Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX

Precisely
PreciselyPrecisely
Continued Innovation
in IBM z/System Sort
Optimization with
Syncsort MFX
Denise Tabor | Senior Product Manager Director
Alissa Margulies | Principal Sales Engineer
Today’s Agenda
• Syncsort MFX overview
• Tips for Syncsort MFX optimization
• Important enhancements for Syncsort MFX
customers
2
Syncsort MFX
IBM® z Integrated Information
Processor (zIIP):
• Helps improve GP utilization
• Allows customers to purchase additional processing
power
• Removes IBM software charges on zIIP capacity
Bottom line:
• Pay once, use often, without any additional cost
• Eligible workloads moved to zIIP reduce license
costs
• HOWEVER, MOST workloads are not enabled to
run on zIIP
4
IBM zIIP Engines
Syncsort MFX
• Exploits z/OS hardware and software AND zIIP
• Reduces CPU utilization for sort operations
• Optimizes I/O activity
Delivers
• Improved processor and DASD capacity
• Reduced cost
5
Syncsort MFX
Reduces CPU time
and I/O activity
Syncsort ZPSaver
zIIP-enabled sort
operations
Syncsort MFX
PipeSort
Reduces elapsed time
6
Syncsort MFX
Almost 50 years of continual
development and enhancements
Performance
The high-performance
sort/copy/join solution
that delivers better
performance and
saves money
Proven Solution
Improves sort performance while
optimizing overall system efficiency
zIIP Offload
Sort workloads can be directed to
the zIIP, thereby lowering the CPU
time and costs
Encryption
Enhanced security and compliance
with regulations such as GDPR
7
Tuning Syncsort MFX
Data Manipulation
Application &
Environment Options
9
Tuning Objectives for Syncsort
MFX
10
• Tuning sort requires evaluation of your current applications and defining
your objectives
• Consider what you are willing to trade in order to get better performance.
• What is the primary outcome/goal?
• Reduced CPU?
• Reduced Elapsed time?
• Avoiding contention with other workloads?
• Reducing DASD contention?
• Are you willing to:
• Trade CPU for elapsed time improvements and visa versa?
• Reschedule your sort jobs?
• Move data sets to isolate DASD to be used by the sort?
• Pass run-time parameters to the sort?
• Change priorities of applications?
Tuning Options and Recommendation
11
Control Statements
• Ensure you are using the optimal
control statements for an
application.
• Evaluate errors in control
statements that can adversely
affect performance
Optimization Mode
• Select the proper mode for
the desired outcome
• Mode selection may affect
other areas of performance.
• Performance outcome
between modes may need
some experimentation.
Virtual Storage
• Most critical resource
in determining how well the
sort will run.
• One of the most dangerous
• Using too much can lead to
system storage shortages and
system outages.
Tuning Options and Recommendation
12
Rescheduling Work
• Evaluate your overall concurrent
workload
• Is your CPU capacity, real
storage and I/O resources
impacted?
• Determine if your sort work needs
to be rescheduled to a quieter
period.
PARASORT
• PARASORT improves the elapsed
time performance for sorts whose
input is a multi-volume tape data set
and/or concatenated tape data set.
• Uses parallel processing of the
SORTIN input volumes.
• Results in up to 33% reduction in
elapsed time
FILESIZE Estimates
• Use of the FILSZ
parameter provides the sort with
an estimate of the amount
of data to be sorted
• Can significantly improve
the optimization
and performance of very
large sorts.
Syncsort MFX Data Manipulation
Reformatting Records
• INREC, OUTREC and OUTFIL OUTREC
control statements.
• Used when information
in the input record is not required
by the applications, or the data
needs to be a different format.
• Allows you to:
• Delete or repeat segments of record
• Insert new field
• Convert data
• Perform arithmetic operations with
numeric fields and/or constants
• Perform MIN/MAX functions on
numeric data
• Change RECFM of output data set
from fixed to variable or the reverse
13
Record Selection
• INPUT PHASE: Selection of records
in the order in which they appear in
the input data set.
• OUTPUT PHASE: Selection of
records seen in sorted sequence.
• These selections can be specified:
• Skip the first “n” number of records
• Stop after processing “n” number
of records
• Include/omit records based
on comparisons of the contents of
one or more fields within the record
• Include a sample of “m”
records after an interval of every “n”
records
• Distribute the records in
rotation among all of the files in an
OUTFIL group
• Create a file that contains
only those records that were
not included in any other OUTFIL
Syncsort MFX Data Manipulation
14
Summation
• Special processing done on records
with equal sort keys.
• Detailed records are replaced with
a summary record, containing sum,
average, maximum or minimum
values
• Detail records can be written to a
separate data set
Report Writer
• Generate ad-hoc or
scheduled reports.
• Easy to use functionality with powerful
record selection and formatting
Join Records
• Records created by joining 2 files
that contain a common join key.
• Join processing produces 3 types
of records:
• Paired records,
• Unpaired from the first file
• Unpaired from the second file
Additional
Syncsort MFX
product
options
Syncsort
PipeSort
Can simultaneously execute up to
eight differently sequenced sorts
from a single pass of the input data
Syncsort
PROCSort
High performance, transparent
replacement for the SAS®-provided
PROC SORT.
Presentation name
15
Syncsort
ZPSaver
A set of enhanced technologies to
offload copy, SMS compression, and
sort processing to zIIP processors
Recent
Enhancements…
IBM Z Sort Accelerator Support
(delivered)
17
• IBM’s Integrated Accelerator for Z Sort
• New coprocessor designed for the z15
• Accelerates internal sorts
• Precisely partnership w/IBM
• Worked with HW architects
• Developed new algorithms in Syncsort MFX
• Results
• Sort performance improvements
IBM Z Pervasive Encryption Support
(delivered)
18
• The IBM Z Pervasive Encryption enables
• Powerful encryption of data in-flight and at-rest
• Highly secure ways to help deal with today’s compliance and
regulatory requirements
• Requires additional resources
• Consumes processor cycles
• Forces Syncsort MFX to use less performant I/O methods
(BSAM)
• Solution
• For Basic/Large datasets, we can continue to use our low-level
I/O
MFX Operational Visibility
19
• Customers need more insights into their sort
workloads
• Unexpected delays
• Diagnostic information is frequently needed
• Rerun a job and just to gather “debug” information
• Provide more visibility
• Expedite/streamline/improve troubleshooting for large/important
jobs.
• Examine both real-time and historical data
• Port the data to an analytical tool for analysis
Questions?
Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX
1 sur 21

