1AMSVIS(1) Linux on Power Service Tools AMSVIS(1)
2
6 amsvis - view Active Memory Sharing (AMS) metrics in a GUI locally or
7 remotely.
8
10 /usr/bin/amsvis -l [-u SEC] [-v]
11 /usr/bin/amsvis -r [-c] [--hosts HOSTNAME[:PORT][,...]] [-u SEC] [-v]
13 /usr/bin/amsvis -r -s [-p PORT] [-v]
15 /usr/bin/amsvis -h
17
19 This program helps visualize operating system data as it relates to
20 running within an Active Memory Sharing (AMS) environment. The tool
21 can be run both locally or remotely. Additionally, it can be run as a
22 local data server, without a graphical user interface (GUI), to provide
23 data for running the GUI remotely.
24 When the GUI is run for a single host a number of panels will be dis‐
26 played for the operating system, these are the System, IO Bus Memory,
27 and IO Devices panels.
28 When the GUI is run for multiple hosts only the System Panel will be
30 displayed for each of the hosts. This is achieved by specifying multi‐
31 ple systems on the command-line via the --hosts option.
32 System Panel
34 In an AMS environment, the amount of memory available will change as
35 memory is loaned to other partitions by the Cooperative Memory Manager
36 (CMM) module. This panel will show how the total amount of memory pro‐
37 vided to the system is being utilized. The graph displays a histogram
38 showing the loaned, used, and free memory; the most recent value is
39 displayed to the right.
40 For the system memory panel there are six values displays beneath the
42 graph, they are:
43 Total This value, expressed in kilobytes, is the maximum amount of
45 memory available to the operating system. This value represents
46 the graph's total height.
47 Loaned This is the amount of memory, in kilobytes, that the operating
49 system has temporarily given back to the platform firmware to be
50 used by other partitions. This value fluctuates to meet the
51 demands of all of the partitions in the shared memory pool of an
52 AMS environment.
53 Used This value, displayed in kilobytes, is the amount of system mem‐
55 ory in-use, not including buffers or cache.
56 Free This value, displayed in kilobytes, is the amount of system mem‐
58 ory which is unused.
59 Faults A fault occurs when an attempt to access memory is made where
61 the platform firmware must suspend the partition and request
62 data from disk. This value is a sum of the number of faults
63 since the system was booted. Increases in this value indicate
64 contention for memory between partitions in the AMS environment.
65 Fault time
67 This value is a running sum of the amount of time, since boot
68 time, that the operating system has been suspended by platform
69 firmware to process faults.
70 IO Bus Memory Panel
72 This panel shows how the memory given to the operating system for IO
73 operations is being used by the bus and the devices present on the bus.
74 The graph will display a histogram for the spare and in use memory with
76 the most recent value on the right side of the graph. The graph is
77 overlayed with a dashed line corresponding in color to the high water
78 value and another dashed line for the reserve pool.
79 For the system's IO bus memory panel there are eight values displayed
81 beneath the graph, they are:
82 Entitled
84 The entitled memory is a specific amount of memory given to the
85 operating system that will be available to be mapped for IO
86 operations; this guarantee is maintained by the platform
87 firmware and can be tuned through the HMC or IVM depending on
88 the particular environment. This value is displayed in kilo‐
89 bytes. The amount of entitled memory is changed by the HMC or
90 IVM in response to hot-plug events or explicit tuning. This
91 value represents the graph's total height.
92 Desired
94 Each device in the system that requires memory for mapping IO
95 operations will specify an amount of memory which is desired for
96 optimal performance. This value is the sum of the requests made
97 by individual devices on the bus and is measured in kilobytes.
98 High water
100 This value, measured in kilobytes, represents the largest amount
101 of memory mapped at any one point in time since the system was
102 booted or the field was last reset. This value can be reset by
103 writing a '0' to the file /sys/bus/vio/cmo_high.
104 In use This value represents the amount of memory, in kilobytes, that
106 is currently mapped by device drivers for IO operations.
107 Spare This value represents the amount of memory, in kilobytes, kept
109 unused to handle hot-plug events, entitlement rebalancing
110 between devices, and module loading.
111 Reserve pool
113 The entitled memory is split into two pools, the first of which
114 is the reserve pool. This value, expressed in kilobytes, repre‐
115 sents the amount of memory guaranteed to individual devices on
116 the bus.
