Best practice metrics in Threshold Advisor

CREGION

CRGNSMFI

SMF ID associated with the z/OS system that is running the CICS region

Target identification.

CREGION

CRGNTARG

Name of the target CICS region, which is the job name or started task ID associated with the region

Target identification.

CREGION

CRGRPMXT

Number of tasks in the CICS region in relation to the maximum task value. This is the defined capacity of the CICS region

As this metric approaches 100%, the ability of the CICS region to absorb excess work is reduced. Once CICS reaches the maximum task value, it will no longer dispatch new work in the region.

CREGION

CRGIRESP

Average response time of all transactions executing in the CICS region during the recording interval

Average response time is an indication of problems that might be causing additional delays in the execution of work in the CICS region. If average response time exceeds its target, you should examine response time factors in task performance history.

CREGION

CRGIEXCR

Number of EXCPs issued by the CICS region during the current sample expressed as the rate per second over the recording interval

This metric is an indication of the amount of I/O being performed by the CICS region against CICS files. An unexpected decrease in this rate might be an indication of file contention.

CREGDSAS

CRGNTARG

Name of the target CICS region, which is the job name or started task ID associated with the CICS region

Target identification.

CREGDSAS

CRGRDSA

Percent of the dynamic storage area (DSALIMIT) that is currently being used

This metric is the percentage of below-the-line storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGREDSA

Percent of the extended dynamic storage area (EDSALIMIT) that is currently being used

This metric is the percentage of above-the-line storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGRGTPCT

Percent of above the bar dynamic storage area (GDSALIMIT) that is currently being used

This metric is the percentage of above-the-bar storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGRMVSPCT

Percent of z/OS storage in use below the 16 MB line

The percentage of z/OS below-the-line storage is an indication of how much available storage is allocated for use by CICS DSA. If this percentage is too high, the region will not be able to allocate sufficient DSA to satisfy below-the-line requests.

CREGION

CRGRCPUP

CPU time consumed by the CICS region expressed as a percentage of the total available time

This number uses a single processor as its base so it is possible that a CICS region could report more than 100 percent CPU.

This metric is an indication of the amount of CPU resources the CICS region is consuming. An increase in CPU utilization might indicate that more work than expected is being pushed through the region due to slowdowns in other regions, or that some transactions have encountered issues with loops. A decrease might indicate the CICS region is experiencing slowdowns due to resource contention.

CREGION

CRGITRTE

Number of transactions per second during the recording interval

This metric is the rate of work being processed by the CICS region. Deviation from a target transaction rate might indicate a slowdown in this CICS region (the rate decreases) or other CICS regions (the rate increases).

CREGION

CRGIRESP

Average response time of all transactions executing in the CICS region during the recording interval

Average response time can be an indication of problems that might cause additional delays in the execution of work in the CICS region. If the average response time exceeds the target, you should examime response time factors in task performance history.

CREGION

CRGIEXCR

Number of EXCPs issued by the CICS region during the current sample expressed as the rate per second over the recording interval

This metric is an indication of the amount of I/O being performed by the CICS region against CICS files. An unexpected decrease in this rate might be an indication of file contention.

CREGDSAS

CRGNTARG

Name of the target CICS region, which is the job name or started task ID associated with the CICS region

Target identification.

CREGDSAS

CRGRDSA

Percent of the dynamic storage area (DSALIMIT) that is currently being used

This metric is the percentage of below-the-line storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGREDSA

Percent of the extended dynamic storage area (EDSALIMIT) that is currently being used

This metric is the percentage of above-the-line storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGRGTPCT

Percent of above the bar dynamic storage area (GDSALIMIT) that is currently being used

This metric is the percentage of above-the-bar storage in use by the CICS region. If this percentage is too high, the CICS region will not be able to handle additional work and might stall.

CREGDSAS

CRGRMVSPCT

Percent of z/OS storage in use below the 16 MB line

The percentage of z/OS below-the-line storage can indicate how much available storage is allocated for use by CICS DSA. If this percentage is too high, the region will not be able to allocate sufficient DSA to satisfy below-the-line requests.

