note carefully that as of 0.11, we no longer clobber anything in IO::Handle<\/i>. You can replace any calls to IO::Handle::fsync()<\/b> with IO::Handle::sync()<\/b>: Doing fsync()<\/b> if the stdio buffers aren\u2019t flushed (with $| or the autoflush<\/i> method) is probably pointless.<\/p>\n Calling sync()<\/b> too often on a multi-user system is slightly antisocial.<\/p>\n
SYNOPSIS<\/a>
DESCRIPTION<\/a>
NOTES<\/a>
AUTHOR<\/a>
SEE ALSO<\/a> <\/p>\n
\nNAME <\/a> <\/h2>\n
SYNOPSIS <\/a> <\/h2>\n
sync();
fsync(*FILEHANDLE) or die “fsync: $!”;
# and if fdatasync() is available on your system:
fdatasync($fh) or die “fdatasync: $!”;
use File::Sync qw(fsync);
use FileHandle;
$fh = new FileHandle(“> \/tmp\/foo”)
or die “new FileHandle: $!”;
…
$fh\u2212>fsync() or die “fsync: $!”;<\/p>\nDESCRIPTION <\/a> <\/h2>\n
https:\/\/rt.cpan.org\/Public\/Bug\/Display.html?id=50418<\/p>\nNOTES <\/a> <\/h2>\n
AUTHOR <\/a> <\/h2>\n