1PMVAL(1) General Commands Manual PMVAL(1)
2
3
4
6 pmval, pmevent - arbitrary performance metrics value dumper
7
9 pmval [-dgLrvVz?] [-a archive] [-A align] [-f N] [-h host] [-i
10 instances] [-K spec] [-n pmnsfile] [-O offset] [-p port] [-s samples]
11 [-S starttime] [-t interval] [-T endtime] [-U archive] [-w width] [-x
12 pattern] [-Z timezone] [--container=name] [--derived=file] metricname
13
14 pmevent ...
15
17 pmval prints current or archived values for the nominated performance
18 metric. The metric of interest is named in the metricname argument,
19 subject to instance qualification with the -i flag as described below.
20
21 Unless directed to another host by the -h option, or to a set of ar‐
22 chives by the -a or -U options, pmval will contact the Performance Met‐
23 rics Collector Daemon (PMCD) on the local host to obtain the required
24 information.
25
26 The metricname argument may also be given in the metric specification
27 syntax, as described in PCPIntro(1), where the source, metric and
28 instance may all be included in the metricname, e.g. thathost:ker‐
29 nel.all.load["1 minute"]. When this format is used, none of the -h or
30 -a or -U options may be specified.
31
32 When using the metric specification syntax, the ``hostname'' @ is
33 treated specially and causes pmval to use a local context to collect
34 metrics from PMDAs on the local host without PMCD. Only some metrics
35 are available in this mode.
36
37 When processing a set of archives, pmval may relinquish its own timing
38 control, and operate under the control of a a pmtime(1) process that
39 uses a GUI dialog to provide timing control. In this case, either the
40 -g option should be used to start pmval as the sole client of a new
41 pmtime(1) instance, or -p should be used to attach pmval to an existing
42 pmtime(1) instance via the IPC channel identified by the port argument.
43
44 The -S, -T, -O and -A options may be used to define a time window to
45 restrict the samples retrieved, set an initial origin within the time
46 window, or specify a ``natural'' alignment of the sample times; refer
47 to PCPIntro(1) for a complete description of these options.
48
49 The output from pmval is directed to standard output. The following
50 symbols may occasionally appear, in place of a metric value, in pmval
51 output: A question mark symbol (?) indicates that a value is no longer
52 available for that metric instance. An exclamation mark (!) indicates
53 that a 64-bit counter wrapped during the sample.
54
55 pmevent is an alias for pmval.
56
58 The available command line options are:
59
60 -a archive, --archive=archive
61 Performance metric values are retrieved from the set of Perfor‐
62 mance Co-Pilot (PCP) archive log files identified by the archive
63 argument, which is a comma-separated list of names, each of which
64 may be the base name of an archive or the name of a directory con‐
65 taining one or more archives. See also -U.
66
67 -A align, --align=align
68 Force the initial sample to be aligned on the boundary of a natu‐
69 ral time unit align. Refer to PCPIntro(1) for a complete descrip‐
70 tion of the syntax for align.
71
72 -d, --delay
73 When replaying from an archive, this option requests that the pre‐
74 vailing real-time delay be applied between samples (see -t) to
75 effect a pause, rather than the default behaviour of replaying at
76 full speed.
77
78 -f precision, --precision=precision
79 Numbers are reported in ``fixed point'' notation, rather than the
80 default scientific notation, using precision digits for precision.
81 Each number will be up to the column width determined by the
82 default heuristics, else the -w option if specified, and include
83 precision digits after the decimal point. So, the options -f 3 -w
84 8 would produce numbers of the form 9999.999. A value of zero for
85 precision omits the decimal point and any fractional digits.
86
87 -g, --guimode
88 Start pmval as the sole client of a new pmtime(1) server process
89 for replay of archived performance data using the pmtime(1) graph‐
90 ical user interface.
91
92 -h host, --host=host
93 Fetch performance metrics from pmcd(1) on host, rather than from
94 the default localhost.
95
96 -i instances, --instances=instances
97 Specify a list of one or more names of instances for the nominated
98 performance metric - just these instances will be retrieved and
99 reported (the default is to report all instances). The list must
100 be a single argument, with elements of the list separated by com‐
101 mas and/or white space.
102
103 The instance name may be quoted with single (') or double (")
104 quotes for those cases where the instance name contains white
105 space or commas.
106
107 Multiple -i options are allowed as an alternative way of specify‐
108 ing more than one instance of interest.
109
110 As an example, the following are all equivalent:
111
112 $ pmval -i "'1 minute','5 minute'" kernel.all.load
113 $ pmval -i '"1 minute","5 minute"' kernel.all.load
114 $ pmval -i "'1 minute' '5 minute'" kernel.all.load
115 $ pmval -i "'1 minute'" -i "'5 minute'" kernel.all.load
116 $ pmval 'localhost:kernel.all.load["1 minute","5 minute"]'
117
118
119 -K spec, --spec-local=spec
120 When fetching metrics from a local context (see -L), the -K option
121 may be used to control the DSO PMDAs that should be made accessi‐
122 ble. The spec argument conforms to the syntax described in
123 pmSpecLocalPMDA(3). More than one -K option may be used.
124
125 -L, --local-PMDA
126 Use a local context to collect metrics from DSO PMDAs on the local
127 host without PMCD. See also -K.
128
129 -n pmnsfile, --namespace=pmnsfile
130 Load an alternative Performance Metrics Name Space (PMNS(5)) from
131 the file pmnsfile.
132
133 -O origin, --origin=origin
134 When reporting archived metrics, start reporting at origin within
135 the time window (see -S and -T). Refer to PCPIntro(1) for a com‐
136 plete description of the syntax for origin.
