parallel sysplex partitioning using bcpii - ibm as bcpii exploiter. 4 © 2010 ibm corporation ......

© 2010 IBM Corporation

Parallel Sysplex PartitioningUsing BCPii

Seattle, Washington Spring 2010

Riaz Ahmad – IBM GaithersburgMark Brooks – IBM Poughkeepsie

Session 2251

© 2010 IBM Corporation2

SHARE – Seattle, Washington Spring 2010

The following are trademarks of the International Business Machines Corporation in the United States and/or other countries.

The following are trademarks or registered trademarks of other companies.InfiniBand® is a registered trademark of the InfiniBand Trade Association (IBTA).Intel is a trademark of the Intel Corporation in the United States and other countries.Linux is a trademark of Linux Torvalds in the United States, other countries, or both. Java and all Java-related trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc., in the United States and other countries.Microsoft, Windows and Windows NT are registered trademarks of Microsoft Corporation.UNIX is a registered trademark of The Open Group in the United States and other countries.All other products may be trademarks or registered trademarks of their respective companies.The Open Group is a registered trademark of The Open Group in the US and other countries.SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.Notes: Performance is in Internal Throughput Rate (ITR) ratio based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput that any user will experience will vary depending upon considerations such as the amount of multiprogramming in the user's job stream, the I/O configuration, the storage configuration, and the workload processed. Therefore, no assurance can be given that an individual user will achieve throughput improvements equivalent to the performance ratios stated here. IBM hardware products are manufactured from new parts, or new and serviceable used parts. Regardless, our warranty terms apply.All customer examples cited or described in this presentation are presented as illustrations of the manner in which some customers have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics will vary depending on individual customer configurations and conditions.This publication was produced in the United States. IBM may not offer the products, services or features discussed in this document in other countries, and the information may be subject to change without notice. Consult your local IBM business contact for information on the product or services available in your area.All statements regarding IBM's future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.Information about non-IBM products is obtained from the manufacturers of those products or their published announcements. IBM has not tested those products and cannot confirm the performance, compatibility, or any other claims related to non-IBM products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products.Prices subject to change without notice. Contact your IBM representative or Business Partner for the most current pricing in your geography.This presentation and the claims outlined in it were reviewed for compliance with US law. Adaptations of these claims for use in other geographies must be reviewed by the local country counsel for compliance with local laws.

* Registered trademarks of IBM Corporation

AIX*AnyNet*CICS*DB2*developerWorks*DFSMSdfpDFSMSdssDFSMShsmDFSMSrmmDFSORT*DRDA*DS8000ESCON*FICON*

FlashCopy*GDPS*HiperSocketsHiperspaceHyperSwapIBM*IBM eServerIBM logo*IMSInfoprint*IP PrintWayiSeriesLanguage Environment*

MVSNotes*OMEGAMON*Open Class*Parallel Sysplex*PR/SMProcessor Resource/ Systems ManagerpSeries*RACF*RedbookRMFSystem iSystem Storage

System zSystem z9System z10SystemPac*Tivoli*VTAM*WebSphere*z/Architecture*z/OS*z/VM*zSeries*

Trademarks



Agenda

What is BCPii ?

Required setup for BCPii

SFM as BCPii exploiter



What is BCPii ?

Base Control Program internal interface (BCPii)

– Allows authorized z/OS applications to have HMC-like control over systems in the process control HMC network

– Complete communication isolation of existing networks (internet/intranet) from the process control (HMC) network

• Communication to the System z support element is completely within z/OS

– Set of authorized APIs

– A new z/OS Address Space - HWIBCPII• A communications transport between z/OS application and the local support

element• Also between other support elements connected to other CPCs routed by the HMC



What is BCPii …

Authorized z/OS application

•Monitor status or capacity changes

•Obtain configuration data related to CPC or image

•Re-IPL or reset an image SE

SESE

HMC

CPC1 CPC2

CPC3

Process Control (HMC)

Network



Who would use BCPii?

