diff options
| author | Santosh Kumar Pradhan <spradhan@redhat.com> | 2013-09-18 14:43:40 +0530 |
|---|---|---|
| committer | Anand Avati <avati@redhat.com> | 2013-09-20 11:29:34 -0700 |
| commit | e2093fb1500f55f58236d37a996609a2a1e1af8e (patch) | |
| tree | bcd8da1f331e0447554ab06513751f655a39a507 /xlators/nfs/server/src/nfs3.h | |
| parent | 90ec0917558e6836d98faa4406b96158b13eacee (diff) | |
gNFS: NFS daemon is limiting IOs to 64KB
Problem:
Gluster NFS server is hard-coding the max rsize/wsize to 64KB
which is very less for NFS running over 10GE NIC. The existing
options nfs.read-size, nfs.write-size are not working as
expected.
FIX:
Make the options nfs.read-size (for rsize) and nfs.write-size
(for wsize) work to tune the NFS I/O size. Value range would
be 4KB(Min)-64KB(Default)-1MB(max).
NB: Credit to "Richard Wareing" for catching it.
Change-Id: I2754ecb0975692304308be8bcf496c713355f1c8
BUG: 1009223
Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com>
Reviewed-on: http://review.gluster.org/5964
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Kaleb KEITHLEY <kkeithle@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Diffstat (limited to 'xlators/nfs/server/src/nfs3.h')
| -rw-r--r-- | xlators/nfs/server/src/nfs3.h | 40 |
1 files changed, 30 insertions, 10 deletions
diff --git a/xlators/nfs/server/src/nfs3.h b/xlators/nfs/server/src/nfs3.h index 6b37cd458bc..19353abea08 100644 --- a/xlators/nfs/server/src/nfs3.h +++ b/xlators/nfs/server/src/nfs3.h @@ -48,16 +48,36 @@ /* Static values used for FSINFO -FIXME: This should be configurable */ -#define GF_NFS3_RTMAX (64 * GF_UNIT_KB) -#define GF_NFS3_RTPREF (64 * GF_UNIT_KB) -#define GF_NFS3_RTMULT (4 * GF_UNIT_KB) -#define GF_NFS3_WTMAX (64 * GF_UNIT_KB) -#define GF_NFS3_WTPREF (64 * GF_UNIT_KB) -#define GF_NFS3_WTMULT (4 * GF_UNIT_KB) -#define GF_NFS3_DTMIN (4 * GF_UNIT_KB) -#define GF_NFS3_DTPREF (64 * GF_UNIT_KB) -#define GF_NFS3_MAXFILE (1 * GF_UNIT_PB) + * To change the maximum rsize and wsize supported by the NFS client, + * adjust GF_NFS3_FILE_IO_SIZE_MAX. The Gluster NFS server defaults to + * 64KB(65536) which is reasonable for 1GE and 10GE. But for better + * performance on 10GE, rsize/wsize can be raised to 1MB(1048576). + * rsize and wsize can be tuned through nfs.read-size and + * nfs.write-size respectively. + * + * NB: For Kernel-NFS, NFS_MAX_FILE_IO_SIZE is 1048576U (1MB). + */ +#define GF_NFS3_FILE_IO_SIZE_MAX (1 * GF_UNIT_MB) /* 1048576 */ +#define GF_NFS3_FILE_IO_SIZE_DEF (64 * GF_UNIT_KB) /* 65536 */ +#define GF_NFS3_FILE_IO_SIZE_MIN (4 * GF_UNIT_KB) /* 4096 */ + +#define GF_NFS3_RTMAX GF_NFS3_FILE_IO_SIZE_MAX +#define GF_NFS3_RTMIN GF_NFS3_FILE_IO_SIZE_MIN +#define GF_NFS3_RTPREF GF_NFS3_FILE_IO_SIZE_DEF +#define GF_NFS3_RTMULT GF_NFS3_FILE_IO_SIZE_MIN + +#define GF_NFS3_WTMAX GF_NFS3_FILE_IO_SIZE_MAX +#define GF_NFS3_WTMIN GF_NFS3_FILE_IO_SIZE_MIN +#define GF_NFS3_WTPREF GF_NFS3_FILE_IO_SIZE_DEF +#define GF_NFS3_WTMULT GF_NFS3_FILE_IO_SIZE_MIN + +/* This can be tuned through nfs.readdir-size */ +#define GF_NFS3_DTMAX GF_NFS3_FILE_IO_SIZE_MAX +#define GF_NFS3_DTMIN GF_NFS3_FILE_IO_SIZE_MIN +#define GF_NFS3_DTPREF GF_NFS3_FILE_IO_SIZE_DEF + +#define GF_NFS3_MAXFILESIZE (1 * GF_UNIT_PB) + /* FIXME: Handle time resolutions */ #define GF_NFS3_TIMEDELTA_SECS {1,0} #define GF_NFS3_TIMEDELTA_NSECS {0,1} |