Connect the Dots: A z13 and z/OS Dispatching Update

Transcription

Connect the Dots: A z13 and z/OS Dispatching Update
3/30/2015
Glenn Anderson
IBM Lab Services and Training
Connect the Dots: A z13 and z/OS
Dispatching Update
9.0
What I hope to cover......
What are dispatchable units of work on z/OS
How WLM manages dispatchable units of work
The role of HiperDispatch and Warning Track
Dispatching work to zIIP engines
z13 Simultaneous Multithreading (SMT)
(c) 2014 IBM Corporation
1
3/30/2015
z13 and z/OS dispatching
zIIPs
SMT
LPAR1
LPAR2
Work Units (TCBs and SRBs)
z/OS Dispatcher
WLM
Logical Processors
Logical Processors
PR/SM Dispatcher
Physical Processors
PR/SM
HiperDispatch
Warning Track
z/OS dispatchable units
There are different types of Dispatchable Units
(DU's) in z/OS
Preemptible Task (TCB)
Non Preemptible Service Request (SRB)
Preemptible Enclave Service Request (enclave SRB)
Independent - a new transaction
Dependent – extend existing address space
Work-dependent – extend existing independent enclave
(c) 2014 IBM Corporation
2
3/30/2015
z/OS dispatching work
What is a WLM transaction?
A WLM transaction represents a WLM "unit of work"
basic workload entity for which WLM collects a resource usage value
foundation for statistics presented in workload activity report
represents a single subsystem "work request"
Subsystems can implement one of three transaction types
Address Space:
WLM transaction measures all resource used by a subsystem request in a
single address space
Used by JES (a batch job), TSO (a TSO command), OMVS (a process),
STC (a started task) and ASCH (single APPC program)
Enclave:
Enclave created and destroyed by subsystem for each work request
WLM transaction measures resources used by a single subsystem
request across multiple address spaces
Exploited by subsystems - Component Broker(WebSphere), DB2, DDF,
IWEB, MQSeries Workflow, LDAP, NETV, TCP
CICS/IMS Transactions
Neither address space or enclave oriented - special type
(c) 2014 IBM Corporation
3
3/30/2015
The WLM view
Address Spaces, and the transactions inside
A/S is dispatchable unit
Enclaves are dispatchable units
Only A/S is dispatchable unit
CICS/IMS tran
Enclave SRB
transactions
transactions
Enclave SRB
CICS/IMS tran
Enclave SRB
transactions
STC
rules
Enclave SRB
CICS/IMS tran
transactions
STCHI
Vel = 50%
Imp=1
STC
rules
CICS/IMS tran
STCHI
Vel = 50%
Imp=1
WLM policy adjustment algorithm
(c) 2014 IBM Corporation
4
3/30/2015
WLM dispatching priority usage
Dispatching in an LPAR environment
LPAR1
LPAR2
Work Units (TCBs and SRBs)
z/OS Dispatcher
Logical Processors
Logical Processors
PR/SM Dispatcher
Physical Processors
PR/SM
(c) 2014 IBM Corporation
5
3/30/2015
HiperDispatch mode
 PR/SM
– Supplies topology information/updates to z/OS
– Ties high priority logicals to physicals (gives 100% share)
– Distributes remaining share to medium priority logicals
– Distributes any additional service to unparked low priority logicals
 z/OS
– Ties tasks to small subsets of logical processors
– Dispatches work to high priority subset of logicals
– Parks low priority processors that are not need or will not get service
 Hardware cache optimization occurs when a given unit of
work is consistently dispatched on the same physical CPU
HiperDispatch: z/OS part
z/OS
Node
Node
Parked
– Whether a logical processor
has high, medium or low
share
– On which book and chip the
logical processor is located
LPAR
Parked
• z/OS obtains the logical to
physical processor mapping
in Hiperdispatch mode
• z/OS creates dispatch nodes
– The idea is to have 4 high
share CPs in one node
– Each node has TCBs and
SRBs assigned to the node
