1rcapstat(1) User Commands rcapstat(1)
2
3
4
6 rcapstat - report resource cap enforcement daemon statistics
7
9 rcapstat [-g] [-p | -z] [-T u | d ] [interval [count]]
10
11
13 The rcapstat command reports on the projects or zones capped by
14 rcapd(1M). Each report contains statistics that pertain to the project
15 or zone and paging statistics. Paging refers to the act of relocating
16 portions of memory, called pages, to or from physical memory. rcapd
17 pages out the most infrequently used pages.
18
19
20 The paging statistics in the first report issued show the activity
21 since the daemon was started. Subsequent reports reflect the activity
22 since the last report was issued.
23
24
25 Reports are issued every interval seconds up to the quantity specified
26 by count, or forever if count is not specified.
27
29 The following options are supported:
30
31 -g Global statistics. Reports the minimum memory utilization
32 for memory cap enforcement (see rcapadm(1M)) and reports
33 current memory utilization as a percentage of installed
34 physical memory.
35
36
37 -p Report statistics for capped projects. This is the default
38 if no option is specified.
39
40
41 -T u | d Display a time stamp.
42
43 Specify u for a printed representation of the internal rep‐
44 resentation of time. See time(2). Specify d for standard
45 date format. See date(1).
46
47
48 -z Report statistics for capped zones.
49
50
52 The following list defines the column headings in the rcapstat report
53 and provides information about how to interpret the report.
54
55 id The project ID or zone ID of the capped project or zone.
56
57
58 project The project name.
59
60
61 zone The zone name.
62
63
64 nproc The number of processes in the project or zone since the
65 last report.
66
67
68 vm The total virtual memory size of the project or zone's pro‐
69 cesses, including all mapped files and devices, in kilobytes
70 (K), megabytes (M), or gigabytes (G).
71
72
73 rss The total resident set size (RSS) of the project or zone's
74 processes, in kilobytes (K), megabytes (M), or gigabytes
75 (G). The count does not account for shared pages.
76
77
78 cap The RSS cap for the project or zone. See rcapd(1M) for
79 information about how to specify memory caps.
80
81
82 at The total amount of memory that rcapd attempted to page out.
83
84 Paging refers to the act of relocating portions of memory,
85 called pages, to or from physical memory. rcapd pages out
86 the most infrequently used pages.
87
88
89 avgat The average amount of memory that rcapd attempted to page
90 out during each sample cycle. The rate at which rcapd sam‐
91 ples RSS can be set with rcapadm(1M).
92
93
94 pg An estimate of the total amount of memory that rcapd suc‐
95 cessfully paged out.
96
97
98 avgpg An estimate of the average amount of memory that rcapd suc‐
99 cessfully paged out during each sample cycle. The rate at
100 which rcapd samples process RSS sizes can be set with rca‐
101 padm.
102
103
105 The following operands are supported:
106
107 interval Specifies the reporting interval in seconds. The default
108 interval is 5 seconds.
109
110
111 count Specifies the number of reports to produce. By default,
112 rcapstat reports statistics until a termination signal is
113 received or until the rcapd process exits.
114
115
117 Example 1 Using rcapstat to Report Cap and Project Information
118
119
120 Caps are defined for two projects associated with two users. user1 has
121 a cap of 50 megabytes and user2 has a cap of 10 megabytes.
122
123
124
125 The following command produces five reports at 5-second sampling inter‐
126 vals.
127
128
129 example# rcapstat 5 5
130 id project nproc vm rss cap at avgat pg avgpg
131 112270 user1 24 123M 35M 50M 50M 0K 3312K 0K
132 78194 user2 1 2368K 1856K 10M 0K 0K 0K 0K
133 id project nproc vm rss cap at avgat pg avgpg
134 112270 user1 24 123M 35M 50M 0K 0K 0K 0K
135 78194 user2 1 2368K 1856K 10M 0K 0K 0K 0K
136 id project nproc vm rss cap at avgat pg avgpg
137 112270 user1 24 123M 35M 50M 0K 0K 0K 0K
138 78194 user2 1 2368K 1928K 10M 0K 0K 0K 0K
139 id project nproc vm rss cap at avgat pg avgpg
140 112270 user1 24 123M 35M 50M 0K 0K 0K 0K
141 78194 user2 1 2368K 1928K 10M 0K 0K 0K 0K
142 id project nproc vm rss cap at avgat pg avgpg
143 112270 user1 24 123M 35M 50M 0K 0K 0K 0K
144 78194 user2 1 2368K 1928K 10M 0K 0K 0K 0K
145
146
147
148
149 The first three lines of output constitute the first report, which con‐
150 tains the cap and project information for the two projects and paging
151 statistics since rcapd was started. The at and pg columns are a number
152 greater than zero for user1 and zero for user2, which indicates that at
153 some time in the daemon's history, user1 exceeded its cap but user2 did
154 not.
