All- I have db volumes on netapp filer on NFS .And when ever disk is accessed I see 99% wait on DISK FOR cpu? I tuned /usr/sbin/nfso -o nfs_iopace_pages=64 /usr/sbin/nfso -o nfs_max_threads=2000 ;; /usr/sbin/nfso -o nfs_rfc1323=1 /usr/sbin/nfso -o nfs_socketsize=800000 /usr/sbin/nfso -o nfs_tcp_socketsize=800000 /usr/sbin/nfso -o nfs_use_reserved_ports=1 /usr/sbin/nfso -o udpchecksum=0 /usr/sbin/no -o arptab_nb=40 /usr/sbin/no -o ipforwarding=1 /usr/sbin/no -o ipqmaxlen=1024 /usr/sbin/no -o rfc1323=1 /usr/sbin/no -o sb_max=6000000 /usr/sbin/no -o somaxconn=2048 /usr/sbin/no -o tcp_keepidle=1000 /usr/sbin/no -o tcp_mssdflt=1448 /usr/sbin/no -o tcp_pmtu_discover=0 /usr/sbin/no -o tcp_recvspace=655360 /usr/sbin/no -o tcp_sendspace=655360 /usr/sbin/no -o tcp_ttl=500 /usr/sbin/no -o udp_pmtu_discover=0 /usr/sbin/no -o udp_recvspace=655360 /usr/sbin/no -o udp_sendspace=65536 /usr/sbin/no -o udp_ttl=500
But still in vain.I also increased biods on NFS but no effect. These are Netapp filers.I don't see nay fragmentation on Filer and its on NFS netapp disks through 100mpbs. The network is tuned to work at 100 full duplex. What might have gone wrong please help. I am totally pussled sine when I use cmd to copy data within Netapp disks from volume to volume on a filer all looks great. What is wrong with NFS over 100mbps private network on switch. Mtu is default1500 on both ends.Please guide me I welcome any research you guys want me to do!!! Thanks Balanand Pinni
