Contents | Previous | Next

Glossary of Performance Terms

Many of the performance terms use postscripts to reflect the sources of the data described in this document. In all cases, the terms presented here are taken directly as written in the text to allow them to be found quickly. Often there will be multiple definitions of the same data field, differing only in the postscript. This allows the precise definition of each data field in terms of its origins. The postscripts are:

<none>

No postscript indicates that the data are obtained from the VM/ESA Realtime Monitor.

(C)

Denotes data from the VSE console timestamps.

(H)

Denotes data from the internal processor instrumentation tools.

(I)

Denotes data from the CP INDICATE USER command.

(Q)

Denotes data from the SFS QUERY FILEPOOL STATUS command.

(QT)

Denotes data from the CP QUERY TIME command.

Server

Indicates that the data are for specific virtual machines, (for example SFS, CRR, or VTAM/VSCS). If there is more than one virtual machine of the same type, these data fields are for all the virtual machines of that type.

(S)

Identifies OS/2 data from the licensed program, System Performance Monitor 2 (SPM2).

(T)

Identifies data from the licensed program, Teleprocessing Network Simulator (TPNS).

(V)

Denotes data from the licensed program VM Performance Reporting Facility.  
 
 
 
 

Absolute Share

An ABSOLUTE share allocates to a virtual machine an absolute percentage of all the available system resources.

AVG FIRST (T)

The average response time in seconds for the first reply that returns to the screen. For non-fullscreen commands this is the command reflect on the screen.  

AVG LAST (T)

The average response time in seconds for the last response to the screen.

AVG THINK (T)

The average think time determined by TPNS for all users.

Bactrian

A two-humped curve used to represent the think times for both active users and users who are logged on but inactive. The distribution includes those long think times that occur when a user is not actively issuing commands. Actual user data were collected and used as input to the creation of the Bactrian distribution.

BFS

Byte File System

CMS BLOCKSIZE

The block size, in bytes, of the users' CMS minidisks.

Command

In the context of reporting performance results, any user interaction with the system being measured.

CP

Control Program. The hypervisor component of VM.

CP/CMD

For the FS7F, FS8F, and VSECICS workloads, this is the average amount of CP processor time used per command in milliseconds. For the PACE workload, this is the average CP processor time per job in seconds.

CP/CMD (H)

See CP/CMD. This is the hardware based measure.

CP CPU/CMD (V) Server

CP processor time, in milliseconds, run in the designated server machine per command.

CPU PCT BUSY (V)

CPU Percent Busy. The percentage of total available processor time used by the designated virtual machine. Total available processor time is the sum of online time for all processors and represents total processor capacity (not processor usage).

CPU SECONDS (V)

Total CPU time, in seconds, used by a given virtual machine.

CPU UTIL (V)

The percentage of time a given virtual machine spends using the CPU.

DASD IO/CMD (V)

The number of real SSCH or RSCH instructions issued to DASD, per job, used by the VSE guest in a PACE measurement.

DASD IO RATE (V)

The number of real SSCH or RSCH instructions per second that are issued to DASD on behalf of a given virtual machine.

For PACE measurements, the number of real SSCH or RSCH instructions per second issued to DASD on behalf of the VSE guest.

DASD IO TOTAL (V)

The number of real SSCH or RSCH instructions issued to DASD used by the VSE guest in a PACE measurement.

DASD PAGE RATE (V)

The number of DASD page reads per second plus DASD page writes per second that occur in a given virtual machine.

DASD RESP TIME (V)

Average DASD response time in milliseconds. This includes DASD service time plus (except for page and spool volumes) any time the I/O request is queued in the host until the requested device becomes available.

DIAGNOSE

An instruction that is used to request CP services by a virtual machine. This instruction causes a SIE interception and returns control to CP.

DIAG/CMD

The total number of DIAGNOSE instructions used per command or job.

DISPATCH LIST

The average over time of the number of virtual machines (including loading virtual machines) in any of the dispatch list queues (Q0, Q1, Q2 and Q3).

DPA

Dynamic Paging Area. The area of real storage used by CP to hold virtual machine pages, pageable CP modules and control blocks.

EDF

Enhanced Disk Format. This refers to the CMS minidisk file system.

Elapsed Time (C)

The total time, in seconds, required to execute the PACE batch workload.  
This is calculated using the timestamps that appear on the console of the VSE/ESA guest virtual machine. The time the first job started is subtracted from the time the last job ended.

