ca mim for z/os - 12.5 scenarios · ca mim for z/os - 12.5 13-jul-2016 3/86 ... how to change a...

13-Jul-2016Date:

CA MIM for z/OS - 12.5Scenarios

CA MIM for z/OS - 12.5

13-Jul-2016 3/86

This Documentation, which includes embedded help systems and electronically distributed materials, (hereinafter referred to as the “Documentation”) is for your informational purposes only and is subject to change or withdrawal by CA at any time. This Documentation is proprietary information of CA and may not be copied, transferred, reproduced, disclosed, modified or duplicated, in whole or in part, without the prior written consent of CA.

If you are a licensed user of the software product(s) addressed in the Documentation, you may print or otherwise make available a reasonable number of copies of the Documentation for internal use by you and your employees in connection with that software, provided that all CA copyright notices and legends are affixed to each reproduced copy.

The right to print or otherwise make available copies of the Documentation is limited to the period during which the applicable license for such software remains in full force and effect. Should the license terminate for any reason, it is your responsibility to certify in writing to CA that all copies and partial copies of the Documentation have been returned to CA or destroyed.

TO THE EXTENT PERMITTED BY APPLICABLE LAW, CA PROVIDES THIS DOCUMENTATION “AS IS” WITHOUT WARRANTY OF ANY KIND, INCLUDING WITHOUT LIMITATION, ANY IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NONINFRINGEMENT. IN NO EVENT WILL CA BE LIABLE TO YOU OR ANY THIRD PARTY FOR ANY LOSS OR DAMAGE, DIRECT OR INDIRECT, FROM THE USE OF THIS DOCUMENTATION, INCLUDING WITHOUT LIMITATION, LOST PROFITS, LOST INVESTMENT, BUSINESS INTERRUPTION, GOODWILL, OR LOST DATA, EVEN IF CA IS EXPRESSLY ADVISED IN ADVANCE OF THE POSSIBILITY OF SUCH LOSS OR DAMAGE.

The use of any software product referenced in the Documentation is governed by the applicable license agreement and such license agreement is not modified in any way by the terms of this notice.

The manufacturer of this Documentation is CA.

Provided with “Restricted Rights.” Use, duplication or disclosure by the United States Government is subject to the restrictions set forth in FAR Sections 12.212, 52.227-14, and 52.227-19(c)(1) - (2) and DFARS Section 252.227-7014(b)(3), as applicable, or their successors.

Copyright © 2016 CA. All rights reserved. All trademarks, trade names, service marks, and logos referenced herein belong to their respective companies.

Scenarios 4

Table of Contents

How to Dynamically Change CA MIM Communication Methods ................ 8Activate XCF Communication when COMMUNICATION=DASDONLY ...................................................... 9

Revert to the Original Communication Method ................................................................................... 11

Activate List Structures when COMMUNICATION=XCF or CTCONLY .................................................... 12

Revert to the Original List Structures .................................................................................................. 14

Activate XCF Communication on a Subset of the Systems ...................................................................... 14

Revert to CTC Communication ........................................................................................................... 18

Add a System to the MIMplex when COMMUNICATION=CTCDASD ...................................................... 19

Revert to the Original MIMplex ............................................................................................................ 23

Add a System when COMMUNICATION=DASDONLY ............................................................................ 23

Add a System when COMMUNICATION=XCF ......................................................................................... 24

How to Offload CA MIM Work to zIIP Engines .......................................... 27Prerequisites ............................................................................................................................................. 27

Before You Begin ................................................................................................................................ 27

About zIIP Technology .............................................................................................................................. 28

How You Offload CA MIM Work to zIIP Engines ....................................................................................... 29

Verify CCS CAIRIM Component Release and Service Levels .................................................................. 30

Verify MIMplex Systems Have Online zIIP Engines .................................................................................. 30

Verify CA MIM is Active ............................................................................................................................. 32

