.\" snlatency.8: auto-generated, DO NOT EDIT
.\"
.\" Copyright 2011-2018. Quantum Corporation. All Rights Reserved.
.\" StorNext is either a trademark or registered trademark of
.\" Quantum Corporation in the US and/or other countries.
.\"
.\" Code start macro
.de Cs
.sp
.ft C
.in +0.3i
.nf
..
.\" Code end macro
.de Ce
.fi
.in -0.3i
.ft R
..
.TH SNLATENCY 8 "January 2018" "Xsan File System"
.SH NAME
snlatency \- Measure Xsan Volume network latency
.SH SYNOPSIS
.na
.nh
.HP
.B snlatency
.I pathname
.RI [ repeat ]
.ad
.hy
.SH DESCRIPTION
The \fBsnlatency\fR program allows measuring the round trip latency from the
Xsan client to  the FSM process on the MDC host.
The file system to measure is selected via the \fIpathname\fR parameter.
.PP
\fBsnlatency\fR uses a special RPC to the MDC in a loop to measure
the time to send the message, the time for the FSM process to service
the request,
and the time to send the response back up to the system call level.
The average of \fIrepeat\fR calls is averaged and the latencies printed
out in micro seconds.
.PP
If the MDC is not local, then the NTP clock algorithm is used to adjust for
the difference between the two host's clocks before calculating the results,
it is not used when the MDC the same host.
The reason for this is that the NTP algorithm assumes the latency in both
directions is the same and in effect averages them out.
Comparing the latencies seen between the client local to the MDC and remote
clients will show how much of the latency is caused by network infrastructure.
.SH EXIT VALUES
\fBsnlatency\fR will return 0 on success and non-zero on failure.