Current IBM BCPii exploiters

– Capacity Provisioning Manager (CPM)• The capability of adding or removing temporary capacity based on goals defined in

the WLM policy• Available in z/OS V1R10 via APAR OA24945

– Sysplex Failure Manager (SFM)• Used to expedite partitioning when a system fails• Used to determine whether a system has been reset• Available in z/OS V1R11

ISV applications

In-house applications– Installations that have desire to manipulate System z resources

• In a more programmatic and automated way



How to obtain BCPii

z/OS 1.11 – Shipped in the base z/OS– Additional functions are available

z/OS 1.10– Separate deliverable– APAR OA25426 (PTF UA47493)

After installation of z/OS 1.11 or OA25426 on z/OS 1.10– BCPii will attempt to start at IPL– If the proper configuration has not been setup , it will terminate



IPL Messages Prior to BCPii Setup in z/OS 1.11IRR813I NO PROFILE WAS FOUND IN THE STARTED CLASS FOR 497

HWIBCPII WITH JOBNAME HWIBCPII. RACF WILL USE ICHRIN03.

IEF196I 2 //HWIBCPII EXEC IEESYSAS,PROG=HWIAMIN2

IEF196I STMT NO. MESSAGE

IEF196I XX* THE IEESYSAS PROCEDURE IS SPECIFIED IN THE

IEF196I 00150000

IEF196I XX* PARAMETER LIST TO IEEMB881 BY MVS COMPONENTS

IEF196I 00200000

IEF196I 2 IEFC001I PROCEDURE IEESYSAS WAS EXPANDED USING SYSTEM

IEE252I MEMBER CTIHWI00 FOUND IN SYS1.PARMLIB

HWI007I BCPII IS ATTEMPTING COMMUNICATION WITH THE LOCAL CENTRAL 965

PROCESSOR COMPLEX (CPC).

HWI014I THE SNMP COMMUNITY NAME FOR BCPII IN THE SECURITY 049

PRODUCT FOR THE LOCAL CPC IS EITHER NOT DEFINED OR IS INCORRECT.

IEA989I SLIP TRAP ID=X13E MATCHED. JOBNAME=HWIBCPII, ASID=001A.

HWIBCPII HWIBCPII IEFPROC 00:00:00 2,117 00

IEF404I IEESYSAS - ENDED - TIME=13.31.03

IEF196I IEF142I IEESYSAS HWIBCPII - STEP WAS EXECUTED - COND CODE 0000

IEF196I IEF376I JOB/HWIBCPII/STOP 2009239.1331 CPU 0MIN 00.02SEC

IEF196I SRB 0MIN 00.00SEC

IEF352I ADDRESS SPACE UNAVAILABLE

IEF196I IEF352I ADDRESS SPACE UNAVAILABLE

IEA989I SLIP TRAP ID=X33E MATCHED. JOBNAME=*UNAVAIL, ASID=001A.

HWI006I BCPII ADDRESS SPACE HAS ENDED.

IXC104I SYSTEM STATUS DETECTION PARTITIONING PROTOCOL ELIGIBILITY: 399

SYSTEM CANNOT TARGET OTHER SYSTEMS.

REASON: BCPII SERVICES NOT AVAILABLE

SYSTEM IS NOT ELIGIBLE TO BE TARGETED BY OTHER SYSTEMS.