155
156
157
158 The subsequent reports show no significant activity.
159
160
161 Example 2 Using rcapstat to Monitor the RSS of a Project
162
163 example% rcapstat 5 5
164 id project nproc vm rss cap at avgat pg avgpg
165 376565 user1 57 209M 46M 10M 440M 220M 5528K 2764K
166 376565 user1 57 209M 44M 10M 394M 131M 4912K 1637K
167 376565 user1 56 207M 43M 10M 440M 147M 6048K 2016K
168 376565 user1 56 207M 42M 10M 522M 174M 4368K 1456K
169 376565 user1 56 207M 44M 10M 482M 161M 3376K 1125K
170
171
172
173
174 The project user1 has an RSS in excess of its physical memory cap. The
175 nonzero values in the pg column indicate that rcapd is consistently
176 paging out memory as it attempts to meet the cap by lowering the physi‐
177 cal memory utilization of the project's processes. However, rcapd is
178 unsuccessful, as indicated by the varying rss values that do not show a
179 corresponding decrease. This means that the application's resident mem‐
180 ory is being actively used, forcing rcapd to affect the working set.
181 Under this condition, the system continues to experience high page
182 fault rates, and associated I/O, until the working set size (WSS) is
183 reduced, the cap is raised, or the application changes its memory
184 access pattern. Notice that a page fault occurs when either a new page
185 must be created, or the system must copy in a page from the swap
186 device.
187
188
189 Example 3 Determining the Working Set Size of a Project
190
191
192 This example is a continuation of Example 1, and it uses the same
193 project.
194
195
196 example% rcapstat 5 5
197 id project nproc vm rss cap at avgat pg avgpg
198 376565 user1 56 207M 44M 10M 381M 191M 15M 7924K
199 376565 user1 56 207M 46M 10M 479M 160M 2696K 898K
200 376565 user1 56 207M 46M 10M 424M 141M 7280K 2426K
201 376565 user1 56 207M 43M 10M 401M 201M 4808K 2404K
202 376565 user1 56 207M 43M 10M 456M 152M 4800K 1600K
203 376565 user1 56 207M 44M 10M 486M 162M 4064K 1354K
204 376565 user1 56 207M 52M 100M 191M 95M 1944K 972K
205 376565 user1 56 207M 55M 100M 0K 0K 0K 0K
206 376565 user1 56 207M 56M 100M 0K 0K 0K 0K
207 376565 user1 56 207M 56M 100M 0K 0K 0K 0K
208 376565 user1 56 207M 56M 100M 0K 0K 0K 0K
209 376565 user1 56 207M 56M 100M 0K 0K 0K 0K
210
211
212
213
214 By inhibiting cap enforcement, either by raising the cap of a project
215 or by changing the minimum physical memory utilization for cap
216 enforcement (see rcapadm(1M)), the resident set can become the working
217 set. The rss column might stabilize to show the project WSS, as shown
218 in the previous example. The WSS is the minimum cap value that allows
219 the project's processes to operate without perpetually incurring page
220 faults.
221
222
224 The following exit values are returned:
225
226 0 Successful completion.
227
228
229 1 An error occurred.
230
231
232 2 Invalid command-line options were specified.
233
234
236 See attributes(5) for descriptions of the following attributes:
237
238
239
240
241 ┌─────────────────────────────┬─────────────────────────────┐
242 │ ATTRIBUTE TYPE │ ATTRIBUTE VALUE │
243 ├─────────────────────────────┼─────────────────────────────┤
244 │Availability │SUNWrcapu │
245 └─────────────────────────────┴─────────────────────────────┘
246
248 rcapadm(1M), rcapd(1M), attributes(5)
249
250
251 Physical Memory Control Using the Resource Capping Daemon in System
252 Administration Guide: Resource Management
253
255 If the interval specified to rcapstat is shorter than the reporting
256 interval specified to rcapd (with rcapadm(1M)), the output for some
257 intervals can be zero. This is because rcapd does not update statistics
258 more frequently than the interval specified with rcapadm, and this
259 interval is independent of (and less precise than) the sampling inter‐
260 val used by rcapstat.
261
262
263
264SunOS 5.11 16 Jun 2009 rcapstat(1)