fslatency-nd.8

.TH fslatency-nd 8  "2015-01-30" "USER COMMANDS"
.SH NAME
fslatency-nd \- Measure file system sync read and write latency distribution. Uses Linux SystemTap (non-debuginfo).
.SH SYNOPSIS
.B fslatency-nd.stp
[interval [count]]
.SH DESCRIPTION
fslatency-nd.stp dynamically traces two common file system functions:
do_sync_read() and do_sync_write(), and reports a histogram distribution of
latency. Many, but not all, file systems and
workloads use these functions. Tracing their time provides one view of
suffered file system latency.

Other file system requests not covered by these functions and this tool include:
asynchronous I/O, directory operations, and file handle operations.
.SH REQUIREMENTS
SystemTap.
.SH OPTIONS
.TP
interval
Interval for reports, in seconds.
.TP
count
Number of reports to output.
.SH EXAMPLES
.TP
Measure (most) file system synchronous read and write latency, and show a distribution histogram on Ctrl-C:
#
.B fslatency-nd.stp
.TP
Measure (most) file system synchronous read and write latency, and show a report every second, five times:
#
.B fslatency-nd.stp 1 5
.SH FIELDS
.TP
value
The minimum value, in nanoseconds, for this histogram bucket. Read this value
as "I/O latency was at least xxx".
.TP
count
How many reads/writes fell into this latency range.
.SH OVERHEAD
The overheads are measurable, as file system requests can be very frequent, and
the small overhead of tracing each (eg, 4 microseconds per call) can begin to
add up. Best case, it won't be noticeable. Worst case, the application may run
2x slower.
.SH SOURCE
This is from the SystemTap lwtools collection.
.IP
https://github.com/brendangregg/systemtap-lwtools
.PP
Also look under the examples directory for a text file containing example
usage, output, and commentary for this tool.
.SH OS
Linux
.SH STABILITY
Unstable - in development.
.SH AUTHOR
Brendan Gregg
.SH SEE ALSO
iostat(1), fsslower-nd.stp(8), biolatency-nd.stp(8)