Lund Performance Solutions

SOS/3000 Pulse Points
Pulse points are the indicators of performance displayed in the SOS Pulse Points screen. For information about Pulse point performance indicators, see "SOS Pulse Points".

The following HP-3000 Pulse points performance ranges are generic for all HP e3000 systems—customizing them for your system is recommended. Please refer to the configuration instructions in "Pulse Points - SOSADVIC File".
Table A.1 SOS/3000 Pulse Points
Performance Indicator
Performance Ranges
Processor Performance
High Pri Busy (%)
The summations of the AQ+BQ+CQ+Memory+Dispatch+ICS/OH. High Priority busy time is a good indicator of CPU saturation.
less than 50
50 to 85
greater than 85
The average number of processes that required use of the CPU in order to continue processing during the last interval. This number is roughly equivalent to the number of processes appearing in the upper right hand column of the SHOWQ command, except that SOS/3000 provides a current and cumulative average.
less than 5
5 to 15
greater than 15
ICS/OH+Dispatch (%)
The time the CPU spends dealing with external device activity and the time the CPU spends on dispatcher activity. Time handling disc I/O completions are included here. Interrupt Control Stack activity ICS requires service time by the CPU.
less than 10
10 to 15
greater than 15
CPU CM (%)
The average number of time the CPU spends in Compatibility Mode program code.
less than 10
10 to 15
greater than 15
The beginning letters (i.e., "A" and "B") signify the particular dispatch subqueue in which the process is executing. See "AL" and "BL" for more information.
less than 5
5 to 8
greater than 8
Memory Performance
The amount of the total CPU capacity consumed by main memory page activity. This counter includes time spent on memory allocations for user processes that cannot acquire the CPU’s attention (launched) until necessary segments are present in memory.
less than 4
4 to 10
greater than 10
Page Fault Rate
The current and cumulative number of times per second that memory page faulting occurs. A Page Fault is counted when a process needs a memory object (code or data) that is absent from main memory. The threshold between the various zones for this particular indicator do tend to vary with your CPU size.
See the approximate ranges for this item in "Page Fault Rates".
less than 10
10 to 25
greater than 25
The ratio of the number of swap-ins to the number of launches occurring for the current interval. This can also be a good indicator of memory pressure.
less than 0.4
0.4 to 0.8
greater than 0.8
Memory Cycles/Hour
This is an activity of the memory manager. When there is a request for memory space, the memory manager begins to search memory where it last left off. The time it takes to cycle through all of main memory is referred to as a "clock cycle". This time is an important performance metric.
less than 10
10 to 25
greater than 25
Disc Performance
The percentage of time the CPU spends waiting for disc I/O’s to complete.
less than 5
5 to 15
greater than 15
Read Hit %
The percentage of time that Disc Read requests were satisfied in main memory. A Read Hit percentage less than 95 could indicate a data locality problem or a shortage of memory.
greater than 95
less than 85
Average Q-Length
The average length of the disc request queue for all disc devices when another disc I/O request arrives. An average queue length of one (1.0) or greater is not acceptable.
less than 0.5
0.5 to 1.0
greater than 1.0
Disc I/O Rate/Sec
The number of disc I/O (reads and writes to disc) per second.
less than 10
10 to 25
greater than 25
Miscellaneous Performance
CM to NM Switches
The number and rate per second (nnn/s) of Compatibility Mode to Native Mode switches performed by the process.
greater than 200
NM to CM Switches
The number and rate per second (nnn/s) of Native Mode to Compatibility Mode switches performed by the process.
greater than 75

Lund Performance Solutions
Voice: (541) 812-7600
Fax: (541) 812-7611