Lines Matching +full:performance +full:- +full:affecting
5 :Authors: Greg Banks <gnb@sgi.com> - 26 Mar 2009
13 program from the nfs-utils distribution provides a helpful command-line
48 packets-arrived
58 However this is a more accurate and less workload-dependent measure
62 sockets-enqueued
68 network-facing work to be done but it couldn't be done immediately,
70 rate of change for this counter is zero; significantly non-zero
71 values may indicate a performance limitation.
74 pool for the NFS workload (the workload is thread-limited), in which
76 performance of the NFS workload.
78 threads-woken
83 network-facing NFS work is being handled quickly, which is a good
85 to but less than the rate of change of the packets-arrived counter.
87 threads-timedout
95 affecting performance. Unfortunately, it's only a clue and not
98 - Currently the rate at which the counter is incremented is quite
103 - It is usually a wise policy to provide some slack,
109 one of three ways. An nfsd thread can be woken (threads-woken counts
111 (sockets-enqueued counts this case), or the packet can be temporarily
116 packets-deferred = packets-arrived - ( sockets-enqueued + threads-woken )