ELIGIBLE LIST

The average over time of the number of virtual machines (including loading virtual machines) in any of the eligible list queues (E0, E1, E2 and E3).

EMUL ITR

Emulation Internal Throughput Rate. The average number of transactions completed per second of emulation time. This is from the EM_ITR field under TOTALITR of the RTM TRANSACT screen.

EMUL/CMD

For the FS7F, FS8F, and VSECICS workloads, this is the amount of processor time spent in emulation mode per command in milliseconds. For the PACE workload, this is the emulation processor time per job in seconds.

EMUL/CMD (H)

See EMUL/CMD. This is the hardware based measurement.

ETR

External Throughput Rate.

ETR (C)

See ETR. The external throughput rate for the VSE guest measurements.

ETR (T)

See ETR. TPNS-based calculation of ETR.

ETR RATIO

This is the ratio of the RTM-based ETR calculation and the TPNS-based ETR calculation.

External TPNS

The TPNS workload driver is run on a different system that is connected to the measured system with a communications link. See Internal TPNS.

FAST CLR/CMD

The number of fast path clears of real storage per command or job. This includes V=R and regular guests.

FCON/ESA

FCON/ESA is a program that is available from IBM that provides performance monitoring capabilities with system console operation in full screen mode. FCON/ESA can provide an immediate view of system performance or post process its own history files or VM/ESA monitor data for selected data. Threshold monitoring and user loop detection is provided. FCON/ESA also has the ability to monitor remote systems.

FREE TOTL/CMD

The number of requests for free storage per command or job. This includes V=R and regular guests.

FREE UTIL

The proportion of the amount of available free storage actually used.

FREEPGS

The total number of pages used for FREE storage (CP control blocks).

FST

File Status Table. The CMS control block that contains information about a file belonging to a minidisk or SFS directory.

GB

Gigabytes. 1024 megabytes.

GUEST SETTING

This field represents the type of VSE guest virtual machine in a PACE measurement. This fields possible values are V=V, V=F or V=R.

GUESTWT/CMD

The number of entries into guest enabled wait state per job.

GUESTWT/SEC

The number of entries into guest enabled wait state per second.  

Hardware Instrumentation

See Processor Instrumentation

IML MODE

This is the hardware IML mode used in VSE guest measurements. The possible values for this field are 370, ESA, or LPAR.

Instruction Path Length

The number of machine instructions used to run a given command, function or piece of code.

Internal Response Time

The response time as seen by CP. This does not include line or terminal delays.

Internal TPNS

The TPNS workload driver is run as one or more virtual machines in the system being measured. When used with TCP/IP, the loopback IP address is used to direct communications between the TPNS and TCP/IP stack virtual machine(s), thus eliminating the need for physical communications hardware. See External TPNS.

ISFC

Inter-System Facility for Communications

ITR

Internal Throughput Rate. This is the number of units of work accomplished per unit of processor busy time in an nonconstrained environment. For the FS7F, FS8F, and VSECICS workloads this is represented as commands per processor second. For the PACE workload, this is represented as jobs per processor minute.

ITR (H)

See ITR. This is the hardware based measure. In this case, ITR is measured in external commands per unit of processor busy time. For the FS7F, FS8F, and VSECICS workloads this is represented as commands per processor second, while for the PACE workload this is represented in jobs per processor minute.

ITR (V)

See ITR. This is the VMPRF-based measure. ITR is measured in external commands per unit of processor busy time.

ITRR

Internal Throughput Rate Ratio. This is the RTM based ITR normalized to a specific run.

ITRR (H)

See ITRR. This is the ITR (H) normalized to a specific run.

ITRR (V)

See ITRR. This is the ITR (V) normalized to a specific run.

IUCV

Inter-User Communication Vehicle. A VM generalized CP interface that helps the transfer of messages either among virtual machines or between CP and a virtual machine.

k

Multiple of 1000.

Kb

Kilobits. One kilobit is 1024 bits.

KB

Kilobytes. One kilobyte is 1024 bytes.

MASTER CP (H)

Total host (CP) state utilization for the master processor. This is calculated as MASTER TOTAL (H) - MASTER EMUL (H).

MASTER CP

Total CP processor utilization for the master processor. This is calculated as MASTER TOTAL - MASTER EMUL.

MASTER EMUL

Total emulation state utilization for the master processor. For uniprocessors this is the same as 'TOTAL EMUL' and is generally not shown.

