On Mon, Nov 4, 2013 at 12:26 PM, m.roth@5-cent.us wrote:
I'm trying to make sense of that timing. Does that mean that pre-6.x, fsync() didn't really wait for the data to be written to disk, or does it somehow take 7 minutes to get 100M onto your disk in the right order? Or is this an artifact of a specific raid controller and what you have to do to flush its cache?
No, this is regardless of what box, old Penguins, newer Dell's with PERC 600 or 700 RAID controllers. Apparently, this "barrier" controls journalling transactions, so that they are in order, or something like that.
I just don't see where that kind of time can go unless it is forcing a flush of a large (and probably mostly unrelated) cache to disk - possibly even in the internal drive caches if there is a way to do that, and waiting for it to complete after each file close.