– Optimizes the execution of
work units on z/OS
PR/SM
Logical to Physical
Mapping
Hardware
Books
(z13 – Drawer / Nodes)
12
(c) 2014 IBM Corporation
6
3/30/2015
RMF CPU activity report
HiperDispatch and LPAR
1
P A R T I T I O N
D A T A
R E P O R T
PAGE
z/OS V1R10
SYSTEM ID LPAR1
CONVERTED TO z/OS V1R12 RMF
DATE 04/29/2011
TIME 19.28.00
3
INTERVAL 14.59.998
CYCLE 1.000 SECONDS
MVS PARTITION NAME
LPAR1
IMAGE CAPACITY
3165
NUMBER OF CONFIGURED PARTITIONS
4
WAIT COMPLETION
NO
DISPATCH INTERVAL
DYNAMIC
--------- PARTITION DATA --------------------MSU---- -CAPPING-NAME
S
WGT DEF
ACT DEF
WLM%
LPAR1
A
494
0
582 NO
0.0
LPAR2
A
446
0
762 NO
0.0
LPAR3
A
59
0
0 NO
0.0
LPAR5
A
1
0
0 NO
0.0
*PHYSICAL*
TOTAL
NUMBER OF PHYSICAL PROCESSORS
CP
IIP
-- LOGICAL PARTITION PROCESSOR DATA -PROCESSOR- ----DISPATCH TIME DATA---NUM
TYPE
EFFECTIVE
TOTAL
32.0
CP
02.17.24.319 02.20.44.154
32.0
CP
03.01.28.607 03.04.05.167
3.0
CP
00.00.00.000 00.00.00.000
1.0
CP
00.00.00.000 00.00.00.000
00.10.58.833
------------ -----------05.18.52.927 05.35.48.155
55
53
2
GROUP NAME
LIMIT
AVAILABLE
N/A
N/A
N/A
-- AVERAGE PROCESSOR UTILIZATION PERCENTAGES -LOGICAL PROCESSORS --- PHYSICAL PROCESSORS --EFFECTIVE
TOTAL LPAR MGMT EFFECTIVE TOTAL
28.63
29.32
0.44
17.96 18.40
37.81
38.35
0.34
23.72 24.06
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
1.44
1.44
--------- ----2.21
41.68 43.90
Total LPAR weight = 1000
LPAR1 494/1000 = .494 * 53 CPs = 26.18 CPs
LPAR2 446/1000 = .446 * 53 CPs = 23.64 CPs
LPAR1 = 25 VH and 2 VM at 59% share (27 logicals unparked)
LPAR2 = 23 VH and 1 VM at 64% share (24 logicals unparked)
____________
51 logicals unparked
(c) 2014 IBM Corporation
53 physicals
7
3/30/2015
Dispatching in an LPAR environment
LPAR1
LPAR2
Work Units (TCBs and SRBs)
z/OS Dispatcher
Logical Processors
Logical Processors
PR/SM Dispatcher
Physical Processors
PR/SM
Warning track
(c) 2014 IBM Corporation
8
3/30/2015
Warning track
Latent demand: LPAR Busy vs MVS Busy
(c) 2014 IBM Corporation
9
3/30/2015
Warning track statistics
WLM Topology Report Tool
20
(c) 2014 IBM Corporation
10
3/30/2015
Dispatching zIIP eligible work
LPAR1
LPAR2
Work
(TCBs and
SRBs)units
zIIPUnits
eligible
work
z/OS Dispatcher
Logical
Processors
zIIP
logical
processors
zIIP
Logicallogical
Processorsprocessors
PR/SM Dispatcher
Physical Processors
zIIP physical
processors
PR/SM
IBM z Integrated Information Processor (zIIP)
on the z13
 The IBM z13 continues to support the z Integrated Information
Processor (zIIP) which can take advantage of the optional simultaneous
multithreading (SMT) technology capability. SMT allows up to two active
instruction streams per core, each dynamically sharing the core's
execution resources.
̶
With the multithreading function enabled, the performance capacity of the
zIIP processor is expected to be up to 1.4 times the capacity of these
processors on the zEC12
 The rule for the CP to zIIP purchase ratio is that for every CP purchased,
up to two zIIPs may be purchased
 zAAP eligible workloads such as Java and XML, can run on zIIPs using
zAAP on zIIP processing
 zAAPs are no longer supported on the z13
(c) 2014 IBM Corporation
11
3/30/2015
Current IBM exploitation of zAAPs and zIIPs
Specialty CP
Eligible
Major Users
zAAP or zIIP
on z13
Any Java Execution Websphere
CICS
Native apps
XMLSS
zIIP
Enclave SRBs
DRDA over TCPIP
DB2 Parallel Query
DB2 Utilities Load, Reorg, Rebuild
DB2 V9 z/OS remote native SQL procedures
TCPIP - IPSEC
XMLSS
zIIP Assisted HiperSockets Multiple Write
Virtual Tape Facility Mainframe (VTFM) Software
z/OS Global Mirror (XRC), System Data Mover (SDM)
z/OS CIM Server
RMF Mon III
OMEGAMON on z/OS and DB2
IMS Ver 8
What is "Needs Help?"