Assess CA MIM Workload to be Offloaded to zIIP Engines ...................................................................... 33

Activate the zIIP Enablement Feature ....................................................................................................... 36

Verify zIIP Enablement Feature is Active .................................................................................................. 37

Assess CA MIM Workload being Offloaded to zIIP Engines ..................................................................... 38

Results of zIIP Enablement Feature Offload ............................................................................................. 42

Summary ................................................................................................................................................... 44

CA MIM HYPERSTAR Feature - A Case Study ........................................ 45How To Read This Case Study ................................................................................................................. 45

About The HYPERSTAR Feature ............................................................................................................. 47

How HYPERSTAR Works ......................................................................................................................... 47

Without HYPERSTAR ......................................................................................................................... 47

Scenarios 5

With HYPERSTAR .............................................................................................................................. 48

Activating HYPERSTAR ............................................................................................................................ 48

Test Results .............................................................................................................................................. 49

General Summary ............................................................................................................................... 49

Statistical Details ................................................................................................................................. 50

Test 1 ......................................................................................................................................... 50

Test 2 ......................................................................................................................................... 52

Test 3 ......................................................................................................................................... 55

Average Cycle Times On The Four Client Systems .................................................................................. 57

How to Set Up Restart Manager ............................................................... 58Understanding How Restart Manager Works ............................................................................................ 59

Review the Restart Manager Prerequisites ............................................................................................... 60

Review the Restart Manager Set Option and Commands ........................................................................ 61

Review the LOADLIB Command ............................................................................................................... 61

DISPLAY LOADLIB ............................................................................................................................. 61

LOADLIB ............................................................................................................................................. 62

How to Use Restart Manager to Apply Maintenance Dynamically ............ 64Review the LOADLIB Command ............................................................................................................... 65

DISPLAY LOADLIB ............................................................................................................................. 65

LOADLIB ............................................................................................................................................. 66

Populate the Alternate Library with the Maintenance PTF ........................................................................ 67

Commit the Alternate Library as the Primary Library ................................................................................ 68

How to Activate MIMQUERY in the MIMplex Dynamically ........................ 69Review the MIMQUERY Prerequisites ...................................................................................................... 71

CA Common Services and CCS for z/OS Component Requirements ................................................ 71

CA MIM Setup Requirements ............................................................................................................. 72

Required PTFs .................................................................................................................................... 72

Review the Environment Setup ................................................................................................................. 73

Verify that You Installed the Maintenance and Activate MIMQUERY Dynamically ................................... 75

Use MIMQUERY to Request Resource Information ................................................................................. 76

How to Use the Restart Manager and Dynamic Refresh Enhancements .....

78

Scenarios 6

How to Change a System Name Without a Control File Format ............... 79

How to Remove a System from the MIMPLEX when

COMMUNICATION=DASDONLY ............................................................. 83

How to Remove a System from the MIMPLEX when

COMMUNICATION=CTCONLY ................................................................ 85


13-Jul-2016 7/86

ScenariosThis section covers the following topics:

How to Dynamically Change CA MIM Communication Methods (see page 8)How to Offload CA MIM Work to zIIP Engines (see page 27)CA MIM HYPERSTAR Feature - A Case Study (see page 45)How to Set Up Restart Manager (see page 58)How to Apply Maintenance Updates Without Restarting Your System Dynamically (see page 64)How to Activate MIMQUERY in the MIMplex Without Restarting Your Systems (see page 69)How to Use the Restart Manager and Dynamic Refresh Enhancements (see page 78)How to Change a System Name Without a Control File Format (see page 79)How to Remove a System from the MIMPLEX when COMMUNICATION=DASDONLY (see page 83)How to Remove a System from the MIMPLEX when COMMUNICATION=CTCONLY (see page 85)


13-Jul-2016 8/86

1.

2.

3.

4.

5.

