Improve the queuing algorithms used by NFS' asynchronous i/o. The

existing mechanism uses a global queue for some buffers and the
vp->b_dirtyblkhd queue for others.  This turns sequential writes into
randomly ordered writes to the server, affecting both read and write
performance.  The existing mechanism also copes badly with hung
servers, tending to block accesses to other servers when all the iods
are waiting for a hung server.

The new mechanism uses a queue for each mount point.  All asynchronous
i/o goes through this queue which preserves the ordering of requests.
A simple mechanism ensures that the iods are shared out fairly between
active mount points.  This removes the sysctl variable vfs.nfs.dwrite
since the new queueing mechanism removes the old delayed write code
completely.

This should go into the 2.2 branch.

1 files changed, 2 insertions, 2 deletions

diff --git a/sys/nfsserver/nfsrvstats.h b/sys/nfsserver/nfsrvstats.h
index 3d9d801740d6..bc2edeb12ac5 100644
--- a/sys/nfsserver/nfsrvstats.h
+++ b/sys/nfsserver/nfsrvstats.h
@@ -34,7 +34,7 @@
  * SUCH DAMAGE.
  *
  *	@(#)nfs.h	8.1 (Berkeley) 6/10/93
- * $Id: nfs.h,v 1.17 1996/01/30 22:59:39 mpp Exp $
+ * $Id: nfs.h,v 1.18 1996/08/21 21:55:44 dyson Exp $
  */
 
 #ifndef _NFS_NFS_H_
@@ -335,7 +335,7 @@ extern TAILQ_HEAD(nfs_reqq, nfsreq) nfs_reqq;
 #define	NWDELAYHASH(sock, f) \
 	(&(sock)->ns_wdelayhashtbl[(*((u_long *)(f))) % NFS_WDELAYHASHSIZ])
 #ifndef NFS_MUIDHASHSIZ
-#define NFS_MUIDHASHSIZ	67	/* Tune the size of nfsmount with this */
+#define NFS_MUIDHASHSIZ	63	/* Tune the size of nfsmount with this */
 #endif
 #define	NMUIDHASH(nmp, uid) \
 	(&(nmp)->nm_uidhashtbl[(uid) % NFS_MUIDHASHSIZ])