CREGDSAS

CRGREMVSPCT

Percent of z/OS storage in use above the 16 MB line

The percentage of z/OS above-the-line storage can indicate how much available storage is allocated for use by CICS DSA. If this percentage is too high, the region will not be able to allocate sufficient EDSA to satisfy above-the-line requests.

COBJ

MONPARM

Name of a product target included in a workload definition, which is used to identify the monitor to be tested

Target identification.

COBJ

MONTARG

Job name of the system that is being monitored

Target identification.

COBJ

MONICVAL

Reported value of data collected by the monitor during the current interval

This metric represents the average response time for CICS transactions that are part of this workload. A increase in the expected average response time for transactions in the workload can be an indication of resource contention causing delays.

COBJ

MONICCNT

Number of recorded transactions from a product workload during the current interval

This metric represents the number of CICS transactions that executed during the current interval as part of this workload.

COBJ

MONIPOBJ

Percentage of transactions in the current interval whose response times are less than or equal to the response time goal set in a workload definition

This metric indicates how well the workload is doing in relation to its response time goals. The goal is a target response time, and this metric indicates what percentage of transactions that ran during the interval met the goal.

CTGSERV

RC1CTGNAME

Job name of the CICS Transaction Gateway (CTG) daemon

Target identification.

CTGSERV

RC1CTGAPPL

APPLID of the CTG daemon.

Target identification.

CTGSERV

RC1GDCHEALT

Current health value for communications between the CTG daemon and CICS

Values range from 0 to 100, with the higher the value the healthier the state.

The health value is an indication of whether the CTG daemon has to wait for threads to process work. If this value is less than 100, CTG is waiting for threads.

CTGSRVTP

RC1CTGAPPL

APPLID of the CTG daemon

Target identification.

CTGSRVTP

RC1ITRANRAT

Calculated transaction rate over the monitored lifetime of CTG

This metric is the rate of work being processed by CTG. Deviation from a target transaction rate can indicate a slowdown in this CTG or the CICS regions where the work is being performed (the rate decreases) or other CTGs (the rate increases).

CTGSRVTP

RC1ITASKAV

Average elapsed time of all transactions that were processed by the CTG daemon during the reporting period

Average elpased time can be an indication of problems that might cause additional delays in the execution of work in the CTG. If the average response time exceeds the target, you should examime response time factors in task performance history

CTGSRVTP

RC1ITASKWAV

Average portion of elapsed time that all transactions (both EXCI and IPIC) spent waiting for threads during the reporting period

Waiting for threads in the CTG indicates a problem. If this value climbs, it indicaties that too much work is being processed through CTG with too few threads available to process the requests. This is an indication of increased workload without an increase in the ability of CTG to manage the load.

STBFRPL

STDB2TARG

DB2 subsystem ID

Target identification.

STBFRPL

SIRPI

Buffer pool page-ins for read

This metric indicates z/OS paging activity due to a shortage of real storage. This value is frequently the result of excessive load on the LPAR or inadequate real storage configuration.

STBFRPL

SIWPI

Buffer pool page-ins for write

This metric indicates z/OS paging activity due to a shortage of real storage. This value is frequently the result of excessive load on the LPAR or inadequate real storage configuration.

STDB2

STDB2TARG

DB2 subsystem ID

Target identification.

STDB2

STATOT

Active threads

Higher than normal transaction volumes can cause constraints to various system resources. In that case, system resources might need to be tuned or the workload might need to be throttled back.

STDB2

STEXTOT

Total exceptions

This metric indicates the existence of potential system-wide problems that should be individually investigated and resolved.

STDB2

SCAVSTOR

DBM1 storage available

DB2 will abend if all available virtual storage below the 2 GB bar is exhausted. Further investigation using DB2STORD might be necessary.

STDB2

STCPUDB2

CPU % - DB2 threads

High consumption of CPU by DB2 threads might indicate more work taking place, inefficient queries, or a looping application.