REASON: BCPII SERVICES NOT AVAILABLE

IXC107E SYSTEM STATUS DETECTION PARTITIONING PROTOCOL CONFIGURATION IS NOT COMPLETE



How to use BCPii

Functions performed using BCPii APIs:

– Obtain the System z topology of the current interconnected CPCs, Images (LPARs) and Capacity Records

– Query various CPC, image (LPAR) and capacity record information

– Issue commands against both the CPC and image to perform hardware and software-related functions

– Listen for various hardware and software events which may take place on various CPC and images

Environment:

– Services available in any address space • Program-authorized, and• SAF-authorized

– C and Assembler programming languages

– Unix System Services callers can receive event notifications thru z/OS UNIX-only services utilizing Common Event Adapter (CEA)



How to use BCPii (the BCPii APIs)

List of available APIs

– HWILIST (BCPii List)

– HWICONN (BCPii Connect)

– HWIDISC (BCPii Disconnect)

– HWIQUERY (BCPii Query)

– HWISET (BCPii Set) – introduced in V1R11

– HWICMD (BCPii Command)

– HWIEVENT (BCPii Event (for non-z/OS UNIX callers))

– HwiBeginEventDelivery, HwiEndEventDelivery, HwiManageEvents, HwiGetEvent (for z/OS UNIX callers)



Information you can query using the BCPii APIsCPC information

– General information• Name, serial, machine type, id, networking info

– Status information• Operating status and other status values

– Capacity information• Various CBU info, Capacity on Demand info, Processor configuration, including IFA, IFL, ICF, IIP

Image information

– General information• Name, OS info

– Status Information• Operating status and other status values

– Capacity information• Defined capacity, Processor LPAR weights

Capacity record information

– General information• Name, Activation and expiration dates, activation days

– Status information• Record status

– Capacity information• The entire Capacity record



Information you can set using the BCPii APIsCPC information

– Acceptable status values– Next Reset activation profile name– Processor Running Time

Image information– Various processor weights

Activation Profile Information (available in V1R11, 1st Quarter 2010 via APAR OA29638)

Any activation profile value

CPC commands• Activate, Deactivate an entire CPC• CBU request Activate or Undo• On/Off Capacity on Demand request

Activate or Undo

Image commands– SysReset, SysReset with IPL Token (V1R11)– Load– Start, Stop all CPs– Add or remove temporary capacity– Issue operating system command




Examples of events that can be listened to:

– Command completions

– Status changes

– Capacity changes

– Disabled waits

– BCPii status changes




Application contains calls like this:

– HWILIST (ListCPCs)

– For each CPC name returned above:• HWICONN (CPC name (input), CPCConnectToken(output))• HWIQUERY (CPCConnectToken (input), QueryParms (HWI_CBUTESTAR))• HWILIST (CPCConnectToken(input), ListImages)• For each image returned above:

HWICONN (CPCConnectToken(input), Image name (input), ImageConnectToken(output))

HWIQUERY (ImageConnectToken(input), QueryParms(HWI_OSNAME,HWI_OSTYPE,HWI_OSLEVEL,HWI_SYSPLEX,HWI_DEFCAP)

HWIDISC(ImageConnectToken)• HWIDISC (CPCConnectToken)



The new BCPii Address Space

The BCPii address space:

– Starts automatically during IPL

– Requires some initial setup before becoming active

– Can be stopped if necessary

– Can be restarted via the HWISTART

User’s Address Space BCPii Address Space

BCPii

API

Processing

BCPii

Transport

Processing

z/Series

APIs

To/From Support ElementPC

BCPii APIs



The new BCPii address space

The bridge between a z/OS application and the support element

– Manages all application connections

– Builds and receives all internal communication requests to the SE

– Provides an infrastructure for storage required by callers and for the transport communicating with the SE

– Provides diagnostic capabilities to help with BCPii problem determination

– Provides security authentication of requests



BCPii Configuration

BCPii setup steps:

– Configure the local SE to support BCPii

– Configure the BCPii address space

– Authorize an application to use BCPii

– Set up the event notification mechanism for z/OS UNIX callers (if required)



BCPii Configuration

Configure the local SE to support BCPii

– Make sure microcode is at the correct level

– Enable BCPii communications on the SE• Turn on “Cross Partition Authority”

– Define the BCPii community name on the SE• Turn on “Enable SNMP APIs”• Add a BCPii community name for the BCPii port

Repeat these same steps above for any additional CPCs communication is desired.