Determination zIIP or zAAP work is being delayed and additional
resources should help process the work
ƒRequires xxPHONORPRIORITY=YES to be set
If help is required:
ƒThe zxxP CP signals waiting zxxP to help
ƒWhen all zxxP CPs are busy the zxxP asks for help from the GCP
–All available speciality engines (of the same type) must be busy before
help is asked of the GCPs
IF the zxxPs needs help and all zxxPs are busy help is obtained from 1 GCP
IF zxxPs continue to need help additional CPs may be asked to help
ƒHelp is always provided in dispatch priority order
(c) 2014 IBM Corporation
12
3/30/2015
Specialty CP work running in a WLM service class
REPORT BY: POLICY=WLMPOL
WORKLOAD=BAT_WKL
SERVICE CLASS=BATSPEC
TRANSACTIONS
TRANS-TIME HHH.MM.SS.TTT --DASD I/O-- ---SERVICE---AVG
0.98 ACTUAL
6.520 SSCHRT 11.5 IOC
8326
MPL
0.98 EXECUTION
6.128 RESP
7.0 CPU
662386
ENDED
10 QUEUED
391 CONN
6.9 MSO
0
END/S
0.17 R/S AFFIN
0 DISC
0.0 SRB
965
#SWAPS
0 INELIGIBLE
0 Q+PEND
0.1 TOT
671677
EXCTD
0 CONVERSION
0 IOSQ
0.0 /SEC
11195
AVG ENC 0.00 STD DEV
0
GOAL: EXECUTION VELOCITY 35.0%
RESPONSE TIME EX
PERF
VELOCITY MIGRATION:
AVG
I/O MGMT
RESOURCE GROUP=BATMAXRG
SERVICE TIMES ---APPL %--CPU
24.7
CP
0.97
SRB
0.0
AAPCP
0.01
RCT
0.0
IIPCP
0.00
IIT
0.0
HST
0.0
AAP
40.27
AAP
24.2
IIP
0.00
IIP
0.0
99.2%
INIT MGMT 92.2%
------ USING% ----- ------------ EXECUTION DELAYS % -------
SYSTEM
SYSD
--N/A--
VEL% INDX ADRSP
CPU
AAP
IIP
I/O
TOT
CPU
99.2
0.8 45.9
0.0
3.9
0.4
0.4
0.4
1.0
RMF report is at 1 minute interval
DB2 parallel query, enclave SRBs and zIIPs
Host 1
Query CP
Parallelism
DU
DU
DU
DU
IO
IO
IO
IO
Have been independent enclave
SRBs to be zIIP eligible. Beginning in
z/OS R11 the child tasks are now
work-dependent enclaves.
Portions of complex query
arrive on participant systems,
classified under "DB2" rules,
and run in enclave SRBs, so
zIIP eligible
PARTITIONED TABLESPACE
Complex
query
originates
here
DU
IO
Sysplex Query Parallelism
Host 1
Host 2
DU
DU
DU
IO
IO
IO
DU
IO
DU
IO
DU
IO
Host 3
DU
DU
IO
IO
DU
IO
DU
IO
DU
IO
PARTITIONED
TABLESPACE
(c) 2014 IBM Corporation
13
3/30/2015
Work-Dependent enclaves
DB2
Address Space
Work-Dependent Enclave
zIIP Offload = Y %
create
Independent Enclave
zIIP Offload = X %
create
Work-Dependent Enclave
zIIP Offload = Z %
Managed as one transaction, represented by Independent
Enclave
Implement a new type of enclave named “Work-Dependent” as an extension
of an Independent Enclave. A Work-Dependent enclave becomes part of the
Independent Enclave’s transaction but allows to have its own set of attributes
(including zIIP offload percentage)
DDF and work-dependent enclaves
ssnmDIST (DDF)
DDF production
requests
Enclave SRB
PC-call to DBM1
In cases where DRDA applications
create extended duration work
threads in DB2, for example through
extensive use of held cursors, the
zIIP utilization levels can become
more variable. DB2 and DDF now
may use work-dependent enclaves
in this situation to control this
variability. See APAR PM28626.
