1 02/26/85 post_purge_meters, ppm
2
3 Syntax as a command: ppm -control_arg
4
5
6 Function: displays information collected at post purge time, if post
7 purging is enabled. The print_tuning_parameters and work_class_meters
8 commands described later in this section are used to determine which
9 work classes, if any, are being post purged.
10
11
12 Control arguments:
13 -reset, -rs
14 resets the metering interval for the invoking process so that the
15 interval begins at the last call with -reset specified. If -reset
16 has never been given in a process, it is equivalent to having been
17 specified at system initialization time.
18 -report_reset, -rr
19 generates a full report and then performs the reset operation.
20
21
22 Access required: This command requires access to phcs_ or
23 metering_gate_.
24
25
26 Notes: If the post_purge_meters command is given with no control
27 argument, it prints a full report.
28
29 The following is a brief description of each of the variables printed
30 out by the post_purge_meters command.
31 Post purge time
32 is the average CPU time per post purge call.
33 Ave list size
34 is the average number of page fault entries found in the per-process
35 page trace list at post purge time.
36
37
38 Ave working set
39 is the average estimated working set. The current estimated working
40 set for each process is computed by the following formula.
41
42 working set = working_set_factor * raw_working_set
43 + working_set_addend
44
45 The raw working set is estimated by page control at post purge time.
46 Working set factor
47 is the current value of the wsf tuning parameter, and can be changed
48 by the change_tuning_parameters command. Increasing the value tends
49 to reduce page thrashing, but may increase multiprogramming idle.
50 Decreasing the value has the opposite effects.
51
52
53 Working set addend
54 is the current value of the wsa tuning parameter, and can be changed
55 by the change_tuning_parameters command. Increasing and decreasing
56 this value has the same effects as noted above.
57 Thrashing percentage
58 is the percentage of page faults that were taken on pages faulted
59 earlier in quantum.
60 Ave post in main memory
61 is the average number of entries in the trace list for which the
62 page was still in main memory at post purge time, and the ratio of
63 incore pages to faulted pages expressed as a percentage.