Recommandé

Unlock Cost Savings for your IBM z Systems Environment par
Unlock Cost Savings for your IBM z Systems EnvironmentUnlock Cost Savings for your IBM z Systems Environment
Unlock Cost Savings for your IBM z Systems EnvironmentPrecisely
49 vues29 diapositives
Evaluate the Benefits of TFP for Your IBM Z System par
Evaluate the Benefits of TFP for Your IBM Z SystemEvaluate the Benefits of TFP for Your IBM Z System
Evaluate the Benefits of TFP for Your IBM Z SystemPrecisely
55 vues24 diapositives
What’s Going on in Your 4HRA? par
What’s Going on in Your 4HRA?What’s Going on in Your 4HRA?
What’s Going on in Your 4HRA?Precisely
358 vues30 diapositives
New Mainframe Developments and How Syncsort Helps You Exploit Them par
New Mainframe Developments and How Syncsort Helps You Exploit ThemNew Mainframe Developments and How Syncsort Helps You Exploit Them
New Mainframe Developments and How Syncsort Helps You Exploit ThemPrecisely
433 vues20 diapositives
Best Practices – Extreme Performance with Data Warehousing on Oracle Database par
Best Practices – Extreme Performance with Data Warehousing on Oracle DatabaseBest Practices – Extreme Performance with Data Warehousing on Oracle Database
Best Practices – Extreme Performance with Data Warehousing on Oracle DatabaseEdgar Alejandro Villegas
2.8K vues47 diapositives
Introducing MFX for z/OS 2.1 & ZPSaver Suite par
Introducing MFX for z/OS 2.1 & ZPSaver SuiteIntroducing MFX for z/OS 2.1 & ZPSaver Suite
Introducing MFX for z/OS 2.1 & ZPSaver SuitePrecisely
2.1K vues17 diapositives

Contenu connexe

Similaire à Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX

Basics of micro controllers for biginners par
Basics of  micro controllers for biginnersBasics of  micro controllers for biginners
Basics of micro controllers for biginnersGerwin Makanyanga
290 vues46 diapositives
Cobol performance tuning paper lessons learned - s8833 tr par
Cobol performance tuning paper   lessons learned - s8833 trCobol performance tuning paper   lessons learned - s8833 tr
Cobol performance tuning paper lessons learned - s8833 trPedro Barros
680 vues35 diapositives
Dynamics CRM high volume systems - lessons from the field par
Dynamics CRM high volume systems - lessons from the fieldDynamics CRM high volume systems - lessons from the field
Dynamics CRM high volume systems - lessons from the fieldStéphane Dorrekens
1.6K vues40 diapositives
IMS04 BMC Software Strategy and Roadmap par
IMS04   BMC Software Strategy and RoadmapIMS04   BMC Software Strategy and Roadmap
IMS04 BMC Software Strategy and RoadmapRobert Hain
791 vues35 diapositives
How to Improve RACF Performance (v0.2 - 2016) par
How to Improve RACF Performance (v0.2 - 2016)How to Improve RACF Performance (v0.2 - 2016)
How to Improve RACF Performance (v0.2 - 2016)Rui Miguel Feio
1.3K vues32 diapositives
Sparc t4 systems customer presentation par
Sparc t4 systems customer presentationSparc t4 systems customer presentation
Sparc t4 systems customer presentationsolarisyougood
937 vues27 diapositives