Create Enclave
Schedule SRB
PC-call to DBM1
DDF default
requests
DDF
rules
DDFPROD
RT=85%, 2s
Imp=1
SMF
72
Enclave SRB
DDFDEF
STC
rules
(c) 2014 IBM Corporation
RT=5s avg
Imp=3
STCHI
Vel = 50%
Imp=1
SMF 72
SMF 30
SMF 72
14
3/30/2015
Work-dependent enclaves in SDSF
zIIP processors and simultaneous multithreading
LPAR1
LPAR2
Work
(TCBs and
SRBs)units
zIIPUnits
eligible
work
z/OS Dispatcher
Logical
Processors
zIIP
logical
cores with
1 or 2 threads
zIIP logical cores with
Logical Processors
1 or 2 threads
PR/SM Dispatcher
PR/SM
(c) 2014 IBM Corporation
zIIP physical
with
Physicalcores
Processors
1 or 2 threads
15
3/30/2015
z13 - Simultaneous Multithreading (SMT)
“Simultaneous multithreading (SMT) permits
multiple independent threads of execution to
better utilize the resources provided by modern
processor architectures.”*
●
With z13, SMT allows up to two instructions per
core to run simultaneously to get better overall
throughput
●
SMT is designed to make better use
of processors
●
80
50
On z/OS, SMT is available for
zIIP processing:
●
–
Two concurrent threads are available per core
and can be turned on or off
–
Capacity (throughput) usually increases
–
Performance may in some cases be superior
using single threading
Two lanes process more
traffic overall
Note: Speed limit signs for illustration only
*Wikipedia®
z13 - SMT Exploitation
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Appl
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Thr
Core
Core
Core
Core
Core
z/OS
Core
z/OS
MT Aware
PR/SM Hypervisor
Appl
Appl
Thr
Thr
Core
Physical Hardware
Thr
Thr
Core
Thr
Thr
MT Ignorant
Appl
Appl
Thr
Thr
Core
Core
• SMT Aware OS informs PR/SM that it intends to exploit SMT
– PR/SM can dispatch any OS core to any physical core
– OS controls the whole core – must follow rules
• Maximize core throughput (Drive cores with high Thread Density [2] )
• Maximize core availability (Meet workload goals using fewest cores )
• SMT is transparent to applications
• LOADxx and IEAOPTxx parmlib options to enable SMT on z/OS:
– LOADxx:
– IEAOPTxx:
PROCVIEW CORE|CPU
MT_ZIIP_MODE={1 | 2}
32
(c) 2014 IBM Corporation
16
3/30/2015
What is new with multithreading support?
New support
LOADxx: PROCVIEW CPU
•
LOADxx: PROCVIEW CORE
This the existing mode that you
• z/OS is core-aware and maps CPUs to cores.
are used to
• Core syntax for CPU/core related commands.
z/OS does only know
IPL required
processors (CPUs)
•
MT1 Mode
MT2 Mode
IEAOPTxx:
MT_ZIIP_MODE = 1
IEAOPTxx:
MT_ZIIP_MODE = 2
SET OPT=yy
Possible on any supported hardware
z13 only
MT1 performance & capacity
MT2 performance & cap.
HiperDispatch=No possible
HiperDispatch=Yes enforced
* Statements regarding IBM future direction and intent are subject to change or withdrawal, and represent goals and objectives only.
33
New terminology for SMT…
■
■
z/OS logical processor (CPU)
 Thread
■ A thread implements (most of) the System z processor architecture
■ z/OS dispatches work units on threads
■ In MT mode two threads are mapped to a logical core
Processor core
 Core
■ PR/SM dispatches logical core on a physical core
■
■
■
MT1 Equivalent Time (MT1ET)
■ z/OS CPU times are normalized to the time it would have taken to run same
work in MT-1 mode
■
■
■
Thread density 1 (TD1) when only a single thread runs on a core
Thread density 2 (TD2) when both threads run on a core
ASCB, ASSB, …, SMF30, SMF32, SMF7x, …
You will usually not see the term MT1ET because it is implied
Several new metrics to describe how efficiently core resources could be utilized…
* Statements regarding IBM future direction and intent are subject to change or withdrawal, and represent goals and objectives only.
(c) 2014 IBM Corporation
17
3/30/2015
…and several new metrics for SMT…
■
New metrics:
■
■
WLM/RMF:
RMF:
Capacity Factor (CF), Maximum Capacity Factor (mCF)
Average Thread Density, Productivity (PROD)
 How are the new metrics derived?