STDB2

STCPUSYS

CPU % - LPAR

A CPU-overloaded LPAR can cause degradation in performance. This might be the result of high DB2 activity, non-DB2 activity, or both.

STDB2

SCDB2CPU

CPU % - DB2 subsystem

This metric refers to DB2 subsystem address space overhead. A certain amount is normal, but it should be investigated if the value gets too high.

STLOCK

STDB2TARG

DB2 subsystem ID

Target identification.

STLOCK

SILOCKR

Locks

Locks are typical of normal application activity, but they might be a factor in elongated application elapsed times if their numbers are higher than usual.

STLOCK

SIUNLK

Unlocks

The number of unlocks should be close to the number of locks; otherwise, applications might be holding locks too long.

STLOCK

SIDRNO

Drains

This metric indicates that utility or command processing is attempting to serialize access to a database resource. This situation might adversely impact normal application activity.

STLOCK

SIDEA

Deadlocks

Deadlocks are a serious form of lock contention that should seldom occur. They are usually the result of faulty application logic.

STLOCK

SITIM

Timeouts

When a timeout occurs, the application abends after the site-specified waiting period has elapsed. Timeouts should not occur very frequently.

DOBJ

MONTARG

DB2 subsystem ID

Target identification.

DOBJ

MONPARM

Workload name

Target identification.

DOBJ

MONAPPL

Composite workload name

Target identification.

DOBJ

MONICVAL

Average response time

Higher than usual response time might indicate the existence of new long-running applications, a performance problem with existing applications, or both.

DOBJ

MONICCNT

Transaction count

Special performance evaluation and tuning might be necessary as the overall workload in DB2 increases.

DOBJ

MONIPOBJ

% Transaction meeting goal

A decrease in this value might mean service levels are no longer being met. If that is the case, immediate tuning action might be required.

STRATES

STDB2TARG

DB2 subsystem ID

Target identification.

STRATES

SISCTHDT

Thread create rate

In a non-batch environment, the lower this number is compared to the commit rate, the more benefit is being realized from thread reuse.

STRATES

SISCOMIT

Commit rate

This is the best metric of transactions per second. This value is not necessarily bad if high, but it might require tuning as suggested by other indicators.

STRATES

SISDMLT

DML SQL rate

Along with the commit rate, this metric might indicate higher transaction volume, more complex applications, or both. Tune as necessary.

STRATES

SISCALL

Stored procedure call rate

Increased stored procedure activity might require WLM tuning, DDF tuning, or both.

STRATES

SISPLGRP

Parallel group rate

A high value is not necessarily a problem, although inefficient queries might be responsible for parallel processing activity. Adequate buffer pool space must be provided.

STRATES

SISDSO

Data set open rate

If workloads are stable, database data sets should never have to be closed and reopened. Adjust the DSMAX ZPARM value, if necessary.

STRATES

SISGET

Getpage rate

A much higher than usual getpage rate might indicate inefficient queries using table scan access instead of index access.

STRATES

SISSWS

Page update rate

This value is governed by application activity. Expect a higher update rate for special batch processing (such as month-end). Logging activity will go hand-in-hand.

STRATES

SISRIO

Sync read I/O rate

A small increase might be nominal; a larger increase might indicate an application problem. Higher rates might require additional buffer pool resources.

STRATES

SISPIO

Async read I/O rate

A higher than normal rate is Often caused by inefficient queries performing full table scans. The situation might also be the result of normal batch or utility processing.

STRATES

SISWRTS

Write I/O rate

Like page update rate, this value is a result of data change activity, but might also indicate that buffer pool tuning is necessary.

STRATES

SISEDML

EDM load rate

This situation should be rare unless applications and databases are initially being accessed. If not, EDM pool tuning might be necessary.

STRATES

SISSUSRT

Lock suspend rate

This metric indicates that applications are experiencing degradation due to lock wait time. This situation should be kept to a minimum.

