| To: | David Chinner <dgc@xxxxxxx> |
|---|---|
| Subject: | Re: [PATCH 3/9] Use _META bio I/O types for metadata I/O |
| From: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
| Date: | Fri, 23 Nov 2007 17:37:23 +0000 |
| Cc: | xfs-oss <xfs@xxxxxxxxxxx>, lkml <linux-kernel@xxxxxxxxxxxxxxx> |
| In-reply-to: | <20071122003512.GI114266761@xxxxxxx> |
| References: | <20071122003512.GI114266761@xxxxxxx> |
| Sender: | xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.4.2.3i |
On Thu, Nov 22, 2007 at 11:35:12AM +1100, David Chinner wrote: > Improve metadata I/O merging in the elevator > > Change all async metadata buffers to use [READ|WRITE]_META I/O types > so that the I/O doesn't get issued immediately. This allows merging > of adjacent metadata requests but still prioritises them over bulk > data. This shows a 10-15% improvement in sequential create speed of > small files. > > Don't include the log buffers in this classification - leave them > as sync types so they are issued immediately. Looks good, and just including the trivial fs.h addition here might be okay aswell. |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: REVIEW: xfs_reno #2, Ruben Porras |
|---|---|
| Next by Date: | Re: [PATCH 4/9] Factor common inode cluster buffer lookup code, Christoph Hellwig |
| Previous by Thread: | [PATCH 3/9] Use _META bio I/O types for metadata I/O, David Chinner |
| Next by Thread: | [PATCH 4/9] Factor common inode cluster buffer lookup code, David Chinner |
| Indexes: | [Date] [Thread] [Top] [All Lists] |