CA MICS Hardware and SCP Analyzer (RMF) Queries › Mainframe MSU Utilization and LPAR Capping › RMFMSU – Daily CPC MSU Use with Stacked LPARs

RMFMSU – Daily CPC MSU Use with Stacked LPARs

Overview

The RMFMSU query shows the average MSU usage for the Central Processing Complex (CPC) at an hourly or RMF interval level, with the individual LPAR usage broken out in each vertical bar. All LPAR MSU usage is accounted for with the exception of LPARs that have only specialty engines, such as Coupling Facility LPARs with ICF engines only or Linux LPARs with IFL engines only. If an LPAR uses at least one general-purpose CP engine, it contributes to the chart. Two MSU utilization data extracts are provided One extract generates daily charts for each CPC, and another extract generates a single chart for each CPC showing multiple days of activity if multiple HARLPC file cycles are input.

The example shows the daily chart for an IBM z/10 2097-712 CPC, with the average MSU usage for each 15-minute RMF interval,

Uses

The processing capacity of mainframe CPCs is expressed in MSUs (Millions of Service Units per Hour). This query provides a CPC level view of MSU usage. The generated MSU charts have the same pattern as the daily CP engine utilization chart by hour generated by query RMFCPD—because MSU usage is a derivative of CP engine usage. These charts help you see LPAR usage patterns relative to each other throughout the day. These charts can be compared to the MSU capacity of the CPC to help determine when an upgrade (or downgrade) is considered.

Data Extracts

Daily MSU Use by CPC Stacked LPARs

Left Y-axis

LPCAVMSU – Avg Number of Consumed MSUs

X-axis

HHMM – Hour : Minute

Data element HHMM is derived by concatenating the HOUR and MINUTE of each RMF interval together, separated by a colon, HH:MM (for example, 14:59 meaning hour 14, 59th minute).

Multiday MSU Use by CPC Stacked LPARs

Left Y-axis

LPCAVMSU – Avg Number of Consumed MSUs

X-axis

DATEHHMM – Date Hour Minute

Data element DATEHHMM is derived by concatenating the DATE, HOUR and MINUTE of each RMF interval together in the format YYYY.MM.DD-HH:MM.

Query Defaults

• HARLPC DAYS timespan file: Cycles 01 - 03
• Filters: PRSMLPTP EQ ‘CP’ or PRSMLPTP EQ ‘PHY’ (select LPARs using CP engines)
• Derivations:

  CPCID

  TRIM(LPCMOD)||'-'||TRIM(CPCMODID)||'-Serial: '||SUBSTR(CPCSEQNB,12,5);
  

  DATE

  DATEPART(ENDTS);
  

  HHMM

  PUT(HOUR(ENDTS),Z2.) || ':' || PUT(MINUTE(ENDTS),Z2.);
  

  DATEHOUR

  DATEHOUR = PUT(DATEPART(ENDTS),YYMMDDD10.) || ':' || PUT(HOUR,Z2.) ;
  

  DATEHHMM

  DATEHHMM = PUT(DATEPART(ENDTS),YYMMDDP10.) || '-' || HHMM ;
  

Modifications

The following modifications can be made to the RMFMSU query:

• If you want to see hourly, rather than interval level MSU usage averages, you can change use the HARLPC DAYS timespan . If so, you may want to substitute the DATEHOUR element for the DATEHHMM element in the SUMMARIZATION and DATA EXTRACT steps.
• If you want to see only yesterday’s data, uncomment the User Code in User Step task in the Input HARLPC File step.
• You can increase or decrease the number of HARLPC file cycles to see more or fewer days of average MSU utilization.
• You can use a filter statement to restrict data to focus on a particular time range (for example, HOUR GE 8 AND HOUR LE 17) or particular time zone (for example, ZONE EQ 1)
• You can add filter statements to select only a single CPC, or set of CPCs using the identifying data elements shown in the CPCID derivation.