Access Hardware Management Console



Cross Partition Authority



Access z Support Element



HMC Single Object Operation



Support Element Settings



Customize API Settings



Community Name



Community Name Information



CEC’s SNA Name



BCPii Configuration

Authorize the application to use BCPii

– Give program authority to application issuing BCPii APIs

– Give general SAF authority

• Give application at least READ access to HWI.APPLNAME.HWISERV profile in the FACILITY resource class

• Make sure the FACILITY class is in RACLIST

Note: For XCF, simply have to ensure that the XCFAS started task is defined in RACF with the TRUSTED attribute



BCPii Configuration …

Authorize the application to use BCPii

– Give SAF authority to a specific resource

• Give at least READ access to the SAF profile name below in the FACILITY resource class (LIST, CONN, DISC, EVENT, QUERY)

CPC Profile: HWI.TARGET.netid.nauImage Profile: HWI.TARGET.netid.nau.imagenameCaprec Profile: HWI.CAPREC.netid.nau.caprec

• Give ALTER access for CMD

Netid.nau represents the 3 to 17 character SNA name of a CEC




Authorize the application to use BCPii – Define community name in SAF for each CPC– Associate the community name already defined on the support element for

BCPii with the SAF CPC profile already definedUse the APPLDATA field to save this

– Can be done at the time of the CPC definition– BCPii propagates this community name to the support element when a

connection is established for validation by the SE.

The local CPC’s community name MUST be defined to SAF or the BCPii address space will fail to initialize




Configure the BCPii address space

– Make sure the high-level-qualifier.SCEERUN2 and high-level-qualifier.SCEERUN data sets are in the link list concatenation

– Optionally, configure the BCPii CTRACE member (CTIHWI00) to set up default CTRACE options for the address space

– Optionally, configure SCHEDxx to modify PPT for BCPii address spaces• HWIAMIN1 and HWAMIN2• IBM Supplied defaults:

NOSWAP - Non SwappableSYST - System TaskKEY(4) - PSW Key for the programLPREF - First level preferred storage

– New z/OS operator commands• START HWISTART• STOP HWIBCPII



Starting HWIBCPII Address Space

HWISTART STARTING HWISTART 00:00:00 2 00

IEF404I HWISTART - ENDED - TIME=11.20.18

$HASP395 HWISTART ENDED

IEA989I SLIP TRAP ID=X33E MATCHED. JOBNAME=*UNAVAIL, ASID=0049.

IRR813I NO PROFILE WAS FOUND IN THE STARTED CLASS FOR 846

HWIBCPII WITH JOBNAME HWIBCPII. RACF WILL USE ICHRIN03.

IEF196I 1 //IEESYSAS JOB MSGLEVEL=1

IEF196I 2 //HWIBCPII EXEC IEESYSAS,PROG=HWIAMIN2

IEF196I STMT NO. MESSAGE

:

:

IEF196I 00200000

IEF196I XX* STARTING FULL FUNCTION SYSTEM ADDRESS SPACES.

IEF196I 00250000

IEF196I IEFC653I SUBSTITUTION JCL - PGM=HWIAMIN2

IEF403I IEESYSAS - STARTED - TIME=11.20.18

IEE252I MEMBER CTIHWI00 FOUND IN SYS1.PARMLIB

IEF196I IEF285I SYS1.PARMLIB KEPT

IEF196I IEF285I VOL SER NOS= OSPPGE.

IEF196I IEF285I SYS1.PARMLIB.IBM KEPT

IEF196I IEF285I VOL SER NOS= ZOBRS1.

HWI016I THE BCPII COMMUNICATION RECOVERY ENVIRONMENT IS 869

NOW ESTABLISHED.

HWI007I BCPII IS ATTEMPTING COMMUNICATION WITH THE LOCAL CENTRAL 870

PROCESSOR COMPLEX (CPC).