How to Dynamically Change CA MIM Communication Methods

Global ENQ response time is dependent on the CA MIM communication method. As a z/OS systems programmer, using an alternate method can improve response time. By supporting dynamic configuration changes, you can more easily try different configurations and measure the results.

MIM--Dynamically Change Communication Methods

The following is a list of supported dynamic procedures.

Activate XCF Communication when COMMUNICATION=DASDONLY (see page 9)

Revert to the Original Communication Method (see page 11)

Activate List Structures when COMMUNICATION=XCF or CTCONLY (see page 12)

Revert to the Original List Structures (see page 14)

Activate XCF Communication on a Subset of Systems (see page 14)

Revert to CTC Communication (see page 18)

Add a System to the MIMplex when COMMUNICATION=CTCDASD (see page 19)

Revert to the Original MIMplex (see page 23)

Add a System when COMMUNICATION=DASDONLY (see page 23)


13-Jul-2016 9/86

6.

1.

Add a System when COMMUNICATION=XCF (see page 24)

Activate XCF Communication when COMMUNICATION=DASDONLY

This procedure dynamically activates the XCF communication and works with either DASD control files or XES list structures. The MIMplex is three systems using XES list structures with the MIM system names of GK03, GK13, and GK62.

Note: Be sure that your COMPATLEVEL is set to 12.0 or higher.

The following steps demonstrate how to dynamically activate the XCF communication when COMMUNICATION=DASDONLY.

Follow these steps:

Set up your XCF communication:

Confirm that all systems in the MIMplex are also within the same sysplex.

Set MIMINIT VCFBUFFERSIZE to 32768 or larger.

Review the XCF configuration and options in the SYS1.PARMLIB(COUPLExx) member.

You do not need to define a special transport class to CA MIM. However, we recommend at least one CLASSLEN of 32768 or larger to avoid possible performance problems. CLASSLEN is an XCF parameter and is coded in the SYS1.PARMLIB(COUPLExx) member.

If some of the LPARS are geographically remote, then IBM recommends that XCF use both CTC and list structures.

Note: You do not need to create the Checkpoint files. However, CA MIM has no issues when you are already using them.


13-Jul-2016 10/86

2.

a.

b.

c.

3.

4.

5.

Activate your XCF communication.

Verify that CA MIM automatically created your XCF paths on GK03:

@D PATH MIM0067I Command DISPLAY 960MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:30:23 ON 2013.171 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 XCF GK13 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0


@D PATHMIM0067I Command DISPLAY 444MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:30:45 ON 2013.171 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 N/A GK13 LOCAL 0 0 0 XCF GK03 USABLE 0 0 0


@D PATH MIM0067I Command DISPLAY 837MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:30:53 ON 2013.171 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK62 LOCAL 0 0 0 XCF GK03 USABLE 0 0 0 XCF GK13 USABLE 0 0 0

CA MIM automatically creates a GLOBALVALUE, when not already coded in the INIT member:

@D GLOBALVALUEMIM0067I Command DISPLAY 956MIM0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=NO NOMASTER=WAIT VCFMASTER=NONE ELIGIBLE MASTER LIST: GK03, GK13, GK62

Order the VCFMASTER list from the busiest system to the least busy system.Performance is improved as it reduces the total amount of I/O.

Enter the DISPLAY IO command on all systems and pick the system with the largest RATE: CYC.In this example, we have determined that GK62 is the busiest, followed by GK13, followed by GK03.

Enter the GLOBALVALUE command on one system only:

@GLOBALVALUE VCFMASTER(GK62,GK13,GK03)MIM0067I Command GLOBALVALUE 263MIM0193I GLOBALVALUE command processing initiatedMIM0363I GLOBALVALUE command changes pending on external systems MIM0362I GLOBALVALUE command processing complete


13-Jul-2016 11/86

6.

7.

8.

9.

1.

2.

