1 02/28/85 traffic_control_meters, tcm
2
3 Syntax as a command: tcm -control_args
4
5
6 Function: prints out the values of various traffic control meters.
7
8
9 Control arguments:
10 -counters, -ct
11 prints out the number and frequency of certain paths through the
12 traffic controller.
13 -gen
14 prints out general traffic control information.
15 -queue, -qu
16 prints out certain resource usage as a function of depth in the
17 eligible queue.
18
19
20 -report_reset, -rr
21 generates a full report and then performs the reset operation.
22 -reset, -rs
23 resets the metering interval for the invoking process so that the
24 interval begins at the last call with -reset specified. If -reset
25 has never been given in a process, it is equivalent to having been
26 specified at system initialization time.
27
28
29 Access required: This command requires access to phcs_ or
30 metering_gate_.
31
32
33 Notes: If the traffic_control_meters command is given with no control
34 arguments, it prints a full report.
35
36 The following meters reflect activity of the traffic controller, and
37 some constants used therein. They are printed if the -gen control
38 argument is specified.
39 Ave queue length
40 is the average number of processes in the eligible and priority
41 queues. This is the average number of ready, waiting, or running
42 processes.
43
44
45 Ave eligible
46 is a recent average of the number of eligible processes.
47 Response time
48 is the average time between a process' receiving an interactive
49 wakeup and the awarding of eligibility to the process. The response
50 time seen by the user is larger than this meter.
51
52
53 The following meters pertain to the number and frequency of certain
54 paths through the traffic controller. They are printed if the -ct
55 control argument is specified.
56 Interactions
57 is a count of, and the average time between, terminal interactions.
58 Loadings
59 is a count of, the average time between, and the number per
60 interaction of process loadings.
61 Blocks
62 is a count of, and the average time between, calls to "block" to
63 block some process.
64 Wakeups
65 is a count of, and the average time between, wakeup signals being
66 sent.
67
68
69 Schedulings
70 is a count of, the average time between, and the number per
71 interaction of trips through the scheduler/rescheduler function that
72 caused priorities to be changed.
73 Lost priority
74 is the number of times the alarm clock went off indicating a
75 priority process that had been running lost its eligibility because
76 it had used up its eligible time; i.e., its eligible time exceeded
77 the CPU quantum that the process remains in the queue. The process
78 reenters the traffic controller to be rescheduled.
79
80
81 Priority boosts
82 is the number of times the alarm clock went off indicating a
83 priority scheduling process on the ready list should be granted high
84 priority; i.e., have its waiting time before rescheduling set to 0.
85 The process is then resorted into the ready list with its new,
86 higher priority.
87 Wait Page
88 is a count of, the average time between, and the number per
89 interaction of calls to force some process to a wait state in order
90 to wait for page transfer.
91 Wait PTL
92 is a count of, the average time between, and the number per
93 interaction of calls to force some process to a wait state in order
94 to wait for the page table lock.
95
96
97 Wait Other
98 is a count of, the average time between, and the number per
99 interaction of calls to force some process to a wait state in order
100 to wait for events other than page control events.
101 Total Waits
102 is a count of, the average time between, and the number per
103 interaction of calls to force some process to a wait state.
104 Notify Page
105 is the number of, and average time between, calls to notify
106 processes waiting for page transfer events.
107 Notify PTL
108 is the number of, and average time between, calls to notify
109 processes waiting for page table unlockings.
110
111
112 Notify Other
113 is the number of, and average time between, calls to notify
114 processes waiting for all other events.
115 Total Notifies
116 is the number of, and the average time between, notify calls i.e.
117 returning a waiting process to the ready state.
118 Get Processor
119 is the number of, and average time between, calls to get_processor.
120 Get_processor is called at notify time to find a CPU on which to run
121 the notified process. An idle process or lower priority running
122 process is preempted.
123 Pre-empts
124 is a count of, average time between, and the number per interaction
125 of process preemptions and timer runout faults.
126
127
128 Getwork
129 is the number of, and average time between, calls to getwork.
130 Getwork is the dispatcher portion of the scheduler; it finds a
131 process to run on the executing CPU.
132 Retry getwork
133 is the number of, and average time between, retries of the getwork
134 function.
135 Extra notifies
136 is the number of, and average time between, notify calls that found
137 no process waiting on the notified event.
138
139
140 Last EN event
141 is the last notified event for which no process was waiting.
142 Notify timeout
143 is the number of times a notify was not received by a waiting
144 process within notify_timeout_interval a tuning parameter. This
145 is printed only if the count is nonzero.
146 Last NTO event
147 is the last event on which a notify timeout occurred.
148
149
150 The following meters pertain to the eligible queue. They are printed
151 if the -qu control argument is specified.
152 Depth
153 is the depth of the process within the eligible queue. A process
154 deep in the eligible queue is run only if processes above it cannot
155 run.
156 %PF
157 is the percentage of page faults that occurred from processes at
158 this depth.
159
160
161 TBPF
162 is the average time between page faults at this depth.
163 %GTW
164 is the percentage of getwork calls being made when a member of this
165 priority relinquishes control.
166 TBS
167 is the average time between getwork calls at this priority level.
168 %CPU
169 is the percentage of CPU time consumed by members of this priority.