 Hardware provides metrics (counters) describing the efficiency of
processor (cache use/misses, number instructions when one or two
threads were active…)
 LPAR level counters are made available to the OS
 MVS HIS component and supervisor collect LPAR level counters. HIS
provides HISMT API to compute average metrics between “previous”
HISMT invocation and “now” (current HISMT invocation)
 System components (WLM/SRM, monitors such as RMF) retrieve
metrics for management and reporting
* Statements regarding IBM future direction and intent are subject to change or withdrawal, and represent goals and objectives only.
z13 - z/OS SMT Metrics
• Capacity Factor (CF)
– How much work core actually completes for a given workload mix at
current utilization - relative to single thread
– MT-1 Capacity Factor is 1.0 (100%)
– MT-2 Capacity Factor is workload dependent
Actual MT-2
Efficiency
• Maximum Capacity Factor (mCF)
– How much work a core can complete for a given
workload mix at most
Estimated max
MT-2 Efficiency
• Core Busy Time
– Time any thread on the core is executing instructions
when core is dispatched to physical core
• Average Thread Density
– Average number of executing threads during Core Busy Time (Range: 1.0 - 2.0)
• Productivity
– Core Busy Time Utilization (percentage of used capacity)
for a given workload mix
– Productivity represents capacity in use (CF) relative to
capacity total (mCF) during Core Busy Time.
• Core Utilization
– Capacity in use relative to capacity total over
some time interval
– Calculated as Core Busy Time x Productivity
(c) 2014 IBM Corporation
% Used MT-2 Core Capacity
during Core Busy Time
% Used MT-2 Core Capacity
during Measurement Interval
36
18
3/30/2015
z13 – SMT: Postprocessor CPU Activity Report
• PP CPU activity report displayed in “old” format when SMT is active
• PP CPU activity report provides new metrics when SMT is active
– MT Productivity and Utilization of each logical core
– MT Multi-Threading Analysis section displays MT Mode, MT Capacity Factors
and average Thread Density
• One data line in PP CPU activity report represents one thread (CPU)
– CPU NUM designates the logical core
• Some metrics like TIME % ONLINE and LPAR BUSY provided at core granularity only
z/OS V2R1
C P U A C T I V I T Y
SYSTEM ID CB8B
DATE 02/02/2015
RPT VERSION V2R1 RMF
TIME 11.00.00
--------------- TIME % -------------- MT % ---- LOG PROC
ONLINE
LPAR BUSY MVS BUSY PARKED
PROD
UTIL
SHARE %
100.00
68.07
67.94
0.00
100.00 68.07 100.0 HIGH
100.00
46.78
46.78
0.00
100.00 46.78
52.9 MED
---CPU--NUM TYPE
0
CP
1
CP
...
TOTAL/AVERAGE
A
IIP
100.00
B
IIP
100.00
8.66
48.15
38.50
54.17
41.70
35.66
32.81
26.47
0.00
0.00
0.00
0.00
100.00
85.84
8.66
41.33
152.9
100.0
HIGH
85.94
33.09
100.0
HIGH
INTERVAL 15.00.004
CYCLE 1.000 SECONDS
--I/O INTERRUPTS–
RATE
% VIA TPI
370.1
13.90
5.29
16.93
375.3
13.95
...