Verify the new VCFMASTER list:

@D GLOBALVMIM0067I Command DISPLAY 273 MIM0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=NO NOMASTER=WAIT VCFMASTER=( GK62, GK13, GK03 ) ELIGIBLE MASTER LIST: GK62, GK13, GK03

Update the initialization member with the new GLOBALVALUE statement.

Use the MIGRATE command and move the control file up to the VCF on GK62.

@MIGRATE MASTER=GK62MIM0067I Command MIGRATE 307MIM0232I VCF migration SCHEDULED - MASTER=GK62MIM0377I VCF migration IN PROGRESS - MASTER=GK62MIM0342I system GK62 VCF activation COMPLETE - MASTER=GK62

Confirm that the Virtual Control file is active on GK62:

@D SYSMIM0067I Command DISPLAY 211MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL ACTIVE zOS 2013.171 14:54:51.19 02 13 GK13 EXTERNAL ACTIVE zOS 2013.171 14:54:51.14 03 62 GK62 LOCAL MASTER zOS 2013.171 14:54:51.22

The XCF communication is activated and GK62 is now the VCFMASTER.

Revert to the Original Communication MethodYou can revert CA MIM to the original DASD file or XES list structure using the MIGRATE command.

Follow these steps:

Use the MIGRATE command:

@MIG CF=00MIM0067I Command MIGRATE 531MIM0227I Migration initiated to control file 00MIM0088I migration to file 00 IN PROGRESSMIM0242I system GK62 VCF deactivation COMPLETEMIM0080I migration to file 00 COMPLETEMIM0638I Global Copy Received

Display the files and confirm that the original DASD file or XES structure is in-use:

@D FILESMIM0067I Command DISPLAY 610MIM0102I Control File Display: File Unit Status Volser Data Set 00 IN-USE CFCCMIM2 KERGL01#TABLE00 01 USABLE CFCCMIM1 KERGL01#TABLE01

You have reverted to the original DASD file or XES structure.


13-Jul-2016 12/86

1.

a.

b.

c.

Activate List Structures when COMMUNICATION=XCF or CTCONLY

This procedure dynamically activates XES list structures when COMMUNICATION=XCF or CTCONLY. The CA MIM system names are GK03, GK13, and GK62.


The following steps create a primary and a secondary list structure. The primary list structure is the size of the largest DISPLAY CFSIZE command. The secondary list structure is about 20 percent larger.

Follow these steps:

Set up your list structures:

Confirm that all systems in the MIMplex are also within the same sysplex.

Enter the DISPLAY CFSIZE command on all systems and determine the optimal size for the new XES list structures.

@D CFSIZEMIM0067I Command DISPLAY 726MIM1154I GDIF CFSIZE Display: The recommended primary GDIF control file size is 10240 KB.

The MIM1154I CFSIZE display indicates that the recommended Control File size is 10240 KB.

Allocate the list structures using the sample JCL contained in the ALLOCSTR member of the MIM.CNTL library.


13-Jul-2016 13/86

2.

a.

b.

c.

d.

3.

Activate the list structures.

Allocate the new XES list structures to the MIM address space by using the ALLOCATE command:

@ALLOC XESFILEID=00 STRNAME=KERGL01#TABLE00IXL014I IXLCONN REQUEST FOR STRUCTURE KERGL01#TABLE00 920WAS SUCCESSFUL. JOBNAME: MIM12MIM ASID: 0088CONNECTOR NAME: MIM12MIMGK62 CFNAME: CFCCMIM2MIM0472I 2049 blocks formatted on KERGL01#TABLE00MIM0067I Command ALLOCATE 922MIM0153I ALLOCATION successful @ALLOC XESFILEID=01 STRNAME=KERGL01#TABLE01IXL014I IXLCONN REQUEST FOR STRUCTURE KERGL01#TABLE01 925WAS SUCCESSFUL. JOBNAME: MIM12MIM ASID: 0088CONNECTOR NAME: MIM12MIMGK62 CFNAME: CFCCMIM1MIM0067I Command ALLOCATE 926MIM0153I ALLOCATION successful

Enter the ALLOCATE commands on all systems.

