Skip to contentSkip to conversation

NFSD: Increase the default max_block_size to 4MB

PATCH In review Chuck Lever 3 participants 3 messages Apr 20

Subsystems (2)
lore
Share & tools
lore

Chuck Lever

author maintainer
#0 lore raw

PATCH 1/1 NFSD: Increase the default max_block_size to 4MB

From: Chuck Lever <[email protected]>

Commit 8a81f16de64f ("NFSD: Add a "default" block size") introduced NFSSVC_DEFBLKSIZE at 1MB, well below the 4MB NFSSVC_MAXBLKSIZE ceiling, with the stated intent that a later change would raise the default.

Raising the default reduces per-RPC overhead on fast networks by amortizing header processing and scheduling costs across larger payloads. The halving loop in nfsd_get_default_max_blksize() constrains the returned value to 1/4096 of available RAM, so the new 4MB default takes effect only on systems with at least 16GB of RAM. Smaller machines continue to receive the same computed value as before. Administrators can still override the computed value through /proc/fs/nfsd/max_block_size.

On systems where the new default takes effect, svc_sock_setbufsize() sizes each service socket's send and receive buffers as nreqs * max_mesg * 2. Quadrupling max_mesg therefore quadruples the per-socket buffer reservation at a fixed thread count, which operators tuning large thread pools should account for.

Note well: Your NFS client implementation must support large read and write size settings to benefit from this change.

---
 fs/nfsd/nfsd.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
fs/nfsd/nfsd.h +2 -3
index a01d70953358..daa63c1b161c 100644
@@ -45,11 +45,10 @@ bool nfsd_support_version(int vers);
 
 /*
  * Default and maximum payload size (NFS READ or WRITE), in bytes.
- * The default is historical, and the maximum is an implementation
- * limit.
+ * The maximum is an implementation limit.
  */
 enum {
-	NFSSVC_DEFBLKSIZE       = 1 * 1024 * 1024,
+	NFSSVC_DEFBLKSIZE       = 4 * 1024 * 1024,
 	NFSSVC_MAXBLKSIZE       = RPCSVC_MAXPAYLOAD,
 };
 
Signature · 1 line
2.53.0
maintainer Reviewed-by on v1 1/1
#1 lore raw
Quoted 39 lines · Chuck Lever

From: Chuck Lever <[email protected]>

Commit 8a81f16de64f ("NFSD: Add a "default" block size") introduced NFSSVC_DEFBLKSIZE at 1MB, well below the 4MB NFSSVC_MAXBLKSIZE ceiling, with the stated intent that a later change would raise the default.

Raising the default reduces per-RPC overhead on fast networks by amortizing header processing and scheduling costs across larger payloads. The halving loop in nfsd_get_default_max_blksize() constrains the returned value to 1/4096 of available RAM, so the new 4MB default takes effect only on systems with at least 16GB of RAM. Smaller machines continue to receive the same computed value as before. Administrators can still override the computed value through /proc/fs/nfsd/max_block_size.

On systems where the new default takes effect, svc_sock_setbufsize() sizes each service socket's send and receive buffers as nreqs * max_mesg * 2. Quadrupling max_mesg therefore quadruples the per-socket buffer reservation at a fixed thread count, which operators tuning large thread pools should account for.

Note well: Your NFS client implementation must support large read and write size settings to benefit from this change.

---
 fs/nfsd/nfsd.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
fs/nfsd/nfsd.h +0 -0
index a01d70953358..daa63c1b161c 100644
@@ -45,11 +45,10 @@ bool nfsd_support_version(int vers);
 
 /*
  * Default and maximum payload size (NFS READ or WRITE), in bytes.

fs/nfsd/nfsd.h bool nfsd_support_version(int vers); · line 54

Quoted 10 lines
fs/nfsd/nfsd.h +2 -3
- * The default is historical, and the maximum is an implementation
- * limit.
+ * The maximum is an implementation limit.
  */
 enum {
-	NFSSVC_DEFBLKSIZE       = 1 * 1024 * 1024,
+	NFSSVC_DEFBLKSIZE       = 4 * 1024 * 1024,
 	NFSSVC_MAXBLKSIZE       = RPCSVC_MAXPAYLOAD,
 };
 

Makes sense, particularly with dynamic threading.

Reviewed-by on v1 1/1
#2 lore raw
In reply to Chuck Lever
Quoted 39 lines · Chuck Lever

From: Chuck Lever <[email protected]>

Commit 8a81f16de64f ("NFSD: Add a "default" block size") introduced NFSSVC_DEFBLKSIZE at 1MB, well below the 4MB NFSSVC_MAXBLKSIZE ceiling, with the stated intent that a later change would raise the default.

Raising the default reduces per-RPC overhead on fast networks by amortizing header processing and scheduling costs across larger payloads. The halving loop in nfsd_get_default_max_blksize() constrains the returned value to 1/4096 of available RAM, so the new 4MB default takes effect only on systems with at least 16GB of RAM. Smaller machines continue to receive the same computed value as before. Administrators can still override the computed value through /proc/fs/nfsd/max_block_size.

On systems where the new default takes effect, svc_sock_setbufsize() sizes each service socket's send and receive buffers as nreqs * max_mesg * 2. Quadrupling max_mesg therefore quadruples the per-socket buffer reservation at a fixed thread count, which operators tuning large thread pools should account for.

Note well: Your NFS client implementation must support large read and write size settings to benefit from this change.

---
 fs/nfsd/nfsd.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
fs/nfsd/nfsd.h +0 -0
index a01d70953358..daa63c1b161c 100644
@@ -45,11 +45,10 @@ bool nfsd_support_version(int vers);
 
 /*

fs/nfsd/nfsd.h bool nfsd_support_version(int vers); · line 53

Quoted 10 lines
fs/nfsd/nfsd.h +2 -3
  * Default and maximum payload size (NFS READ or WRITE), in bytes.
- * The default is historical, and the maximum is an implementation
- * limit.
+ * The maximum is an implementation limit.
  */
 enum {
-       NFSSVC_DEFBLKSIZE       = 1 * 1024 * 1024,
+       NFSSVC_DEFBLKSIZE       = 4 * 1024 * 1024,
        NFSSVC_MAXBLKSIZE       = RPCSVC_MAXPAYLOAD,
 };

Works with ms-nfs41-client.

----

Bye,
Roland

Signature · 5 lines
  __ .  . __
 (o.\ \/ /.o) [email protected]
  \__\/\/__/  MPEG specialist, C&&JAVA&&Sun&&Unix programmer
  /O /==\ O\  TEL +49 641 3992797
 (;O/ \/ \O;)