MT-2 core
TOTAL/AVERAGE
29.48
23.23
86.47 25.39 386.7
capacity used
------------ MULTI-THREADING ANALYSIS --------------CPU TYPE
MODE
MAX CF
CF
AVG TD
CP
1
1.000
1.000
1.000
Productivity of logical core while
IIP
2
1.485
1.279
1.576
dispatched to physical core
37
z13 – SMT: Postprocessor Workload Activity Report
W O R K L O A D
z/OS V2R1
SYSPLEX UTCPLXCB
RPT VERSION V2R1 RMF
A C T I V I T Y
DATE 02/02/2015
TIME 11.00.00
INTERVAL 15.00.004
MODE = GOAL
REPORT BY: POLICY=BASEPOL
-TRANSACTIONSAVG
790.12
MPL
790.12
ENDED
9173
END/S
10.19
#SWAPS
6087
EXCTD
15860
AVG ENC
4.00
REM ENC
0.00
MS ENC
0.00
►
TRANS-TIME HHH.MM.SS.TTT
ACTUAL
27.787
EXECUTION
15.761
QUEUED
1
R/S AFFIN
0
INELIGIBLE
0
CONVERSION
0
STD DEV
8.40.915
--DASD I/O-SSCHRT 3975
RESP
2.8
CONN
1.4
DISC
1.2
Q+PEND
0.1
IOSQ
0.0
MT-1 equivalent service units
and service times
---SERVICE--IOC
4233K
CPU
306468K
MSO
0
SRB
97415K
TOT
408116K
/SEC
453461
ABSRPTN
TRX SERV
SERVICE TIME
CPU 4659.039
SRB 1513.624
RCT
0.622
IIT
21.116
HST
0.000
AAP
N/A
IIP 1308.346
---APPL %--CP
542.89
AAPCP
0.00
IIPCP
2.00
AAP
IIP
N/A
97.84
574
574
Pre SMT:
 Service time : Logical processor CPU time
 APPL%:
Percentage of logical processor capacity
►
SMT mode active:
 Service time:
 APPL% :
MT-1 equivalent CPU time
Percentage of maximum core capacity calculated as
For MT Mode = 1  mCF = 1
(c) 2014 IBM Corporation
Percentage of maximum
core capacity used
3/30/2015
38
19
3/30/2015
Transitioning into MT2 mode: WLM considerations (1)
• Less overflow from zIIP to CPs may occur because
• zIIP capacity increases, and
• number of zIIP CPUs double
• CPU time and CPU service variability may increase, because
• Threads which are running on a core at the same time influence each
other
• Threads may be dispatched at TD1 or TD2
• Sysplex workload routing: routing recommendation may change because
• zIIP capacity will be adjusted with the mCF to reflect MT2 capacity
• mCF may change as workload or workload mix changes
* Statements regarding IBM future direction and intent are subject to change or withdrawal, and represent goals and objectives only.
Transitioning into MT2 mode: WLM Considerations (2)
• Goals should be verified for zIIP-intensive work, because
• The number of zIIP CPUs double and the achieved velocity may
change
• “Chatty” (frequent dispatches) workloads may profit because there
is a chance of more timely dispatching
• More capacity is available
• Any single thread will effectively run at a reduced speed and the
achieved velocity will be lower.
Affects processor speed bound work, such as single threaded Java
batch
* Statements regarding IBM future direction and intent are subject to change or withdrawal, and represent goals and objectives only.
(c) 2014 IBM Corporation
20
3/30/2015
Work unit queue distribution: Mon I CPU report
SYSTEM ADDRESS SPACE AND WORK UNIT ANALYSIS
---------NUMBER OF ADDRESS SPACES--------QUEUE TYPES
MIN
MAX
AVG
IN
IN READY
OUT READY
OUT WAIT
LOGICAL OUT RDY
LOGICAL OUT WAIT
550
4
0
0
0
178
1,008
438
1
0
628
634
594.8
22.7
0.0
0.0
10.3
589.4
281
708
97
0
43
284
763
98
1
97
282.0
736.4
97.9
0.0
68.0
ADDRESS SPACE TYPES
BATCH
STC
TSO
ASCH
OMVS
---------NUMBER OF WORK UNITS------------CPU TYPES
MIN
MAX
AVG
CP
AAP
IIP
444
22
0
888
33
0
555.5
28.8
0.0
-----------------------DISTRIBUTION OF IN-READY WORK UNIT QUEUE------------NUMBER OF
0
10
20
30
40
50
60
70
80
90
100
WORK UNITS
(%)
|....|....|....|....|....|....|....|....|....|....|
<=
=
=
=
<=
<=
<=
<=
<=
<=
<=
<=
<=
<=
<=
>
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
1
2
3
5
10
15
20
30
40
60
80
100
120
150
150
55.5
4.4
4.0
3.7
7.1
14.7
8.0
1.9
0.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>
>>>
>>
>>>>
>>>>>>>>
>>>>>
>
>
16 Buckets representing
the work unit count with
regard to the number of
online processors
N = NUMBER OF PROCESSORS ONLINE UNPARKED (22.4 ON AVG)
Work unit count on
CPU type level
What I hope I covered…..
What are dispatchable units of work on z/OS
How WLM manages dispatchable units of work
The role of HiperDispatch and Warning Track
Dispatching work to zIIP engines
z13 and Simultaneous Multithreading (SMT)
(c) 2014 IBM Corporation
21
3/30/2015
IBM Notice Regarding Specialty Engines
(c) 2014 IBM Corporation
22