Confirm the successful allocation with the following DISPLAY FILES command on all systems:

@D FILESMIM0067I Command DISPLAY 937MIM0102I Control File Display: File Unit Status Volser Data Set 00 USABLE CFCCMIM2 KERGL01#TABLE00 01 USABLE CFCCMIM1 KERGL01#TABLE01

Use the MIGRATE command for migrating to the XES list structure.

@MIG CF=00 MIM0067I Command MIGRATE 767MIM0227I Migration initiated to control file 00MIM0088I migration to file 00 IN PROGRESSMIM0242I system GK62 VCF deactivation COMPLETEMIM0080I migration to file 00 COMPLETEMIM0638I Global Copy Received

The migration is now complete.

Verify that CA MIM is using the proper file by issuing the following DISPLAY FILES command:

@D FILESMIM0067I Command DISPLAY 780MIM0102I Control File Display: File Unit Status Volser Data Set 00 IN-USE CFCCMIM2 KERGL01#TABLE00 01 USABLE CFCCMIM1 KERGL01#TABLE01MIM0104I Checkpoint File Display: File Unit Status Volser Data Set 00 2064 IN-USE TSO23C KERGL01.XE62.CKPT00 01 2064 USABLE TSO23C KERGL01.XE62.CKPT01

The list structures are activated and file 00 is being used, which is correct.


13-Jul-2016 14/86

1.

2.

Revert to the Original List StructuresYou see a problem and want CA MIM to revert to use the XCF signaling(if COMMUNICATION=XCF) or the CTC communication (if COMMUNICATION=CTCONLY).

Follow these steps:

Issue the MIGRATE command:

@MIGRATE MASTER=GK03MIM0067I Command MIGRATE 813MIM0232I VCF migration SCHEDULED - MASTER=GK03MIM0377I VCF migration IN PROGRESS - MASTER=GK03MIM0342I system GK62 VCF activation COMPLETE - MASTER=GK03MIM0241I VCF migration COMPLETE - MASTER=GK03MIM0637I Global Copy Sent to GK03 GK13

Confirm that GK03 is now the VCFMASTER by using the DISPLAY SYSTEMS command:

@D SYSTEMSMIM0067I Command DISPLAY 825MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL MASTER zOS 2013.172 11:39:35.68 02 13 GK13 EXTERNAL ACTIVE zOS 2013.172 11:39:35.61 03 62 GK62 LOCAL ACTIVE zOS 2013.172 11:39:35.69

Activate XCF Communication on a Subset of the Systems

This procedure dynamically activates XCF communication on a subset of the MIMplex when COMMUNICATION=CTCDASD or CTCONLY. The MIMplex is four systems with system names of GK03, GK13, GK23, and GK62. GK03, GK13, and GK62 are in one sysplex. GK23 is in a different sysplex.


Follow these steps:


13-Jul-2016 15/86

1.

2.

Set up your XCF communication on a subset of systems:

Set MIMINIT VCFBUFFERSIZE to 32768 or larger.

Review the XCF configuration and options in the SYS1.PARMLIB(COUPLExx) member.

You do not need to define a special transport class to CA MIM. However, we recommend at least one CLASSLEN of 32768 or larger to avoid possible performance problems. CLASSLEN is an XCF parameter and is coded in the SYS1.PARMLIB(COUPLExx) member.

If some of the LPARS are geographically remote, then IBM recommends that XCF use both CTC and list structures.

Note: You do not need to create the Checkpoint files. However, if you are already using them CA MIM has no issues.

Confirm the current paths by entering the DISPLAY PATH command on all four systems.