Similaire à Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX(20)

Cobol performance tuning paper lessons learned - s8833 tr par Pedro Barros
Cobol performance tuning paper   lessons learned - s8833 trCobol performance tuning paper   lessons learned - s8833 tr
Cobol performance tuning paper lessons learned - s8833 tr
Pedro Barros680 vues
Dynamics CRM high volume systems - lessons from the field par Stéphane Dorrekens
Dynamics CRM high volume systems - lessons from the fieldDynamics CRM high volume systems - lessons from the field
Dynamics CRM high volume systems - lessons from the field
IMS04 BMC Software Strategy and Roadmap par Robert Hain
IMS04   BMC Software Strategy and RoadmapIMS04   BMC Software Strategy and Roadmap
IMS04 BMC Software Strategy and Roadmap
Robert Hain791 vues
How to Improve RACF Performance (v0.2 - 2016) par Rui Miguel Feio
How to Improve RACF Performance (v0.2 - 2016)How to Improve RACF Performance (v0.2 - 2016)
How to Improve RACF Performance (v0.2 - 2016)
Rui Miguel Feio1.3K vues
Sparc t4 systems customer presentation par solarisyougood
Sparc t4 systems customer presentationSparc t4 systems customer presentation
Sparc t4 systems customer presentation
solarisyougood937 vues
Session 6638 - The One-Day CICS Transaction Server Upgrade: Migration Conside... par nick_garrod
Session 6638 - The One-Day CICS Transaction Server Upgrade: Migration Conside...Session 6638 - The One-Day CICS Transaction Server Upgrade: Migration Conside...
Session 6638 - The One-Day CICS Transaction Server Upgrade: Migration Conside...
nick_garrod1.7K vues
Maaz Anjum - IOUG Collaborate 2013 - An Insight into Space Realization on ODA... par Maaz Anjum
Maaz Anjum - IOUG Collaborate 2013 - An Insight into Space Realization on ODA...Maaz Anjum - IOUG Collaborate 2013 - An Insight into Space Realization on ODA...
Maaz Anjum - IOUG Collaborate 2013 - An Insight into Space Realization on ODA...
Maaz Anjum638 vues
Approximation techniques used for general purpose algorithms par Sabidur Rahman
Approximation techniques used for general purpose algorithmsApproximation techniques used for general purpose algorithms
Approximation techniques used for general purpose algorithms
Sabidur Rahman476 vues
Ashnik EnterpriseDB PostgreSQL - A real alternative to Oracle par Ashnikbiz
Ashnik EnterpriseDB PostgreSQL - A real alternative to Oracle Ashnik EnterpriseDB PostgreSQL - A real alternative to Oracle
Ashnik EnterpriseDB PostgreSQL - A real alternative to Oracle
Ashnikbiz4.6K vues
PostgreSQL 10: What to Look For par Amit Langote
PostgreSQL 10: What to Look ForPostgreSQL 10: What to Look For
PostgreSQL 10: What to Look For
Amit Langote847 vues
Maximizing the Value of IBM's New Mainframe Pricing Model with Syncsort Elevate par Precisely
Maximizing the Value of IBM's New Mainframe Pricing Model with Syncsort Elevate Maximizing the Value of IBM's New Mainframe Pricing Model with Syncsort Elevate
Maximizing the Value of IBM's New Mainframe Pricing Model with Syncsort Elevate
Precisely284 vues
Data Virtualization Reference Architectures: Correctly Architecting your Solu... par Denodo
Data Virtualization Reference Architectures: Correctly Architecting your Solu...Data Virtualization Reference Architectures: Correctly Architecting your Solu...
Data Virtualization Reference Architectures: Correctly Architecting your Solu...
Denodo 847 vues
Presentation db2 best practices for optimal performance par solarisyougood
Presentation   db2 best practices for optimal performancePresentation   db2 best practices for optimal performance
Presentation db2 best practices for optimal performance
solarisyougood4.8K vues
AWS re:Invent 2016| DAT318 | Migrating from RDBMS to NoSQL: How Sony Moved fr... par Amazon Web Services
AWS re:Invent 2016| DAT318 | Migrating from RDBMS to NoSQL: How Sony Moved fr...AWS re:Invent 2016| DAT318 | Migrating from RDBMS to NoSQL: How Sony Moved fr...
AWS re:Invent 2016| DAT318 | Migrating from RDBMS to NoSQL: How Sony Moved fr...
InfiniFlux vs_RDBMS par InfiniFlux
InfiniFlux vs_RDBMSInfiniFlux vs_RDBMS
InfiniFlux vs_RDBMS
InfiniFlux593 vues