137
138 -p port, --guiport=port
139 Attach pmval to an existing pmtime(1) time control process
140 instance via the IPC channel identified by the port argument.
141 This option is normally only used by other tools, e.g.
142 pmchart(1), when they launch pmval with synchronized time control.
143
144 -r, --raw
145 Print raw values for cumulative counter metrics. Normally cumula‐
146 tive counter metrics are converted to rates. For example, disk
147 transfers are reported as number of disk transfers per second dur‐
148 ing the preceding sample interval, rather than the raw value of
149 number of disk transfers since the machine was booted. If you
150 specify this option, the raw metric values are printed.
151
152 -s samples, --samples=samples
153 The samples argument defines the number of samples to be retrieved
154 and reported. If samples is 0 or -s is not specified, pmval will
155 sample and report continuously (in real time mode) or until the
156 end of the set of PCP archives (in archive mode).
157
158 -S starttime, --start=starttime
159 When reporting archived metrics, the report will be restricted to
160 those records logged at or after starttime. Refer to PCPIntro(1)
161 for a complete description of the syntax for starttime.
162
163 -t interval, --interval=interval
164 Set the reporting interval to something other than the default 1
165 second. The interval argument follows the syntax described in
166 PCPIntro(1), and in the simplest form may be an unsigned integer
167 (the implied units in this case are seconds).
168
169 -T endtime, --finish=endtime
170 When reporting archived metrics, the report will be restricted to
171 those records logged before or at endtime. Refer to PCPIntro(1)
172 for a complete description of the syntax for endtime.
173
174 -U archive, --nointerp=archive
175 Performance metric values are retrieved from the Performance Co-
176 Pilot (PCP) archive. The argument is a comma-separated list of
177 names, each of which may be the base name of an archive or the
178 name of a directory containing one or more archives. However,
179 unlike -a every recorded value in the archive for the selected
180 metric and instances is reported (so no interpolation mode, and
181 the sample interval (-t option) is ignored. See also -a.
182
183 At most one of the options -a and -U may be specified.
184
185 -v, --verbose
186 Enable verbose mode.
187
188 -V, --version
189 Display version number and exit.
190
191 -w width, --width=width
192 Set the width of each column of output to be width columns. If
193 not specified columns are wide enough to accommodate the largest
194 value of the type being printed.
195
196 -x pattern, --filter=pattern
197 The given pattern is sent to the performance metric domain agent
198 for the requested metricname before any values are requested.
199 This serves two purposes. Firstly, it provides a mechanism for
200 server-side event filtering that is customisable for individual
201 event streams. In addition, some performance metrics domain
202 agents also use the PMCD store mechanism to provide a basic secu‐
203 rity model (e.g. for sensitive log files, only a client host with
204 pmStore(3) access would be able to access the event stream).
205
206 As pattern may be processed by regcomp(3) it should be a non-empty
207 string. Use . (dot) for a “match all” pattern.
208
209 -z, --hostzone
210 Use the local timezone of the host that is the source of the per‐
211 formance metrics, as identified by either the -h or the -a or the
212 -U options. The default is to use the timezone of the local host.
213
214 -Z timezone, --timezone=timezone
215 Use timezone for the date and time. Timezone is in the format of
216 the environment variable TZ as described in environ(7).
217
218 -?, --help
219 Display usage message and exit.
220
221 --container=container
222 Specify an individual container to be queried.
223
224 --derived=file
225 Load derived metric definitions from file.
226
228 By default, pmval attempts to display non-integer numeric values in a
229 way that does not distort the inherent precision (rarely more than 4
230 significant digits), and tries to maintain a tabular format in the out‐
231 put. These goals are sometimes in conflict.
232
233 In the absence of the -f option (described above), the following table
234 describes the formats used for different ranges of numeric values for
235 any metric that is of type PM_TYPE_FLOAT or PM_TYPE_DOUBLE, or any met‐
236 ric that has the semantics of a counter (for which pmval reports the
237 rate converted value):
238
239 ┌──────────┬──────────────────────┐
240 │ Format │ Value Range │
241 ├──────────┼──────────────────────┤
242 │ ! │ No values available │
243 │9.999E-99 │ < 0.1 │
244 │ 0.0 │ 0 │
245 │ 9.9999 │ > 0 and <= 0.9999 │
246 │ 9.999 │ > 0.9999 and < 9.999 │
247 │ 99.99 │ > 9.999 and < 99.99 │
248 │ 999.9 │ > 99.99 and < 999.9 │
249 │9999. │ > 999.9 and < 9999 │
250 │9.999E+99 │ > 9999 │
251 └──────────┴──────────────────────┘
253 All are generated on standard error and are intended to be self-
254 explanatory.
255
257 Environment variables with the prefix PCP_ are used to parameterize the
258 file and directory names used by PCP. On each installation, the file
259 /etc/pcp.conf contains the local values for these variables. The
260 $PCP_CONF variable may be used to specify an alternative configuration
261 file, as described in pcp.conf(5).
262
263 For environment variables affecting PCP tools, see pmGetOptions(3).
264
266 PCPIntro(1), pmcd(1), pmchart(1), pmdumplog(1), pmdumptext(1),
267 pminfo(1), pmlogger(1), pmrep(1), pmtime(1), PMAPI(3), pmStore(3),
268 pmSpecLocalPMDA(3), pcp.conf(5), pcp.env(5) and PMNS(5).
269
270
271
272Performance Co-Pilot PCP PMVAL(1)