On GK03:

@D PATHMIM0067I Command DISPLAY 422MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:10 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 XCF GK13 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0.000382 3,269 184A GK62 USABLE IDLE 3,268 3,268 0 180A GK62 USABLE IDLE 0 0 0 0B4A GK62 USABLE IDLE 0 0 0 186A GK23 USABLE IDLE 3,268 3,268 0 182A GK23 USABLE IDLE 0 0 0 0B6A GK23 USABLE IDLE 0 0 0 183A GK13 USABLE IDLE 3,254 3,254 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0

In the previous display, CA MIM automatically created the XCF paths to external systems that are in the same sysplex as the local system. No actions were required to accomplish this task. Also notice that CA MIM does not show the path currently IN-USE. Issue the DISPLAY PATH command on client systems to determine the current path.


13-Jul-2016 16/86

2.

On GK13:

@D PATHMIM0067I Command DISPLAY 673MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:20 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 N/A GK13 LOCAL 0 0 0 187A GK62 USABLE IDLE 0 0 0 0BFA GK62 USABLE IDLE 0 0 0 0B7A GK62 USABLE IDLE 0 0 0 185A GK23 USABLE IDLE 0 0 0 181A GK23 USABLE IDLE 0 0 0 0B5A GK23 USABLE IDLE 0 0 0>183A GK03 IN-USE BUSY 3,540 3,540 0.001925 3,540 RESERVE 154A GK03 USABLE IDLE 0 0 0 0B3A GK03 USABLE IDLE 0 0 0 XCF GK03 USABLE 0 0 0

Note the XCF paths to GK03 and GK62 in the previous display.

On GK23:

@D PATHMIM0067I Command DISPLAY 327MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:29 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK23 LOCAL 0 0 0 183A GK62 USABLE IDLE 0 0 0 154A GK62 USABLE IDLE 0 0 0 0B3A GK62 USABLE IDLE 0 0 0 185A GK13 USABLE IDLE 0 0 0 181A GK13 USABLE IDLE 0 0 0 0B5A GK13 USABLE IDLE 0 0 0>186A GK03 IN-USE IDLE 3,657 3,657 0.002053 3,657 182A GK03 USABLE IDLE 0 0 0 0B6A GK03 USABLE IDLE 0 0 0

The previous display shows no XCF paths to GK03, GK13, and GK62 because GK23 is in a different sysplex.

On GK62:

@D PATHMIM0067I Command DISPLAY 136MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:36 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK62 LOCAL 0 0 0 183A GK23 USABLE IDLE 0 0 0 154A GK23 USABLE IDLE 0 0 0 0B3A GK23 USABLE IDLE 0 0 0 187A GK13 USABLE IDLE 0 0 0 0BFA GK13 USABLE IDLE 0 0 0 0B7A GK13 USABLE IDLE 0 0 0 >184A GK03 IN-USE IDLE 3,796 3,796 0.002015 3,796 180A GK03 USABLE IDLE 0 0 0 0B4A GK03 USABLE IDLE 0 0 0 XCF GK03 USABLE 0 0 0 XCF GK13 USABLE 0 0 0


13-Jul-2016 17/86

2.

3.

4.

5.

So, all four systems are currently using CTCs to communicate, but systems GK03, GK13, and GK62 can now communicate through XCF. Use the SET MIM VCFPREFERENCE=CTC/XCF command for specifying whether CA MIM uses the CTC or XCF connection. The default is CTC when COMMUNICATION=CTCONLY or CTCDASD.

Activate your XCF communication

Issue the SET MIM VCFPREFERENCE=XCF command on GK13 and GK62 to switch from CTC to XCF. The command is not needed on the current VCFMASTER(GK03) because it detects the switch from the clients.

On GK13:

@SET MIM VCFPREF=XCFMIM0067I Command SETOPTION 169MIM0052I SETOPTION MIM processing completeMIM0292I VCF switched to alternate VCF path XCF

On GK62:

@SET MIM VCFPREF=XCFMIM0067I Command SETOPTION 169MIM0052I SETOPTION MIM processing completeMIM0292I VCF switched to alternate VCF path XCF

The activation is now complete.