Plus de Precisely

Prozesseffizienz im Finanzbereich durch Automatisierung - am Beispiel "Comp... par
Prozesseffizienz im Finanzbereich durch Automatisierung -   am Beispiel "Comp...Prozesseffizienz im Finanzbereich durch Automatisierung -   am Beispiel "Comp...
Prozesseffizienz im Finanzbereich durch Automatisierung - am Beispiel "Comp...Precisely
3 vues22 diapositives
Data Integrity for Banking and Financial Services par
Data Integrity for Banking and Financial ServicesData Integrity for Banking and Financial Services
Data Integrity for Banking and Financial ServicesPrecisely
8 vues26 diapositives
Unlock the Power of Mainframe Data for Democratized Cloud Analytics par
Unlock the Power of Mainframe Data for Democratized Cloud AnalyticsUnlock the Power of Mainframe Data for Democratized Cloud Analytics
Unlock the Power of Mainframe Data for Democratized Cloud AnalyticsPrecisely
7 vues25 diapositives
Optimierung von Stammdatenprozessen - ein Bericht aus der Praxis par
Optimierung von Stammdatenprozessen - ein Bericht aus der PraxisOptimierung von Stammdatenprozessen - ein Bericht aus der Praxis
Optimierung von Stammdatenprozessen - ein Bericht aus der PraxisPrecisely
9 vues29 diapositives
Data Integrity for Banking and Financial Services par
Data Integrity for Banking and Financial ServicesData Integrity for Banking and Financial Services
Data Integrity for Banking and Financial ServicesPrecisely
78 vues26 diapositives
Automate Evolve Training: Excel Workflow for Automatic Routing par
Automate Evolve Training: Excel Workflow for Automatic RoutingAutomate Evolve Training: Excel Workflow for Automatic Routing
Automate Evolve Training: Excel Workflow for Automatic RoutingPrecisely
12 vues28 diapositives

Plus de Precisely(20)

Prozesseffizienz im Finanzbereich durch Automatisierung - am Beispiel "Comp... par Precisely
Prozesseffizienz im Finanzbereich durch Automatisierung -   am Beispiel "Comp...Prozesseffizienz im Finanzbereich durch Automatisierung -   am Beispiel "Comp...
Prozesseffizienz im Finanzbereich durch Automatisierung - am Beispiel "Comp...
Precisely3 vues
Data Integrity for Banking and Financial Services par Precisely
Data Integrity for Banking and Financial ServicesData Integrity for Banking and Financial Services
Data Integrity for Banking and Financial Services
Precisely8 vues
Unlock the Power of Mainframe Data for Democratized Cloud Analytics par Precisely
Unlock the Power of Mainframe Data for Democratized Cloud AnalyticsUnlock the Power of Mainframe Data for Democratized Cloud Analytics
Unlock the Power of Mainframe Data for Democratized Cloud Analytics
Precisely7 vues
Optimierung von Stammdatenprozessen - ein Bericht aus der Praxis par Precisely
Optimierung von Stammdatenprozessen - ein Bericht aus der PraxisOptimierung von Stammdatenprozessen - ein Bericht aus der Praxis
Optimierung von Stammdatenprozessen - ein Bericht aus der Praxis
Precisely9 vues
Data Integrity for Banking and Financial Services par Precisely
Data Integrity for Banking and Financial ServicesData Integrity for Banking and Financial Services
Data Integrity for Banking and Financial Services
Precisely78 vues
Automate Evolve Training: Excel Workflow for Automatic Routing par Precisely
Automate Evolve Training: Excel Workflow for Automatic RoutingAutomate Evolve Training: Excel Workflow for Automatic Routing
Automate Evolve Training: Excel Workflow for Automatic Routing
Precisely12 vues
Automatizando de procesos SAP par Precisely
Automatizando de procesos SAPAutomatizando de procesos SAP
Automatizando de procesos SAP
Precisely6 vues
Improve IT operations management with ServiceNow and Ironstream par Precisely
Improve IT operations management with ServiceNow and IronstreamImprove IT operations management with ServiceNow and Ironstream
Improve IT operations management with ServiceNow and Ironstream
Precisely19 vues
Supercharging AI with Data Enrichment par Precisely
Supercharging AI with Data EnrichmentSupercharging AI with Data Enrichment
Supercharging AI with Data Enrichment
Precisely18 vues
Protecting Your Power Systems with Cloud-based HA/DR par Precisely
Protecting Your Power Systems with Cloud-based HA/DRProtecting Your Power Systems with Cloud-based HA/DR
Protecting Your Power Systems with Cloud-based HA/DR
Precisely23 vues
Hyperautomation & AI/ML: Keys to Digital Transformation Success par Precisely
Hyperautomation & AI/ML: Keys to Digital Transformation SuccessHyperautomation & AI/ML: Keys to Digital Transformation Success
Hyperautomation & AI/ML: Keys to Digital Transformation Success
Precisely34 vues
Hyperautomation, AI & machine learning: a strategy for digital transformation... par Precisely
Hyperautomation, AI & machine learning: a strategy for digital transformation...Hyperautomation, AI & machine learning: a strategy for digital transformation...
Hyperautomation, AI & machine learning: a strategy for digital transformation...
Precisely22 vues
Obtain a 360 - degree view of Your IT Infrastructure par Precisely
Obtain a 360 - degree view of Your IT InfrastructureObtain a 360 - degree view of Your IT Infrastructure
Obtain a 360 - degree view of Your IT Infrastructure
Precisely13 vues
Streamline Your SAP-centric Business Processes with API-Driven Automation par Precisely
Streamline Your SAP-centric Business Processes with API-Driven AutomationStreamline Your SAP-centric Business Processes with API-Driven Automation
Streamline Your SAP-centric Business Processes with API-Driven Automation
Precisely43 vues
Governance as a "painkiller": A Business First Approach to Data Governance par Precisely
Governance as a "painkiller": A Business First Approach to Data GovernanceGovernance as a "painkiller": A Business First Approach to Data Governance
Governance as a "painkiller": A Business First Approach to Data Governance
Precisely13 vues
Democratized Data & Analytics for the Cloud​ par Precisely
Democratized Data & Analytics for the Cloud​Democratized Data & Analytics for the Cloud​
Democratized Data & Analytics for the Cloud​
Precisely22 vues
Looking Ahead: The Future Of Automation and Live Q&A par Precisely
Looking Ahead: The Future Of Automation and Live Q&ALooking Ahead: The Future Of Automation and Live Q&A
Looking Ahead: The Future Of Automation and Live Q&A
Precisely8 vues
Maximize Your Investments with the Precisely Strategic Services Team par Precisely
Maximize Your Investments with the Precisely Strategic Services TeamMaximize Your Investments with the Precisely Strategic Services Team
Maximize Your Investments with the Precisely Strategic Services Team
Precisely13 vues
Hyperautomation and AI/ ML: A Strategy for Digital Transformation Success par Precisely
 Hyperautomation and AI/ ML: A Strategy for Digital Transformation Success Hyperautomation and AI/ ML: A Strategy for Digital Transformation Success
Hyperautomation and AI/ ML: A Strategy for Digital Transformation Success
Precisely18 vues
The State of Automation: New Trends and Insights par Precisely
The State of Automation: New Trends and InsightsThe State of Automation: New Trends and Insights
The State of Automation: New Trends and Insights
Precisely14 vues

