1
diff -ruN numactl-2.0.8-rc5.orig/numastat.8 numactl-2.0.8-rc5.new/numastat.8
2
--- numactl-2.0.8-rc5.orig/numastat.8 2012-08-23 15:50:37.000000000 -0400
3
+++ numactl-2.0.8-rc5.new/numastat.8 2012-10-07 00:05:46.676484265 -0400
6
-.\" Copyright 2004 Andi Kleen, SuSE Labs.
8
-.\" Permission is granted to make and distribute verbatim copies of this
9
-.\" manual provided the copyright notice and this permission notice are
10
-.\" preserved on all copies.
12
-.\" Permission is granted to copy and distribute modified versions of this
13
-.\" manual under the conditions for verbatim copying, provided that the
14
-.\" entire resulting derived work is distributed under the terms of a
15
-.\" permission notice identical to this one.
17
-.\" Since the Linux kernel and libraries are constantly changing, this
18
-.\" manual page may be incorrect or out-of-date. The author(s) assume no
19
-.\" responsibility for errors or omissions, or for damages resulting from
20
-.\" the use of the information contained herein.
22
-.\" Formatted or processed versions of this manual, if unaccompanied by
23
-.\" the source, must acknowledge the copyright and authors of this work.
24
-.TH NUMACTL 8 "Nov 2004" "SuSE Labs" "Linux Administrator's Manual"
26
-numastat \- Print statistics about NUMA memory allocation
30
+.TH "numastat" "8" "1.0.0" "Bill Gray" "Administration"
33
+\fBnumastat\fP \- Show per-NUMA-node memory statistics for processes and the operating system
39
+\fBnumastat\fP [\fI\-V\fP]
42
+\fBnumastat\fP [\fI\<PID>|<pattern>...\fP]
45
+\fBnumastat\fP [\fI\-c\fP] [\fI\-m\fP] [\fI\-n\fP] [\fI\-p <PID>|<pattern>\fP] [\fI\-s[<node>]\fP] [\fI\-v\fP] [\fI\-z\fP] [\fI\<PID>|<pattern>...\fP]
50
-displays NUMA allocations statistics from the kernel memory allocator.
51
-Each process has NUMA policies that specifies on which node pages
56
-on details of the available policies.
57
-The numastat counters keep track on what nodes memory is finally allocated.
59
-The counters are separated for each node. Each count event is the allocation
62
+with no command options or arguments at all, displays per-node NUMA hit and
63
+miss system statistics from the kernel memory allocator. This default
64
+\fBnumastat\fP behavior is strictly compatible with the previous long-standing
65
+\fBnumastat\fP perl script, written by Andi Kleen. The default \fBnumastat\fP
66
+statistics shows per-node numbers (in units of pages of memory) in these categories:
69
-is the number of allocations where an allocation was intended for
70
-that node and succeeded there.
72
+is memory successfully allocated on this node as intended.
75
-shows how often an allocation was intended for this node, but ended up
76
-on another node due to low memory.
78
+is memory allocated on this node despite the process preferring some different node. Each
85
-is the number of allocations that were intended for another node,
86
-but ended up on this node. Each
87
+is memory intended for this node, but actually allocated on some different node. Each
96
-is the number of interleave policy allocations that were intended for a
97
-specific node and succeeded there.
99
+is interleaved memory successfully allocated on this node as intended.
102
-is incremented when a process running on the node allocated
103
-memory on the same node.
105
+is memory allocated on this node while a process was running on it.
108
-is incremented when a process running on another node allocated memory on that node.
113
+is memory allocated on this node while a process was running on some other node.
115
+Any supplied options or arguments with the \fBnumastat\fP command will
116
+significantly change both the content and the format of the display. Specified
117
+options will cause display units to change to megabytes of memory, and will
118
+change other specific behaviors of \fBnumastat\fP as described below.
123
+Minimize table display width by dynamically shrinking column widths based on
124
+data contents. With this option, amounts of memory will be rounded to the
125
+nearest megabyte (rather than the usual display with two decimal places).