MASTER EMUL (H)

Total emulation state utilization for the master processor. For uniprocessors this is the same as 'TOTAL EMUL' and is generally not shown. This is the hardware based calculation.

MASTER TOTAL

Total utilization of the master processor. For uniprocessor this is the same as 'TOTAL' and is generally not shown.

MASTER TOTAL (H)

Total utilization of the master processor. For uniprocessor this is the same as 'TOTAL (H)' and is generally not shown. This is the hardware based calculation.

MB

Megabytes. One megabyte is 1,048,576 bytes.

MDC AVOID

The number of DASD read I/Os per second that were avoided through the use of minidisk caching.

MDC HIT RATIO

Minidisk Cache Hit Ratio.

MDC MODS

Minidisk Cache Modifications. The number of times per second blocks were written in the cache, excluding the writes that occurred as a result of minidisk cache misses. This measure only applies to VM releases prior to VM/ESA 1.2.2.  

MDC READS (blks)

Minidisk Cache Reads. The number of times per second blocks were found in the cache as the result of a read operation. This measure only applies to VM releases prior to VM/ESA 1.2.2.  

MDC READS (I/Os)

Minidisk Cache Reads. The total number of virtual read I/Os per second that read data from the minidisk cache. This measure does not apply to VM releases prior to VM/ESA 1.2.2.  

MDC REAL SIZE (MB)

The size, in megabytes, of the minidisk cache in real storage. This measure does not apply to VM releases prior to VM/ESA 1.2.2.  

MDC WRITES (blks)

Minidisk Cache Writes. The number of CMS Blocks moved per second from main storage to expanded storage. This measure only applies to VM releases prior to VM/ESA 1.2.2.  

MDC WRITES (I/Os)

Minidisk Cache Writes. The total number of virtual write I/Os per second that write data into the minidisk cache. This measure does not apply to VM releases prior to VM/ESA 1.2.2.

MDC XSTOR SIZE (MB)

The size, in megabytes, of the minidisk cache in expanded storage.

Millisecond

One one-thousandth of a second.

Minidisk Caching

Refers to a CP facility that uses a portion of storage as a read cache of DASD blocks. It is used to help eliminate I/O bottlenecks and improve system response time by reducing the number of DASD read I/Os. Prior to VM/ESA 1.2.2, the minidisk cache could only reside in expanded storage and only applied to 4KB-formatted CMS minidisks accessed via diagnose or *BLOCKIO interfaces. Minidisk caching was redesigned in VM/ESA 1.2.2 to remove these restrictions. With VM/ESA 1.2.2, the minidisk cache can reside in real and/or expanded storage and the minidisk can be in any format. In addition to the diagnose and *BLOCKIO interfaces, minidisk caching now also applies to DASD accesses that are done using SSCH, SIO, or SIOF.

Minidisk File Cache

A buffer used by CMS when a file is read or written to sequentially. When a file is read sequentially, CMS reads ahead as many blocks as will fit into the cache. When a file is written sequentially, completed blocks are accumulated until the cache is filled and then are written out together.

ms

Millisecond.

Native

Refers to the case where an operating system is run directly on the hardware as opposed to being run as a guest on VM.

Non-shared Storage

The portion of a virtual machine's storage that is unique to that virtual machine, (as opposed to shared storage such as a saved segment that is shared among virtual machines). This is usually represented in pages.

NONPAGE RIO/CMD (V)

The number of real SSCH and RSCH instructions issued per command for purposes other than paging.

NONTRIV INT

Non-trivial Internal response time in seconds. The average response time for transactions that completed with more than one drop from Q1 or one or more drops from Q0, Q2, or Q3 per second.

Non-Spool I/Os (I)

Non-spool I/Os done by a given virtual machine. This is calculated from INDICATE USER data obtained before and after the activity being measured. The value shown is final IO - initial IO.

NPDS

No Page Data-Set. A VSE/ESA option, when running on VM/ESA as a V=V guest, that eliminates paging by VSE/ESA for improved efficiency. All paging is done by VM/ESA.

NUCLEUS SIZE (V)

The resident CP nucleus size in kilobytes.  
 
This is from the <K bytes> column on the Total Resident Nucleus line in the VMPRF System Configuration Report.  

OSA

IBM S/390 Open Systems Adapter. An integrated S/390 hardware feature that provides an S/390 system with direct access to Token Ring, Ethernet, and FDDI local area networks.

PAGE/CMD

The number of pages moved between real storage and DASD per command or job.

PAGE IO RATE (V)

The number of real SSCH or RSCH instructions issued on behalf of system paging.