Dernier

Netmera Presentation.pdf par
Netmera Presentation.pdfNetmera Presentation.pdf
Netmera Presentation.pdfMustafa Kuğu
22 vues50 diapositives
Deep Tech and the Amplified Organisation: Core Concepts par
Deep Tech and the Amplified Organisation: Core ConceptsDeep Tech and the Amplified Organisation: Core Concepts
Deep Tech and the Amplified Organisation: Core ConceptsHolonomics
17 vues21 diapositives
MVP and prioritization.pdf par
MVP and prioritization.pdfMVP and prioritization.pdf
MVP and prioritization.pdfrahuldharwal141
39 vues8 diapositives
KubeConNA23 Recap.pdf par
KubeConNA23 Recap.pdfKubeConNA23 Recap.pdf
KubeConNA23 Recap.pdfMichaelOLeary82
24 vues27 diapositives
Business Analyst Series 2023 - Week 4 Session 8 par
Business Analyst Series 2023 -  Week 4 Session 8Business Analyst Series 2023 -  Week 4 Session 8
Business Analyst Series 2023 - Week 4 Session 8DianaGray10
145 vues13 diapositives
LLMs in Production: Tooling, Process, and Team Structure par
LLMs in Production: Tooling, Process, and Team StructureLLMs in Production: Tooling, Process, and Team Structure
LLMs in Production: Tooling, Process, and Team StructureAggregage
57 vues77 diapositives

Dernier(20)