117 Excess pool
119 This pool contains any amount of entitled memory not placed in
120 the reserve pool; the value is displayed in kilobytes. Any
121 device can map memory from this pool for IO operations, after
122 having used up its own allocation from the reserve pool, until
123 the excess pool is exhausted.
124 Excess free
126 This value represents the amount of memory, in kilobytes, cur‐
127 rently available for devices to use from the excess pool.
128 IO Devices Panel
130 There will be one panel for each VIO device present in the system that
131 maps memory for IO operations. Panels will be added and removed for
132 devices as a result of DLPAR hot-plug operations and module load/unload
133 events.
134 The graph will display a histogram for the values of allocated memory,
136 with the most recent value on the right side of the graph. The graph
137 is overlaid with a dashed line that represents the latest value of
138 entitled memory. The area of the graph above the dashed line repre‐
139 sents the excess Pool of IO bus memory, any portion of this area that
140 is grayed out has been mapped by other devices and is currently
141 unavailable to the particular device.
142 For each device panel there are five values displayed beneath the
144 graph, they are:
145 Device Name
147 The name displayed for the device is its location on the bus.
148 Entitled
150 The amount of memory, in kilobytes, that the device is guaran‐
151 teed that it may map for IO operations. The bus will guarantee
152 that the device can map this number of kilobytes of memory for
153 IO operations. The allocation of this entitled memory is made
154 to the device from the bus' reserve pool.
155 Desired
157 The amount of memory, in kilobytes, that the device has
158 requested from the bus to provide optimal performance. The
159 amount of entitled memory will not exceed this amount. The
160 device may receive a memory entitlement less than its desired
161 level in cases where resources are limited.
162 Allocated
164 The amount of memory, in kilobytes, that the device has cur‐
165 rently mapped for IO operations.
166 Allocs failed
168 When the amount of memory allocated has exhausted both the enti‐
169 tled memory and the bus' excess pool, memory mapping failures
170 will occur. For each failed attempt, the value displayed here
171 will increase by 1. Large changes in this value would indicate
172 resource contention that may require system tuning. The device
173 drivers are written such that these failures, while impacting
174 performance, do not impede functionality. To reset this
175 counter, the value '0' can be written to the file
176 /sys/bus/vio/devices/<device name>/cmo_allocs_failed.
177
179 -c, --client
180 When running with remote data, function as the client to display
181 the GUI. [default]
182 -h, --help
184 Display a help message and exit.
185 --hosts=HOSTNAME[:PORT][,HOSTNAME[:PORT],...]
187 Hostname(s) with optional port number, the default is local‐
188 host:50000. If more than one hostname is specified, only the
189 System Panel is displayed for each host.
190 -l, --local
192 Run on the local system to gather data and display it in a GUI.
193 [default]
194 -pPORT, --port=PORT
196 Network port number while running as a server, the default is
197 50000.
198 -r, --remote
200 Run as a network client/server to gather data or display it in a
201 GUI.
202 -s, --server
204 When running remotely, function as the server to provide AMS
205 data to clients. No GUI is displayed.
206 -uSECONDS, --update=SECONDS
208 Specify the number of seconds between updates of the GUI, the
209 default is 5.
210 -v, --verbose
212 Enable verbose output.
213
215 Running locally:
216 amsvis --local --update=10
218 Display the AMS data for the local system, providing an update
219 every 10 seconds.
220 Running remotely:
222 amsvis --remote --server --port=50000
224 Start a server for AMS data on port 50000.
225 amsvis --remote --client --hostname=server.example.org --port=50000
227 Display the AMS data from the remote system, providing an update
228 every 10 seconds.
229
231 /proc/sys/meminfo
232 /proc/ppc64/lparcfg
233 /sys/devices/system/cmm/cmm0/
234 /sys/bus/vio/cmo_*
235 /sys/bus/vio/devices/*/cmo_*
236
238 Though not a bug, the name of a number of files and variables are of
239 the form cmo* rather than Active Memory Sharing (AMS). The kernel code
240 for this feature was committed under the name Cooperative Memory Over‐
241 commitment (CMO). Kernel support is enabled using the kernel config
242 option CONFIG_PPC_SMLPAR, for shared memory logical partition support.
243
245 $Revision: 1.1 $
246Linux January 2009 AMSVIS(1)