HWI001I BCPII IS ACTIVE.



Starting HWIBCPII Address Space on z9IEF196I IEF285I VOL SER NOS= ZOBRS1.

HWI001I BCPII IS ACTIVE.

IXC103I SYSTEM IDENTIFICATION INFORMATION 754

CONNECTION STATUS: CONNECTED

SYSTEM NAME: SYSB

SYSTEM NUMBER: 01000563

IMAGE NAME: SOSP2

NODE DESCRIPTOR: 002094.IBM.02.00000004B10E

PARTITION NUMBER: 02

CPC ID: 00

NETWORK ADDRESS: IBM390PS.SSYS

IPL TOKEN: NOT AVAILABLE

IXC104I SYSTEM STATUS DETECTION PARTITIONING PROTOCOL ELIGIBILITY: 755

SYSTEM CAN TARGET OTHER SYSTEMS.

SYSTEM IS NOT ELIGIBLE TO BE TARGETED BY OTHER SYSTEMS.

REASON: NOT SUPPORTED BY HARDWARE

IXC111I LOGICAL PARTITION REMOTE CONNECTION INFORMATION 756


SYSTEM NAME: SYSC


IMAGE NAME: TOSP1

NETWORK ADDRESS: IBM390PS.TSYS

IPL TOKEN: C4D59D59 C0C7F8B7

DIAG INFO: N/A

IXC111I LOGICAL PARTITION REMOTE CONNECTION INFORMATION 757


SYSTEM NAME: SYSD


IMAGE NAME: TOSP2

NETWORK ADDRESS: IBM390PS.TSYS

IPL TOKEN: C4D59E6A A2989016

DIAG INFO: N/A



BCPii Diagnostics

API Return Codes

CTRACE– BCPii cuts CTRACE records using SYSBCPII CTRACE comp– Default CTRACE CTIHWI00 PARMLIB member shipped– Two CTRACE options:

• Min• All

– Dump is taken whenever CTRACE is turned off

Symptom Records– Limited first failure data capture for select problems

Support Element Tracing

TRACEOPTS ON OPTIONS('ALL')

© 2010 IBM Corporation

System Status Detection Protocoland

BCPii



Overview

In a z/OS Parallel Sysplex, unresponsive systems must be partitioned from the sysplex in a timely manner

– Both to avoid corruption of shared data, -and-– To avoid “sympathy sickness” in the surviving systems of the sysplex

The removal of unresponsive systems from the sysplex is the responsibility of XCF component of z/OS

– The Sysplex Failure Management (SFM) sub-component of XCF provides a policy-based mechanism to enable XCF to automatically deal with unresponsive systems



Overview - Background

A weakness of current sysplex partitioning support is that today, XCF does not really know the state of the system that is in an unknown, “status-update missing” state

All that is known is that the system looks unresponsive because it is not performing system status updates

– Thus XCF monitors the system for an interval of time (the failure detection interval) to allow the system some time in which to resume normal operation

– The penalty for waiting to see if the system will resume normal operation is that the sysplex may suffer some workload processing interruption for as long as the system’s failure detection interval



Overview - Problem

There are limitations which constrain XCF’s ability to remove an ailing system from the sysplex in an expeditious manner

– When operator intervention is required, the delay can extend to many minutes

XCF cannot unambiguously determine that a system is truly dead

– XCF must infer that it is appropriate to partition a system based on the amount of time elapsed since that system has updated its heartbeat in the sysplex couple data set (CDS) and/or sent XCF signals

XCF relies on the coupling facility (CF) ISOLATE command to fence a target system from the channel subsystem

– Isolation is needed to ensure data integrity when partitioning a system– Isolation may fail for many reasons, and when it does, XCF has no choice but to

prompt the operator to perform a manual system reset



Solution

The BCPii support implemented in z/OS V1R11 provides mechanism by which XCF can, in most cases, eliminate these constraints