STRATES

SISLCTOT

Latch suspend rate

This value is the result of internal DB2 activity and is typically inconsequential. If the value is significant, investigate with IBM.

STRATES

SISLKOUT

Lockout rate

This metric indicates that deadlocks and timeouts are occurring. This situation should be considered abnormal and a cause for investigation.

STRATES

SISGCONT

Global lock suspend rate

This metric is the same as lock suspend rate except the scope is between members of a data sharing group.

STDIST

STDB2TARG

DB2 subsystem ID

Target identification.

STDIST

SIQDBT

DBATs queued - MAXDBAT reached

Distributed transactions are having to wait to be processed within DB2. The MAXDBAT ZPARM value might need to be increased.

STDIST

SIQCRT

Connections deallocated - CONDBAT reached

More distributed connections to DB2 are being requested than the CONDBAT ZPARM value allows. The value might require adjustment.

STDIST

SCCNAT

Current DBATs active and pooled

If the overall distributed transaction load is growing significantly, consider throttling with ZPARMs or performing DB2 subsystem tuning.

STDIST

SCMQR2

HWM connections queued for DBAT

The maximum number of waiting connections can be used along with other indicators to determine if distributed connection tuning is necessary.

STCACHE

STDB2TARG

DB2 subsystem ID

Target identification.

STCACHE

SIPFND

Global cache hit ratio

As this ratio degrades from 100%, reusable dynamic SQL statements must be reloaded and reprepared.

STCACHE

SIDEXP

Statements discarded

The least recently used SQL statements were discarded from the cache because the MAXKEEPD ZPARM value had been reached. Consider increasing the MAXKEEPD value or the cache itself.

STCACHE

SIHNPRP

Local cache hit ratio

As this ratio degrades from 100%, reusable dynamic SQL statements must be reloaded and reprepared.

STCACHE

SIDINV

Statements purged from cache

This situation is often the result of a DROP ALTER or REVOKE statement against a dependent object. Such configuration changes should be tightly controlled in a production environment.

STDSA

STDB2TARG

DB2 subsystem ID

Target identification.

STDSA

SCDSOPN

Data sets open

This metric represents the number of concurrently open database data sets. If this number grows significantly, make sure adequate virtual storage is available in the DBM1 address space.

STDSA

SIPCCT

Convert RW to RO

This metric ndicates that infrequently used database objects have been converted to read-only to reduce the possible impact on recovery of large numbers of objects in read-write status.

STDSA

SIDSDRN

Datasets closed by drain

This metric is incremented when a data set has not been recently used and DB2 has reached 99% of the DSMAX ZPARM value. The value can be increased, if adequate DBM1 storage exists.

STEDMP

SIPHDBD

DBD hit ratio

If this value degrades significantly in a stable production environment, consider increasing the size of the DBD cache EDMDBDC ZPARM.

STEDMP

SIPHCT

CT hit ratio

If this value degrades significantly in a stable production environment, consider increasing the size of the EDMPOOL ZPARM.

STEDMP

SIPHPT

PT hit ratio

If this value degrades significantly in a stable production environment, consider increasing the size of the EDMPOOL ZPARM.

STGBLLK

STDB2TARG

DB2 subsystem ID

Target identification.

STGBLLK

SIGCONTP

% Global contention

In a data sharing environment, the degree of lock contention between members should be kept to a minimum. Standard lock tuning applies, except between members of a data sharing group.

STLOG

STDB2TARG

DB2 subsystem ID

Target identification.

STLOG

SILSUS

Suspends log writes

The number of times applications had to wait for log records to be written to the active log. A moderate value here is not necessarily a problem; a higher value might mean log tuning is needed.

STLOG

SIBPAG

Log output buffers paged in

This metric indicates z/OS paging activity due to a shortage of real storage. This situation should be a very rare occurrence.

STLOG

SISCIWR

Active log bytes written per second

This value is the result of normal application processing. If excessive log data set I/O occurs, tuning might be necessary.

STSERV

STDB2TARG

DB2 Subsystem ID

Target identification.

