Messages in this thread | | | Date | Wed, 11 Jan 2012 13:17:00 -0600 (CST) | From | Christoph Lameter <> | Subject | Re: [PATCH -mm] make swapin readahead skip over holes |
| |
On Wed, 11 Jan 2012, John Stoffel wrote:
> KOSAKI> so, we can eat free lunch up to 7MB ~= 60(MB/sec) * 1000 / 8.5(ms). > > What if the disk is busy doing other writeout or readin during this > time? You can't assume you have the full disk bandwidth available, > esp when you hit a swap storm like this.
The assumptions by Kosaki are quite conservative.
What if one did not get a disk from the garbage heap but instead has a state of the art storage cluster or simply an SSD (in particular relevant now since HDs are in short supply given the situation in Asia)?
| |