HP 3000 Manuals

Job/Session Screen [ HP GlancePlus User's Manual (for MPE/iX Systems) ] MPE/iX 5.0 Documentation


HP GlancePlus User's Manual (for MPE/iX Systems)

Job/Session Screen 

If you select a specific job or session (using the #, J, or S commands),
the Job/Session screen appears containing information about how the job
or session is utilizing system resources.  The job or session number is
displayed in the Banner Line as part of the screen identifier.  You may
also select all system processes that do not belong to any job or session
by using the # command and specifying SYS at the prompt.  In this case,
System Processes will display as the screen identifier.  The Job/Session
screen can be used to track the resource usage of several programs that
execute in a job without having to change screens.  It is shown below.

The detailed data area of the display is divided into two parts:

   *   Job/Session Bars and Summary Fields.

   *   Interesting Processes.

[]
Figure 5-3. Job/Session Screen Job/Session Bars and Summary Fields
[]
The Job/Session Bars appear below the Global bars (CPU, Disc, and Memory) and show additional information about how the selected job or session is using CPU and Disc resources. The Job/Session Bars and Summary Fields contain the following information: CPU This CPU bar shows the percentage of central processing unit time used by the job or session during the last interval. The codes in this bar correspond to the codes used in the Global CPU bar (see CPU Bars under "Reading the Screen Display" in chapter 3). The number at the right of the bar is the percentage of time the CPU is busy (not counting idle and paused-for-disc time). An asterisk (*) represents the average usage of this resource during this execution of HP GlancePlus. You can use the Z command to reset the averages to zero and recompute them beginning with the current interval values. DISC This bar shows the job or session's disc transfer rate (transfers per second) during the interval. The codes in this bar correspond to the codes used in the Global Disc bar (see "Disc" under "Global Screen" earlier in this chapter). The overall transfer rate appears at the right of the bar. An asterisk (*) represents the average usage of this resource during this execution of HP GlancePlus. You can use the Z command to reset the averages to zero and recompute them beginning with the current interval values. Table 5-1. (cont.) COUNT The count (in parentheses) is the number of processes associated with the current job or session. There are four processes shown in the previous example. LOGON LOGON is the logon string (jobsessionname,user.account). The logon string in the previous example is "DIANNE.EVAL". ELAPSED ELAPSED is the total clock time during which a job or session has been logged on. CPU-TOT CPU-TOT is the total amount of processor time used during the life of the job or session (see note). DISC-TOT DISC-TOT is the total number of physical Disc transfers completed during the life of the job or session (see note). A number followed by the letter K represents thousands of transfers; a number followed by the letter M represents millions of transfers.
NOTE Disc totals are not collected until either HP GlancePlus or another performance tool enables the MPE measurement facility to collect process data. This means that the DISC-TOT value will not include physical transfers that were completed by a process before the MPE measurement facility was enabled. Upon entry to the Job/Session screen, CPU-TOT and DISC-TOT values represent CPU and disc totals for the processes, associated with the job or session, that are still alive. To prevent these values from decrementing as processes terminate, HP GlancePlus maintains counters for the job or session as long as it is monitored in the Job/Session screen. These counters are lost as soon as you enter another screen or select a different job or session.
Interesting Processes Interesting processes related to the job or session you select appear below the Job/Session Bars and Summary Fields. In addition, you can use the SYS option of the Select command to display all system processes not related to a job or session.
[]
Figure 5-4. Job/Session Screen--Interesting Processes Interesting Job Processes belonging to this Job (#Jnnnn) or or Session Session (#Snnnn) are considered interesting Processes here for the same reasons as in the Global screen (see "Interesting Processes" under "Global Screen" earlier in this chapter). Any job or session process that is not waiting for another process (father-, son-, or message-waited) will also be displayed. For example, the screen will display the command interpreter (CI) process for a job or session at the top of the Interesting Processes section of the screen. If a son process starts running, both processes will be displayed while both are running. If the CI process must wait for the son process to complete, the CI will disappear from the screen during the wait interval, but it will return when the wait period ends. This allows you to see the most interesting part of a job or session without having to manually track the process numbers as they are created and terminated. You can see all processes belonging to the job or session by entering the A command. Interesting Use the Select command's SYS option to System Processes display only system processes that do not belong to a specific job or session. Since there are usually many of these processes left in different waiting states, system processes will only be displayed if they are "interesting" (see "Interesting Processes" under "Global Screen" earlier in this chapter). You can see all system processes by entering the A command. Interesting Process Components Each process related to the job or session is displayed in a single line. The following components make up a line of interesting process data: PIN PIN is the process identification number for the process. PROGRAM PROGRAM refers to the name of the program being run. If the program is the MPE Command Interpreter (CI), then this field displays a colon (:) followed by the first characters of the last CI command issued. PRI PRI is the execution queue (where: A=AS, B=BS, C=CS, D=DS, E=ES) followed by the absolute execution priority number of the process (1-255). If a process has just terminated, then this field will contain the word DIED. CPU% CPU% is the percentage of the central processing unit used by this process during the last interval. DISC DISC is the disc transfer rate (transfers per second) for this process during the last interval. Table 5-1. (cont.) TRN TRN is the rate at which the Return or Enter key was pressed during the last interval. A rate is displayed to maintain independence from the time interval; thus, one transaction in a 30-second interval equates to a rate of 2 transactions per minute. The number of transactions is 0 for most batch jobs because generally they do not read from terminals. RESP RESP is the average time (in seconds) between when the Return or Enter key was pressed and the computer was ready to accept more input. This number is 0.0 if there were no transactions during the last interval. WAIT WAIT is the reason a process was waiting at the end of the last interval. See table 5-1 for definitions of wait reasons and appendix B, "Wait Reasons--Expanded Definitions." More detailed Wait reasons are defined for the Process screen (see appendix C). ELAPSED ELAPSED is the length of time a process was alive. CPU-TOT CPU-TOT is the amount of CPU time used during the life of the process. DISC-TOT DISC-TOT is the total number of physical disc transfers completed during the life of the process. A number followed by the letter K represents thousands of transfers; a number followed by the letter M represents millions of transfers. P-PIN PIN is the process identification number for the parent (or father) of this process.
NOTE Disc totals are not collected until either HP GlancePlus or another performance tool enables the MPE measurement facility to collect process data. This means that the DISC-TOT value will not include physical transfers that were completed by a process before the MPE measurement facility was enabled.


MPE/iX 5.0 Documentation