STSERV

SICTHW

Create threads queued

This value represents the number of times a user thread had to wait to be processed because the CTHREAD ZPARM limit had been reached. Increase the CTHREAD limit, if adequate resources are available.

STSERV

SIHWIB

Batch threads HWM

Batch jobs can often negatively impact the performance of transaction processing by locking users out of key database resources. Be aware of batch jobs running when they shouldn't be.

STSERV

SIHWIF

TSO threads HWM

Like batch jobs, TSO threads sometimes represent administrative activity that could increase locking contention among normal production transactions.

STSERV

SIABRT

Aborts

Rollbacks, whether the result of abends or initiated by applications themselves, indicate abnormal activity. Further investigation is warranted, if this value increases.

STSERV

SIHWCT

Allied threads HWM

This metric epresents the overall application workload within DB2. If the workload grows significantly, DB2 tuning as well as z/OS tuning and adjustments to hardware resources might be necessary.

IMSDASHR

IMSSIMSI

IMS ID

Target identification.

IMSDASHR

IMSFFMSQ

Number of messages queued for full function databases

A high value could indicate a lack of processing regions, workload flucuation, or both.

IMSDASHR

IMSFFAVG

Average response time of full function database messages

This metric is a measure of response time health for online transactions.

IMSDASHR

IMSFFRAT

Current number of full-function messages arriving (enqueues) per second

This metric is a measure of the speed of IMS online transaction workloads and fluctuations.

IMSDASHR

IMSFPMSQ

Number of messages queued for Fast Path databases

A high value could indicate a lack of processing regions, workload fluctuation, or both.

IMSDASHR

IMSFPAVG

Average response time of Fast Path database messages

This metric is a measure of response time health for online transactions.

IMSDASHR

IMSFPRAT

Current number of Fast Path messages arriving (enqueues) per second

This metric is a measure of the speed of IMS online transaction workloads and fluctuations.

IMSDASHR

IMSBMMSQ

Number of batch message processing (BMP) messages queued

This value is the number of messages queued for batch processing.

IMSDASHR

IMSBMRAT

Current number of BMP messages arriving (enqueues) per second

This metric is a measure of the speed of BMP transaction workloads and fluctuations.

IMSDASHR

IMSCTLCP

Current CPU utilization of the IMS control region

Excessive CPU usage by IMS could effect all other work on z/OS images.

IMSDASHR

IMSIRLCP

Current CPU utilization of the IRLM to which IMS has identified itself

A high value could indicate an excessive amount of locking in database access.

IMSDASHR

IMSDLSCP

Current CPU utilization of the DLISAS to which IMS has identified itself

This metric could indicate an increase or decrease in accessing databases; a significant increase could indicate a possible loop in an application.

IMSDASHR

IMSDBRCP

Current CPU utilization of the DBRC address space to which IMS has identified itself

A high value could indicate an increase of tasks or jobs accessing RECONs, or disorganized RECONs.

IMSDASHR

IMSMVSCP

Current CPU utilization of z/OS

This metric could indicate an increase or decrease in workloads.

IMSDASHR

IMSRECAU

Current percentage of RECA (receive any buffers VTAM) pool utilization

A high value might indicate a shortage of pool space and cause for IMS disapatching slowdowns.

IMSDASHR

IMSDBBFU

Current percentage of DBBF pool utilization

A high value might indicate a pool shortage and possible failure in the scheduling of MPRs.

IMSDASHR

IMSPSBWU

Percentage of PSBW pool utilization

A high value might indicate a possible pool shortage and might prevent transactions from scheduling in MPRs.

IMSDASHR

IMSPIWU

Percentage of program isolation pool utilization

A high value might indicate excessive database locking contention and might result in pool shortages.

IMSDASHR

IMSDMBWU

Percentage of DMBW pool utilization

A high value might indicate more databases are being opened or closed too often and might result in pool shortages.

IMSDASHR

IMSOLDPC

Current percentage of OLDS that are full

A high value might indicate an OLDS shortage and possible DASD contention or application looping.