– BCPii’s query capabilities allows XCF to determine whether a system is demised, without waiting for installation specified intervals to elapse

– BCPii’s remote command capability provides XCF an alternative to isolate, allowing XCF to reset an image when fencing fails

A system is demised when:• Enters a non-restartable wait-state• Reset• Reloaded• Deactivated• Check stopped



Overview - Solution

The intent of this support is to allow XCF to obtain certain knowledge of the state of another image in the sysplex, and to thereby eliminate these issues with the sysplex partitioning process

These enhancements to removing a system from the sysplex are called the System Status Detection (SSD) Partitioning Protocol

– When SSD is enabled, the monitoring system detects changes in target system status and may thereby be able to bypass certain delays that occur in the status monitoring and partitioning processes

– This support also enables a system to initiate actions, such as reset-normal, against the peer system to ensure that it is no longer actively processing



Overview

When the System Status Detection Partitioning Protocol is enabled XCF will use the BCPii APIs, to expedite and enhance the sysplexpartitioning process in the following areas:

– Bypass the Failure Detection Interval (FDI)

– Bypass system isolation (fencing)

– Avoid IXC402D manual intervention

– Bypass the cleanup interval

– Avoid IXC102A manual intervention

– Validate a “DOWN” response• Messages IXC102A and IXC402D



Overview

Bypass the Failure Detection Interval (FDI)

– FDI is used to indicate how long a system may appear unresponsive before other systems in the sysplex should start to take disruptive actions against that system

– XCF will use the BCPii remote query capabilities to discover when another system has become demised

– XCF will immediately initiate sysplex partitioning without waiting for the FDI to elapse if the system can be determined to be demised

• This will typically speed up partitioning by a few minutes



Overview

Bypass system isolation (fencing)

– A partitioned system must be isolated from shared resources to ensure data integrity

– XCF uses the BCPii remote query capabilities to discover when another system is isolated from shared resources • For example: reset, check-stopped, powered off, deactivated

– If the system is determined to be incapable of accessing shared resources, XCF can bypass system isolation processing

– In cases where traditional system isolation fails, XCF can use BCPii services to reset the subject system and thereby isolate the system from shared resources



Overview

Avoid IXC402D manual intervention

– IXC402D is normally issued when a system has neither updated its “heartbeat” in the sysplex CDS nor sent XCF signals within the FDI, and the SFM policy specifies an indeterminate status action of PROMPT

– When XCF can determine through BCPii APIs that the unresponsive system is truly demised, XCF may bypass the FDI and immediately proceed with partitioning without issuing WTOR

– NOTE: IXC402D is not eliminated !• System could be unresponsive but not demised (e.g. spin loop)• Or BCPii services might fail



Overview

Avoid IXC102A manual intervention

– IXC102A may be issued when SFM policy specifies PROMPT

• XCF will not issue the WTOR if it can use the BCPii remote query capability to determine that the system is demised

– IXC102A may be issued if SFM cannot successfully isolate the system image that is being partitioned

• XCF will not issue the WTOR if it can use the BCPii remote reset capability to perform a system reset against the system

– NOTE: IXC102A is not ELIMINATED !• The BCPii services might fail, and XCF will fall back to the WTOR



Overview

Bypass the cleanup interval

– The CLEANUP interval determines how long XCF is to wait for exploiters to perform cleanup processing before a system shuts down

– When varying a system out of the sysplex in response to operator command, XCF will use the BCPii query capability to determine when the target system becomes demised• At that point, it is no longer necessary to wait for the cleanup interval to elapse, as no

further cleanup is taking place on the target

– XCF can immediately continue with partitioning processing



Overview

Validate a “DOWN” response

– Messages IXC102A and IXC402D invite the operator to reset an unresponsive system and to reply “DOWN” when the reset is complete

• It is critical that the operator actually perform the reset before responding to the WTOR, because as soon as the response is received, XCF proceeds with partitioning, freeing other systems to take over resources previously owned by the outgoing system

• If the target system has not in fact been reset, and is still manipulating those resources, data integrity problems can result

– XCF will use BCPii query services to verify that the target system has in fact been reset before proceeding with partitioning, and will re-prompt the operator if it has not

• Note: if BCPii verification is not possible, XCF continues to trust the operator

If operator can reply: likely SSD is not functional or operator is wrong !