126
+Column width and inter-column spacing will be somewhat unpredictable with this
127
+option, but the more dense display will be very useful on systems with many
131
+Show the meminfo-like system-wide memory usage information. This option
132
+produces a per-node breakdown of memory usage information similar to that found
136
+Show the original \fBnumastat\fP statistics info. This will show the same
137
+information as the default \fBnumastat\fP behavior but the units will be megabytes of
138
+memory, and there will be other formatting and layout changes versus the
139
+original \fBnumastat\fP behavior.
141
+\fB\-p\fR <\fBPID\fP> or <\fBpattern\fP>
142
+Show per-node memory allocation information for the specified PID or pattern.
143
+If the \-p argument is only digits, it is assumed to be a numerical PID. If
144
+the argument characters are not only digits, it is assumed to be a text
145
+fragment pattern to search for in process command lines. For example,
146
+\fBnumastat -p qemu\fP will attempt to find and show information for processes
147
+with "qemu" in the command line. Any command line arguments remaining after
148
+\fBnumastat\fP option flag processing is completed, are assumed to be
149
+additional <\fBPID\fP> or <\fBpattern\fP> process specifiers. In this sense,
150
+the \fB\-p\fP option flag is optional: \fBnumastat qemu\fP is equivalent to
151
+\fBnumastat -p qemu\fP
154
+Sort the table data in descending order before displaying it, so the biggest
155
+memory consumers are listed first. With no specified <node>, the table will be
156
+sorted by the total column. If the optional <node> argument is supplied, the
157
+data will be sorted by the <node> column. Note that <node> must follow the
158
+\fB\-s\fP immediately with no intermediate white space (e.g., \fBnumastat
162
+Make some reports more verbose. In particular, process information for
163
+multiple processes will display detailed information for each process.
164
+Normally when per-node information for multiple processes is displayed, only
165
+the total lines are shown.
168
+Display \fBnumastat\fP version information and exit.
171
+Skip display of table rows and columns of only zero valuess. This can be used
172
+to greatly reduce the amount of uninteresting zero data on systems with many
173
+NUMA nodes. Note that when rows or columns of zeros are still displayed with
174
+this option, that probably means there is at least one value in the row or
175
+column that is actually non-zero, but rounded to zero for display.
177
-numastat output is only available on NUMA systems.
179
-numastat assumes the output terminal has a width of 80 characters
180
-and tries to format the output accordingly.
182
-.I watch -n1 numastat
183
+\fBnumastat\fP attempts to fold each table display so it will be conveniently
184
+readable on the output terminal. Normally a terminal width of 80 characters is
185
+assumed. When the \fBresize\fP command is available, \fBnumastat\fP attempts
186
+to dynamically determine and fine tune the output tty width from \fBresize\fP
187
+output. If \fBnumastat\fP output is not to a tty, very long output lines can
188
+be produced, depending on how many NUMA nodes are present. In all cases,
189
+output width can be explicitly specified via the \fBNUMASTAT_WIDTH\fP
190
+environment variable. For example, \fBNUMASTAT_WIDTH=100 numastat\fP. On
191
+systems with many NUMA nodes, \fBnumastat \-c \-z ....\fP can be very helpful
192
+to selectively reduce the amount of displayed information.
193
+.SH "ENVIRONMENT VARIABLES"
199
+\fI/proc/*/numa_maps\fP
201
+\fI/sys/devices/system/node/node*/meminfo\fP
203
+\fI/sys/devices/system/node/node*/numastat\fP
205
+.I numastat \-c \-z \-m \-n
207
+.I numastat \-czs libvirt kvm qemu
209
+.I watch \-n1 numastat
211
-.I watch -n1 --differences=accumulative numastat
213
-/sys/devices/system/node/node*/numastat
215
-The output formatting on machines with a large number of nodes
217
+.I watch \-n1 \-\-differences=cumulative numastat
220
+The original \fBnumastat\fP perl script was written circa 2003 by Andi Kleen
221
+<andi.kleen@intel.com>. The current \fBnumastat\fP program was written in 2012
222
+by Bill Gray <bgray@redhat.com> to be compatible by default with the original,
223
+and to add options to display per-node system memory usage and per-node process
228
+.BR set_mempolicy( 2),