IMSDASHR

IMSOLDAV

Number of available OLDS

This metric can indicate whether extra OLDS should be allocated to prevent a shortage.

IMSDASHR

IMSLCHWC

Current number of latch waits

An increase in this value might indicate an internal IMS latching problem.

IMSDASHR

IMSVSAVG

Average response time for VSAM buffer pools

An increase in this value might indicate the need for more buffers or separating sizes of VSAM buffer pools.

IMSDASHR

IMSOSAVG

Average response time for OSAM buffer pools

An increase in this value might indicate the need for more buffers or separating sizes of OSAM buffer pools.

IMSDASHR

IMSDEAVG

Average response time for DEDB databases

An increase in this value might indicate a pool shortage or disorganized DEDB databases.

IMSDASHR

IMSBFULS

Number of times the OLDS buffers are full when the system checkpoint is not in progress

An increase in this value indicates the number of OLDS buffers should be increased.

IMSDASHR

IMSLKWPI

Number of program isolation waiters

Program isolation (PI) pool size might be too small, or application processes are not taking enough checkpoints.

IMSDASHR

IMSLKWIR

Number of IRLM waiters

A high value is a possible indication that IRLM locks are being used up by too many requestors.

IMSDASHR

IMSQLGUT

Percentage of long message queue data set utilization

A high value is an indication of a large number of long messages being enqueued to the LGMSGQ data set.

IMSDASHR

IMSQSHUT

Percentage of short message queue data set utilization

A high value is an indication of a large number of short messages being enqueued to the SHMSGQ data set.

IMSDASHR

IMSQBKUT

Percentage of QBLKS message queue data set utilization

A high value is an indication of a large amount of inserts to the SHMSGQ, LGMSGQ, or both.

IMSDASHR

IMSFAILP

Percentage of transaction failures compared to total activity

A high value could indicate pool shortages, program conflicts, or intent failures at scheduling time.

IRGOCCR

RGNOROCC

Percentage of time the region is busy processing transactions/scheduling

This metric can indicate if class assignment is balanced, or a possible overuse or underuse of a processing region.

QMMVS

QMNAMES

Queue manager name

Target identification.

QMMVS

QMNQMES

Normal queue messages

Number of normal queue messages.

QMMVS

QMXQMES

Transmission queue messages

Number of transmission queue messages.

QMMVS

QMIPUTR

Interval put rate

The rate of put requests during the current interval.

QMMVS

QMIGETR

Interval get rate

The rate of get requests during the current interval.

BP

SBPQMSN

Queue manager name

Target identification.

BP

SBPBUFID

Buffer pool ID

Target identification.

BP

SBPRPRR

Page read ratio

Shows the efficiency of page retrieval within the buffer pool during the most recent sample.

BP

SBPUPCT

Percent used

Shows percentage used of the buffer pool.

CF

CFQMGRN

Queue manager name

Target identification.

CF

CFSTRNAM

CF structure name

Target identification.

CF

CFSTRIUP

CF entries in use percent

Percentage of list entries that are in use for the structure.

LM

LMQMGRN

Queue manager name

Target identification.

LM

LMRRDLY

Realtime archive log ratio

If this number is greater than 0.01 (1%), you should increase the MAXALLC value in CSQ6LOGP.

LOGID

LOGQMGR

Queue manager name

Target identification.

LOGID

LOGUSED1

Log used for copy 1

The percentage of the first copy of the active log DSN that has been used.

LOGID

LOGUSED2

Log used for copy 2

The percentage of the second copy of the active log DSN that has been used.

PS

SPSPSID

Page set ID

Target identification.

PS

SPFREE

Percent free

The percentage of pages available for use.

CFOVER

CFGCFNM

CF name

Target identification.

CFOVER

CFIPRCPC

CPU utilization %

This metric indicates how much coupling facility (CF) CPU is consumed by CF activity. When close to 100%, tasks have to wait and application delays might occur.

CFOVER

CFISTGPC

Storage utilization %

A high value indicates CF-dedicated storage is overcommitted, which might cause failure to allocate additional structures or structure rebuild from another CF.