PAGE IO/CMD (V)

The number of real SSCH and RSCH instructions issued per command on behalf of system paging.

Path length

See Instruction Path Length

PBT/CMD

For the FS7F, FS8F, and VSECICS workloads, this is the number of milliseconds of processor activity per command. For the PACE workload, this is the number of seconds of processor activity per job.

PBT/CMD (H)

See PBT/CMD. This is the hardware based measure.  

PCI

Program controlled interrupt.

PERCENT CP (H)

The percentage of all CPU time that is used by CP. This is calculated as CP/CMD (H) / PBT/CMD (H).

PGBLPGS

The number of system pageable pages available.

PGBLPGS/USER

The number of system pageable pages available per user.

Privileged Operation

Any instruction that must be run in supervisor state.

PRIVOP/CMD

The number of virtual machine privileged instructions simulated per command or job. This does not include DIAGNOSE instructions.

PRIVOPS (Privileged Operations)

See Privileged Operation.

Processor Instrumentation

An IBM internal tool used to obtain hardware-related data such as processor utilizations.

Processor Utilization

The percent of time that a processor is not idle.

Processors

The data field denoting the number of processors that were active during a measurement.

QUICKDSP ON

When a virtual machine is assigned this option, it bypasses the normal scheduler algorithm and is placed on the dispatch list immediately when it has work to do. It does not spend time in the eligible lists. QUICKDSP can be specified either via a CP command or in the CP directory entry.

RAID

Redundant array of independent DASD.

RAMAC

A family of IBM storage products based on RAID technology. These include the RAMAC Array Subsystem and the RAMAC Array DASD.

READS/SEC

The number of pages read per second done for system paging.

Real Storage

The amount of real storage (central storage) used for a particular measurement.

Relative Share

A relative share allocates to a virtual machine a portion of the total system resources minus those resources allocated to virtual machines with an ABSOLUTE share. A virtual machine with a RELATIVE share receives access to system resources that is proportional with respect to other virtual machines with RELATIVE shares.

RESERVE

See SET RESERVED

RESIDENT PAGES (V)

The average number of nonshared pages of central storage that are held by a given virtual machine. This is the Resid Storage Pages column in VMPRF's USER_RESOURCE_UTIL report.

RFC

Request for comments. In the context of this report, an RFC is an online document that describes a TCP/IP standard (proposed or adopted).

RIO/CMD (V)

The number of real SSCH and RSCH instructions issued per command.

RIO RATE (V)

The number of real SSCH and RSCH instructions issued per second.

RSU

Recommend Service Upgrade

RTM

Real Time Monitor A licensed program realtime monitor and diagnostic tool for performance monitoring, analysis, and problem solving.

Run ID

An internal use only name used to identify a performance measurement.

SET RESERVED (Option)

This is a CP command that can be used to allow a V=V virtual machine to have a specified minimum number of pages resident in real storage. It is used to reduce paging and improve performance for a given virtual machine.

SHARE

The virtual machine's SHARE setting. The SET SHARE command and the SHARE directory statement allow control of the percentage of system resources a virtual machine receives. These resources include processors, real storage and paging I/O capability. A virtual machine receives its proportion of these resources according to its SHARE setting. See Relative and Absolute Share.

Shared Storage

The portion of a virtual machines storage that is shared among other virtual machines (such as saved segments). This is usually represented in pages.

SHRPGS

The number of shared frames currently resident.

SIE

ESA Architecture instruction to Start Interpretive Execution. This instruction is used to run a virtual machine in emulation mode.

SIE INTCPT/CMD

The number of exits from SIE which are SIE interceptions per command or job. SIE is exited either by interception or interruption. An intercept is caused by any condition that requires CP interaction such as I/O or an instruction that has to be simulated by CP.

SIE/CMD

SIE instructions used per command or job.

S/390 Real Storage

On an IBM PC Server 500 system, the amount of real storage that is available to the System/390 processor.

TOT CPU/CMD (V) Server

The total amount of processor time, in milliseconds, for the server virtual machine(s).

TOT INT

Total Internal Response Time in seconds. Internal response time averaged over all trivial and non-trivial transactions.  

TOT INT ADJ

Total internal response time (TOT INT) reported by RTM, adjusted to reflect what the response time would have been had CP seen the actual command rate (as recorded by TPNS). This is a more accurate measure of internal response time than TOT INT. In addition, TOT INT ADJ can be directly compared to external response time (AVG LAST (T)) as they are both based on the same, TPNS-based measure of command rate.

