> On Mon, 18 Jan 1999, Simon Kirby wrote:
>
> > Okay...I'm still seeing the problem on the mail server that I saw before,
> > but the example I had with the floppy has been fixed, so it seems like
> > there might be something more that we're missing.
>
> Most mail programs do an fsync() on mailboxes at various points in their
> execution. Looking at fs/ext2/fsync.c, it uses get_hash_table
Everything on the machine has either got fsync()/sync() options disabled
or is using an LD_PRELOAD wrapper so that it doesn't. Trust me, I even
patched the kernel to report processes that call fsync()/sync() to track
everything down...It should be only up the OS to flush buffers IMHO.
> extensively, which is probably leading to the needless waiting. Also, it
> doesn't look like any readahead is done on the indirect blocks (everything
> is syncronous, leading to lots of scattered disk io -- dragging the whole
> system's performance down)... If you try replacing ext2_sync_file in
> fs/ext2/file.c with file_fsync, is performance reasonable (I expect it
> should be)?
Hmm...What would this do? Allow flushing to be done asynchronously?
Is there anything that could be unsafe with this change?
Simon-
| Simon Kirby | Systems Administration |
| mailto:sim@netnation.com | NetNation Communications |
| http://www.netnation.com/ | Tech: (604) 684-6892 |
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/