CFOVER

CFICRQRT

Request rate per second

This metric is the rate per second for requests originating from this system to all structures in this CF. An unusually high value might cause application delays.

CSASUM

CSRECSUP

CSA in use %

A high value indicates CSA is nearing its maximum.

CSASUM

CSREECUP

ECSA in use %

A high value indicates ECSA is nearing its maximum.

CSASUM

CSRESQUP

SQA in use %

A high value indicates SQA is nearing its maximum.

CSASUM

CSREESUP

ESQA in use %

A high value indicates ESQA is nearing its maximum.

CSASUM

CSRECSCP

Converted CSA to SQA %

A non-zero value indicates SQA has been exhausted.

CSASUM

CSREECCP

Converted ECSA to ESQA %

A non-zero value indicates ESQA has been exhausted.

SYSOVER

SUCCPUB

CPU busy %

This metric represents the utilization of standard CPs. An unusually high value might cause application delays or indicate high usage by particular jobs.

SYSOVER

SUCZAAB

zAAP busy %

This metric represents the utilization of zAAPs. An unusually high value might cause application delays or indicate high usage by particular jobs.

SYSOVER

SUCZIIB

zIIP busy %

This metric represents the utilization of zIIPs. An unusually high value might cause application delays or indicate high usage by particular jobs.

SYSOVER

SCIAIRW

Average CPU queue length

This metric represents the average number of address spaces that had to be queued before being dispatched because a processor was not available. An unusually high value might cause CPU delays for applications.

JCPU

ASGNAME

Job name

Target identification.

JCPU

ASGJBID

JES job number

Target identification.

JCPU

ASICPPAU

CPU utilization %

The unadjusted CPU utilization percent shows the percentage of a single processor used by the address space. An unusually high value might indicate an issue within a job that could potentially affect other jobs.

LPARCAP

LPGNAME

Partition name

Target identification.

LPARCAP

LPREMSU

Rolling 4-hour MSU

This metric represents standard CP usage in MSU per hour over the last four hours. For a subcapacity-eligible product, the combined peak four-hour rolling MSU/hour during a month of the LPARs that the product runs in is used to determine the licensing cost for the product in that month.

LPARCAP

LPG4HRUT

Rolling 4-hour MSU as % of CEC capacity

This metric represents standard CP usage of the LPAR as % of CEC capacity over the last four hours. When the CEC capacity usage is unusually high, this value helps idenfity the LPARs that are contributing to the increase.

LPARCAP

LPREC4P2

Rolling 4-hour MSU as % of LPAR defined capacity

This metric represents standard CP usage of the LPAR as % of defined capacity over the last four hours when the LPAR has a defined capacity. When exceeding 100%, the LPAR is soft capped, causing CPU delay to applications.

LPARCAP

LPICPUB

Interval MSU

This metric represents standard CP usage in MSU per hour during the interval. Values trending upward through several intervals might lead to increasing rolling 4-Hour MSU/hour for the LPAR.

LPARCAP

LPGCECUT

Interval MSU as % of CEC capacity

This metric represents standard CP usage of the LPAR as % of CEC capacity during the interval. When the CEC capacity usage is unusually high, this value helps idenfity the LPARs that are contributing to the increase.

LPARCAP

LPGINTGU

Interval MSU as % of group capacity

This metric represents standard CP usage of the LPAR as % of group capacity during the interval when the LPAR belongs to a capacity group. Values trending upward through several intervals might lead to increasing rolling 4-Hour MSU/hour for the capacity group. When the group capacity value is exceeded, all the LPARS of the group are soft capped, causing CPU delay to applications.

LPARCAP

LPG4HRGU

Rolling 4-hour MSU as % of group capacity

This metric represents standard CP usage of the LPAR as % of group capacity over the last four hours when the LPAR belongs to a capacity group. When the sum for all LPARs of the group exceeds 100%, all the LPARS of the group are soft capped, causing CPU delay to applications.