TOT PAGES/USER

The total number of pages that are associated, on average, with each end user virtual machine. This is taken from VMPRF report UCLASS_RESOURCE UTIL and is the sum of resident storage pages, expanded storage pages, and DASD page slots for the "Users" class. This is a measure of how many unique pages are touched during execution of the workload by the average end user.

TOTAL

The total processor utilization for a given measurement summed over all processors.  

TOTAL (H)

See TOTAL. This is the hardware based measurement.

Total CPU (I)

Total CPU time, in seconds, used by a given virtual machine. This is calculated from INDICATE USER data obtained before and after the activity being measured. The value shown is final TTIME - initial TTIME.

TOTAL EMUL

The total emulation state time for all users across all online processors. This indicates the percentage of time the processors are in emulation state.

TOTAL EMUL (H)

The total emulation state time for all users across all online processors. This indicates the percentage of time the processors are in emulation state.

TPNS

Teleprocessing Network Simulator. A licensed program terminal and network simulation tool that provides system performance and response time information.

Transaction

A user/system interaction as counted by CP. For a single-user virtual machine a transaction should roughly correspond to a command. It does not include network or transmission delays and may include false transactions. False transactions can be those that wait for an external event, causing them to be counted as multiple transactions, or those that process more than one command without dropping from queue, causing multiple transactions to be counted as one.

TRACE TABLE (V)

The size in kilobytes of the CP trace table.  
This is the value of the <K bytes> column on the Trace Table line in the VMPRF System Configuration Report.  

Transaction (T)

This is the interval from the time the command is issued until the last receive prior to the next send. This includes clear screens as a result of an intervening MORE... or HOLDING condition.

TRIV INT

Trivial Internal Response Time in seconds. The average response time for transactions that complete with one and only one drop from Q1 and no drops from Q0, Q2, and Q3.

TVR

Total to Virtual Ratio. This is the ratio of total processor utilization to virtual processor utilization.

TVR (H)

See TVR. Total to Virtual Ratio measured by the hardware monitor.

T/V Ratio

See TVR

Users

The number of virtual machines logged on to the system during a measurement interval that are associated with simulated end users. This includes active and inactive virtual machines but does not include service machines.

UTIL/PROC (H)

Per processor utilization reported by the hardware.

UTIL/PROC (V)

Average utilization per processor reported VMPRF.

VIO RATE

The total number of all virtual I/O requests per second for all users in the system.

VIO/CMD

The average number of virtual I/O requests per command or job for all users in the system.

VIRT CPU/CMD (V) Server

Virtual processor time, in milliseconds, run in the designated server(s) machine per command.

VM Mode

This field is the virtual machine setting (370, XA or ESA) of the VSE guest virtual machine in PACE measurements.

VM Size

This field is the virtual machine storage size of the VSE guest virtual machine in PACE measurements.

VMPAF

Virtual Machine Performance Analysis Facility. A tool used for performance analysis of VM systems.

VMPRF

VM Performance Reporting Facility. A licensed program that produces performance reports and history files from VM/ESA or z/VM monitor data.

VSCSs

The number of virtual machines running VSCS external to VTAM during a measurement interval.

VSE Supervisor

This field is the VSE supervisor mode used in a PACE measurement.

VTAMs

The number of virtual machines running VTAM during a measurement interval.

V=F

Virtual equals fixed machine. A virtual machine that has a fixed, contiguous area of real storage. Unlike V=R, storage does not begin at page 0. For guests running V=F, CP does not page this area. Requires the PR/SM hardware feature to be installed.

V=R

Virtual equals real machine. Virtual machine that has fixed, contiguous area of real storage starting at page 0. CP does not page this area.

V=V

Virtual equals virtual machine. Default storage processing. CP pages the storage of a V=V machine in and out of real storage.

WKSET (V)

The average working set size. This is the scheduler's estimate of the amount of storage the average user will require, in pages.

WKSET (V) Server

Total working set of a related group of server virtual machine(s).

WRITES/SEC

The number of page writes per second done for system paging.

XSTOR IN/SEC

The number of pages per second read into main storage from expanded storage. This includes fastpath and non-fastpath pages.

XSTOR OUT/SEC

The number of pages per second written from main storage into expanded storage.

XSTOR/CMD

The number of pages read into main storage from expanded storage and written to expanded storage from main storage per command or job.

Contents | Previous | Next