Software and Hardware Dependencies

Hardware Dependencies– z10 EC GA2, z10 BC GA1 processors or later– With MCL/EC for BCPii HWICMD (F85906.116)

New Sysplex Couple Data Set format

z/OS V1R11

BCPii properly configured and services active

By default, the SSD function will be exploited as soon as the necessary environmental requirements are met

Note: SSD Partitioning Protocol does not apply to VM-Guest Sysplexes



New Sysplex Couple Data Sets

SYSSTATDETECT protocol requires a new sysplex CDS format

– Formatting a sysplex couple data set to support the SSD protocol will increment the version number associated with the couple data set and expand the System Record

//AHMAD1 JOB (67YC0000),CLASS=A,REGION=0M, // MSGCLASS=O,NOTIFY=AHMAD //STEP1 EXEC PGM=IXCL1DSU //STEPLIB DD DSN=SYS1.MIGLIB,DISP=SHR //SYSPRINT DD SYSOUT=* //SYSIN DD *

DEFINEDS SYSPLEX(WSCZPLEX) DSN(SYS1.XCFCDS.PRI) VOLSER(XCFCD1)MAXSYSTEM(8) CATALOG

DATA TYPE(SYSPLEX) ITEM NAME(GROUP) NUMBER(80) ITEM NAME(MEMBER) NUMBER(350) ITEM NAME(GRS) NUMBER(1) ITEM NAME(CLUSTER) NUMBER(1) ITEM NAME(SSTATDET) NUMBER(1)



New Sysplex Couple Data Sets …

Toleration APAR OA26037– Allows down level systems to tolerate the new sysplex CDS format– Available for z/OS V1R9 and z/OS V1R10– Does NOT enable the down level systems to use SSD protocols

Planning Caveats– The new CDS cannot be placed into service if there are down level systems without

the toleration support in the sysplex– Down level systems without the toleration support will not be able to join a sysplex

that is using the new CDS– Once the new CDS is placed into service, a sysplex wide IPL would be needed to

fall back to a sysplex CDS with an older format

Place the new Sysplex CDS (two) into service in the usual way– SETXCF COUPLE,ACOUPLE command to define CDS as new alternate– SETXCF COUPLE,PSWITCH command to make it primary– SETXCF COUPLE,ACOUPLE to avoid the single point of failure– Remember to update COUPLExx members accordingly



Disabling/Enabling SSD Protocols

New SYSSTATDETECT switch for XCF FUNCTIONS parameter

– Use this parameter to enable the new protocol, or to disable it and revert to the existing partitioning protocol

– Controls whether this system can employ the system status detection partitioning protocol when removing other systems from the sysplex

– Controls whether other systems can employ the system status detection partitioning protocol when removing this system from the sysplex

SETXCF FUNCTIONS,{ENABLE|DISABLE}=SYSSTATDETECT

COUPLExx:– FUNCTIONS ENABLE (SYSSTATDETECT)

DISABLE (SYSSTATDETECT)

By default, the SSD protocols are enabled and will be automatically employed when the necessary environmental requirements are met



D XCF,COUPLE - IXC357IDUPLEX LOCK SYNC/ASYNC THRESHOLD: 45

CF REQUEST TIME ORDERING FUNCTION: INSTALLED

SYSTEM STATUS DETECTION PARTITIONING PROTOCOL ELIGIBILITY: SYSTEM CAN TARGET OTHER SYSTEMS. SYSTEM IS ELIGIBLE TO BE TARGETED BY OTHER SYSTEMS.