Deep Tech and the Amplified Organisation: Core Concepts par Holonomics
Deep Tech and the Amplified Organisation: Core ConceptsDeep Tech and the Amplified Organisation: Core Concepts
Deep Tech and the Amplified Organisation: Core Concepts
Holonomics17 vues
Business Analyst Series 2023 - Week 4 Session 8 par DianaGray10
Business Analyst Series 2023 -  Week 4 Session 8Business Analyst Series 2023 -  Week 4 Session 8
Business Analyst Series 2023 - Week 4 Session 8
DianaGray10145 vues
LLMs in Production: Tooling, Process, and Team Structure par Aggregage
LLMs in Production: Tooling, Process, and Team StructureLLMs in Production: Tooling, Process, and Team Structure
LLMs in Production: Tooling, Process, and Team Structure
Aggregage57 vues
Initiating and Advancing Your Strategic GIS Governance Strategy par Safe Software
Initiating and Advancing Your Strategic GIS Governance StrategyInitiating and Advancing Your Strategic GIS Governance Strategy
Initiating and Advancing Your Strategic GIS Governance Strategy
Safe Software184 vues
"Package management in monorepos", Zoltan Kochan par Fwdays
"Package management in monorepos", Zoltan Kochan"Package management in monorepos", Zoltan Kochan
"Package management in monorepos", Zoltan Kochan
Fwdays34 vues
PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」 par PC Cluster Consortium
PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」
PCCC23:日本AMD株式会社 テーマ2「AMD EPYC™ プロセッサーを用いたAIソリューション」
The Power of Heat Decarbonisation Plans in the Built Environment par IES VE
The Power of Heat Decarbonisation Plans in the Built EnvironmentThe Power of Heat Decarbonisation Plans in the Built Environment
The Power of Heat Decarbonisation Plans in the Built Environment
IES VE84 vues
Future of AR - Facebook Presentation par Rob McCarty
Future of AR - Facebook PresentationFuture of AR - Facebook Presentation
Future of AR - Facebook Presentation
Rob McCarty65 vues
Digital Personal Data Protection (DPDP) Practical Approach For CISOs par Priyanka Aash
Digital Personal Data Protection (DPDP) Practical Approach For CISOsDigital Personal Data Protection (DPDP) Practical Approach For CISOs
Digital Personal Data Protection (DPDP) Practical Approach For CISOs
Priyanka Aash162 vues
What is Authentication Active Directory_.pptx par HeenaMehta35
What is Authentication Active Directory_.pptxWhat is Authentication Active Directory_.pptx
What is Authentication Active Directory_.pptx
HeenaMehta3515 vues
NTGapps NTG LowCode Platform par Mustafa Kuğu
NTGapps NTG LowCode Platform NTGapps NTG LowCode Platform
NTGapps NTG LowCode Platform
Mustafa Kuğu437 vues
"Running students' code in isolation. The hard way", Yurii Holiuk par Fwdays
"Running students' code in isolation. The hard way", Yurii Holiuk "Running students' code in isolation. The hard way", Yurii Holiuk
"Running students' code in isolation. The hard way", Yurii Holiuk
Fwdays36 vues
Transcript: Redefining the book supply chain: A glimpse into the future - Tec... par BookNet Canada
Transcript: Redefining the book supply chain: A glimpse into the future - Tec...Transcript: Redefining the book supply chain: A glimpse into the future - Tec...
Transcript: Redefining the book supply chain: A glimpse into the future - Tec...
BookNet Canada41 vues

Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX

  • 1. Continued Innovation in IBM z/System Sort Optimization with Syncsort MFX Denise Tabor | Senior Product Manager Director Alissa Margulies | Principal Sales Engineer
  • 2. Today’s Agenda • Syncsort MFX overview • Tips for Syncsort MFX optimization • Important enhancements for Syncsort MFX customers 2
  • 4. IBM® z Integrated Information Processor (zIIP): • Helps improve GP utilization • Allows customers to purchase additional processing power • Removes IBM software charges on zIIP capacity Bottom line: • Pay once, use often, without any additional cost • Eligible workloads moved to zIIP reduce license costs • HOWEVER, MOST workloads are not enabled to run on zIIP 4 IBM zIIP Engines
  • 5. Syncsort MFX • Exploits z/OS hardware and software AND zIIP • Reduces CPU utilization for sort operations • Optimizes I/O activity Delivers • Improved processor and DASD capacity • Reduced cost 5
  • 6. Syncsort MFX Reduces CPU time and I/O activity Syncsort ZPSaver zIIP-enabled sort operations Syncsort MFX PipeSort Reduces elapsed time 6
  • 7. Syncsort MFX Almost 50 years of continual development and enhancements Performance The high-performance sort/copy/join solution that delivers better performance and saves money Proven Solution Improves sort performance while optimizing overall system efficiency zIIP Offload Sort workloads can be directed to the zIIP, thereby lowering the CPU time and costs Encryption Enhanced security and compliance with regulations such as GDPR 7
  • 10. Tuning Objectives for Syncsort MFX 10 • Tuning sort requires evaluation of your current applications and defining your objectives • Consider what you are willing to trade in order to get better performance. • What is the primary outcome/goal? • Reduced CPU? • Reduced Elapsed time? • Avoiding contention with other workloads? • Reducing DASD contention? • Are you willing to: • Trade CPU for elapsed time improvements and visa versa? • Reschedule your sort jobs? • Move data sets to isolate DASD to be used by the sort? • Pass run-time parameters to the sort? • Change priorities of applications?
  • 11. Tuning Options and Recommendation 11 Control Statements • Ensure you are using the optimal control statements for an application. • Evaluate errors in control statements that can adversely affect performance Optimization Mode • Select the proper mode for the desired outcome • Mode selection may affect other areas of performance. • Performance outcome between modes may need some experimentation. Virtual Storage • Most critical resource in determining how well the sort will run. • One of the most dangerous • Using too much can lead to system storage shortages and system outages.
  • 12. Tuning Options and Recommendation 12 Rescheduling Work • Evaluate your overall concurrent workload • Is your CPU capacity, real storage and I/O resources impacted? • Determine if your sort work needs to be rescheduled to a quieter period. PARASORT • PARASORT improves the elapsed time performance for sorts whose input is a multi-volume tape data set and/or concatenated tape data set. • Uses parallel processing of the SORTIN input volumes. • Results in up to 33% reduction in elapsed time FILESIZE Estimates • Use of the FILSZ parameter provides the sort with an estimate of the amount of data to be sorted • Can significantly improve the optimization and performance of very large sorts.
  • 13. Syncsort MFX Data Manipulation Reformatting Records • INREC, OUTREC and OUTFIL OUTREC control statements. • Used when information in the input record is not required by the applications, or the data needs to be a different format. • Allows you to: • Delete or repeat segments of record • Insert new field • Convert data • Perform arithmetic operations with numeric fields and/or constants • Perform MIN/MAX functions on numeric data • Change RECFM of output data set from fixed to variable or the reverse 13 Record Selection • INPUT PHASE: Selection of records in the order in which they appear in the input data set. • OUTPUT PHASE: Selection of records seen in sorted sequence. • These selections can be specified: • Skip the first “n” number of records • Stop after processing “n” number of records • Include/omit records based on comparisons of the contents of one or more fields within the record • Include a sample of “m” records after an interval of every “n” records • Distribute the records in rotation among all of the files in an OUTFIL group • Create a file that contains only those records that were not included in any other OUTFIL
  • 14. Syncsort MFX Data Manipulation 14 Summation • Special processing done on records with equal sort keys. • Detailed records are replaced with a summary record, containing sum, average, maximum or minimum values • Detail records can be written to a separate data set Report Writer • Generate ad-hoc or scheduled reports. • Easy to use functionality with powerful record selection and formatting Join Records • Records created by joining 2 files that contain a common join key. • Join processing produces 3 types of records: • Paired records, • Unpaired from the first file • Unpaired from the second file
  • 15. Additional Syncsort MFX product options Syncsort PipeSort Can simultaneously execute up to eight differently sequenced sorts from a single pass of the input data Syncsort PROCSort High performance, transparent replacement for the SAS®-provided PROC SORT. Presentation name 15 Syncsort ZPSaver A set of enhanced technologies to offload copy, SMS compression, and sort processing to zIIP processors
  • 17. IBM Z Sort Accelerator Support (delivered) 17 • IBM’s Integrated Accelerator for Z Sort • New coprocessor designed for the z15 • Accelerates internal sorts • Precisely partnership w/IBM • Worked with HW architects • Developed new algorithms in Syncsort MFX • Results • Sort performance improvements
  • 18. IBM Z Pervasive Encryption Support (delivered) 18 • The IBM Z Pervasive Encryption enables • Powerful encryption of data in-flight and at-rest • Highly secure ways to help deal with today’s compliance and regulatory requirements • Requires additional resources • Consumes processor cycles • Forces Syncsort MFX to use less performant I/O methods (BSAM) • Solution • For Basic/Large datasets, we can continue to use our low-level I/O
  • 19. MFX Operational Visibility 19 • Customers need more insights into their sort workloads • Unexpected delays • Diagnostic information is frequently needed • Rerun a job and just to gather “debug” information • Provide more visibility • Expedite/streamline/improve troubleshooting for large/important jobs. • Examine both real-time and historical data • Port the data to an analytical tool for analysis