SYSTEM NODE DESCRIPTOR: 002094.IBM.02.00000004B10E PARTITION: 02 CPCID: 00

SYSTEM IDENTIFIER: B10E2094 0200053C

NETWORK ADDRESS: USIBMSC.SCZP201

PARTITION IMAGE NAME: A01

IPL TOKEN: C4B6E90D BCA7040E

COUPLEXX PARMLIB MEMBER USED AT IPL: COUPLEZB

OPTIONAL FUNCTION STATUS: FUNCTION NAME STATUS DEFAULT DUPLEXCF16 DISABLED DISABLEDSYSSTATDETECT ENABLED ENABLEDUSERINTERVAL DISABLED DISABLED

SYSPLEX COUPLE DATA SETS PRIMARY DSN: SYS1.XCFCDS.PRI

VOLSER: XCFCD1 DEVN: 6808 FORMAT TOD MAXSYSTEM MAXGROUP(PEAK) MAXMEMBER(PEAK)08/21/2009 15:11:41 8 80 (51) 351 (32)ADDITIONAL INFORMATION...SYSTEM STATUS DETECTION PROTOCOL IS SUPPORTED



Migration

SysStatDetect Protocol:

– The SSD Partitioning Protocol support shipped in base z/OS V1R11

• There are no plans to roll back the new function provided by this support• As soon as two or more systems are running z/OS 1.11, those systems can begin exploiting the

SSD partitioning protocol, assuming hardware environmental requirements are met• However, the new function will not be completely effective until all systems in the sysplex are at

or above z/OS 1.11

– Toleration APAR OA26037 enables systems below z/OS V1R11 to operate with a sysplex CDS formatted to support System Status Detection Partitioning Protocol (SSTATDET)

• Down level (below V1R11) systems will not be able to participate in the SSD Protocols



Migration & Coexistence Considerations

New Default of ISOLATETIME(0)

– The SFM policy allows you to specify how a system is to respond to a Status Update Missing condition

– The “system default” is used when no action is specified for the system or when no SFM policy is active

• Before z/OS 1.11, the system default action is PROMPT, which prompts the operator when a system enters a status update missing condition

• As of z/OS 1.11, the system default action has changed to ISOLATETIME(0), which allows other systems to take immediate action to isolate the failed system

– The release level of monitoring system determines which system default action is used– D XCF,COUPLE shows what the system expects, but the monitor may do something else

If manual intervention is still required when a system goes status update missing you will need to set up a SFM Policy to specify PROMPT

(But ISOLATETIME(0) is considered best practice)



New and Updated Messages

Updated MessagesIXC101IIXC105IIXC335IIXC357IIXC358IIXC208IIXC409DIXC422IIXC602I IXC603I IXC611IIXC373I

New MessagesIXC103I - HardcopyIXC104IIXC106I IXC107EIXC108I - HardcopyIXC109I - HardcopyIXC111I - HardcopyIXC112I - HardcopyIXC113I - HardcopyIXC114I - Hardcopy



Publications

z/OS MVS Programming: Callable Services for High-Level Languages – Primary BCPii documentation including:

• Installation instructions• BCPii API documentation

z/OS MVS Programming: Authorized Assembler Services Reference, Volume 2 (EDT-IXG) – BCPii’s ENF68 documentation

z/OS MVS System Commands– START HWISTART and STOP HWIBCPII commands

z/OS MVS Diagnosis: Tools and Service Aids – BCPii’s CTRACE documentation

z/OS MVS Initialization and Tuning Reference – Miscellaneous documentation

z/OS MVS System Codes – BCPii abend ‘042’x documentation

z/OS MVS Setting Up a Sysplex



Thank You !

parallel sysplex partitioning using bcpii - ibm as bcpii exploiter. 4 © 2010 ibm corporation ......

Documents