Notes de l'éditeur

  1. IBM® z Integrated Information Processor (zIIP): A purpose-built processor designed to operate asynchronously with the general processors in the mainframe to help improve utilization of computing capacity and control costs.  zIIPs allow customers to purchase additional processing power without affecting the total million service units (MSU) rating IBM does not impose IBM software charges on zIIP capacity, but charges apply when additional general purpose engines capacity is used Bottom line: Once zIIPs are purchased they are used without any additional cost Moving eligible workloads to zIIP reduces general purpose engine utilization and can save SW license costs The big challenge is that most workloads are not enabled to run on zIIP
  2. Offers a high-performance sort, copy, and join utility designed to exploit the advanced facilities of the z/OS operating system and IBM Z, including zIIP engines. Can significantly reduce CPU utilization for sort operations and optimize I/O activity to reduce contention to control software costs and delay CPU upgrades Frees up general purpose MIPS for handling increased data volumes and new workloads Drive cost reduction strategies by delaying CPU upgrades and reducing software charges.
  3. Sort operations are resource intensive Syncsort MFX uses less CPU time than DFSORT Syncsort MFX optimizes I/O better than DFSORT Syncsort MFX encrypts sort work for required security and compliance Faster execution means more work can execute in the same amount of time === Significantly reduces general processor operation Makes use of under-utilized zIIP engines Helps control MLC costs and delay/prevent CPU upgrades Allows sort work encryption to be completed on zIIP, reducing CPU utilization dramatically and reducing costs === Simultaneously executes up to eight differently sequenced sorts from a single pass of the input data Uses advanced parallel sorting technology Cuts total elapsed time by more than 50% compared to running separate sorts
  4. In certain circumstances, sort may be unable to get a good filesize estimate, which may be an issue when the amount of data to be sorted is over 1 gigabyte. PARASORT requires additional tape drives and will automatically manage the tape drives and optimize their usage.
  5. Notes on Record Selection: Once again it isimportant to understand the flow of the sort and where the particular feature fits into this flow. The resulting output from the sort can be dramatically different if the record selection is performed in the output phase instead of the input phase. Notes on Join Records: The disposition of records in each category can be controlled independently. The output from join processing can contain any combination of these record types. This feature is similar to the join processing found in relational data bases.
  6. JOIN RECORDS: If there are “m” number of records from the first file and “n” records in the second file that contain the same value in the join key, m*n records will be created. SUMMATION: This processing will produce a data set that has only 1 record per sort key value unless an overflow occurs If summation or averaging is requested and the resulting calculation will cause an overflow condition the summation or averaging is not done Notes on Record Selection: Once again it isimportant to understand the flow of the sort and where the particular feature fits into this flow. The resulting output from the sort can be dramatically different if the record selection is performed in the output phase instead of the input phase. Notes on Join Records: The disposition of records in each category can be controlled independently. The output from join processing can contain any combination of these record types. This feature is similar to the join processing found in relational data bases. USE CASES for REPORTS: An example of a report that provides three separate invoice status reports. These reports, titled “Unpaid Invoices,” “Partially Paid Invoices,” and “Fully Paid Invoices,” represent three output files generated with a single pass of the sort. For this example, you can print these reports or write them to disk or tape.
  7. Syncsort ZPSaver is a set of enhanced technologies for MFX to offload copy, SMS compression, and sort sort processing to zIIP processors, effectively reducing the workload on the main CPU. ZPSaver can reduce TCB CPU time up to 95% in eligible Syncsort MFX applications
  8. IBM’s Integrated Accelerator for Z Sort New coprocessor designed for the z15 that reduces CPU usage and improves elapsed time Speeds up sorting, shortens batch windows, and improves select database functions. Precisely worked closely with IBM Worked with HW architects in z/OS Poughkeepsie team to develop support for the sort accelerator. Developed new algorithms in Syncsort MFX to take advantage of the coprocessor Customers can expect to see dramatic improvements to batch sort job performance. In our lab, we are seeing CPU and elapsed time improvements of up to 35% depending on the key length, record length, file size and some other factors.
  9. The IBM Z Pervasive Encryption enables Powerful encryption of data in-flight and at-rest Highly secure ways to help deal with today’s compliance and regulatory requirements Customers love the pervasive encryption approach, no one likes the additional resource consumption The challenge to implementing data encryption is it consumes processor cycles, so it doesn’t come without penalty. For Syncsort MFX users, the input or output data set is encrypted, BSAM must be used instead of our high performant low level IO access methods, and there is an extra cost from that perspective. Once again working closely with IBM, we have identified a way to continue to use our low-level IO for encrypted data sets and improve encryption performance. We have seen great performance improvement from our benchmark testing Although mainframe customers love the pervasive encryption approach, no one likes the additional resource consumption. . Syncsort MFX alone, we have seen up to 45% CPU and 40% elapsed savings, Combing Syncsort MFX with the Syncsort ZPSaver capability we have seen up to 80% CPU and 40% elapsed time savings.
  10. Are you faced with the challenge of providing the right diagnostic information when reporting a failure? Would you like a way to expedite the troubleshooting process? Are you constrained by where you can store job information for further analysis? What types of repositories or tools do you currently use for collecting job information for analysis? Would you like more visibility into your sorting workloads? What kind of analytical tools are you utilizing in other areas of the business to do analysis? Splunk? Elastic? What type of information would you like to see when evaluating your workloads? Would you benefit from a solution that: Identifies the key pieces of information to submit to support for review? Limits or specifies the jobs or groups of jobs to analyze? Provides examples of ways to view the information, or define your own reports/dashboards? What other types of details would you like us to know?