Update the commands member with:

IFSYS GK03,GK13,GK62SET MIM VCFPREFERENCE=XCFENDIF

Note: GK23 was not included in the command as we want it to continue using CTC which is the default for COMMUNICATION=CTCDASD or CTCONLY.

Verify the current connectivity by entering the DISPLAY PATH command on all client systems.

On GK62:

@D PATHMIM0067I Command DISPLAY 180MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:36 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK62 LOCAL 0 0 0 183A GK23 USABLE IDLE 0 0 0 154A GK23 USABLE IDLE 0 0 0 0B3A GK23 USABLE IDLE 0 0 0 187A GK13 USABLE IDLE 0 0 0 0BFA GK13 USABLE IDLE 0 0 0 0B7A GK13 USABLE IDLE 0 0 0 184A GK03 USABLE IDLE 4,769 4,769 0.001856 4,769 180A GK03 USABLE IDLE 0 0 0 0B4A GK03 USABLE IDLE 0 0 0>XCF GK03 IN-USE 95 95 0.000408 95 XCF GK13 USABLE 0 0 0


13-Jul-2016 18/86

5.

1.

Notice how the XCF path is now being used between GK62 and GK03.

On GK13:

@D PATHMIM0067I Command DISPLAY 776MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:20 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 N/A GK13 LOCAL 0 0 0 187A GK62 USABLE IDLE 0 0 0 0BFA GK62 USABLE IDLE 0 0 0 0B7A GK62 USABLE IDLE 0 0 0 185A GK23 USABLE IDLE 0 0 0 181A GK23 USABLE IDLE 0 0 0 0B5A GK23 USABLE IDLE 0 0 0 183A GK03 USABLE IDLE 5,266 5,266 0.001632 5,266 154A GK03 USABLE IDLE 0 0 0 0B3A GK03 USABLE IDLE 0 0 0>XCF GK03 IN-USE 105 105 0.001399 105

Notice how the XCF path is now being used between GK13 and GK03.

On GK23:

MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:29 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK23 LOCAL 0 0 0 183A GK62 USABLE IDLE 0 0 0 154A GK62 USABLE IDLE 0 0 0 0B3A GK62 USABLE IDLE 0 0 0 185A GK13 USABLE IDLE 0 0 0 181A GK13 USABLE IDLE 0 0 0 0B5A GK13 USABLE IDLE 0 0 0>186A GK03 IN-USE IDLE 3,657 3,657 0.002053 3,657 182A GK03 USABLE IDLE 0 0 0 0B6A GK03 USABLE IDLE 0 0 0

GK23 is still using a CTC connection to GK03 because there are no XCF paths between GK23 and GK03.

Revert to CTC CommunicationYou decide to revert CA MIM back to using the CTC communication.

Follow these steps:

Use the SET MIM VCFPREFERENCE=CTC command.

On GK13:

@SET MIM VCFPREFERENCE=CTCMIM0067I Command SETOPTION 035MIM0052I SETOPTION MIM processing completeMIM0292I VCF switched to alternate VCF path 183A@D PATHMIM0067I Command DISPLAY 039MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:51:19 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES

N/A GK13 LOCAL 0 0 0


13-Jul-2016 19/86

1.

1.

N/A GK13 LOCAL 0 0 0 187A GK62 USABLE IDLE 0 0 0 0BFA GK62 USABLE IDLE 0 0 0 0B7A GK62 USABLE IDLE 0 0 0 185A GK23 USABLE IDLE 0 0 0 181A GK23 USABLE IDLE 0 0 0 0B5A GK23 USABLE IDLE 0 0 0>183A GK03 IN-USE IDLE 29 29 0.000692 29 154A GK03 USABLE IDLE 0 0 0 0B3A GK03 USABLE IDLE 0 0 0 XCF GK03 USABLE 46,770 46,770 0.002763 46,770 XCF GK62 USABLE 0 0 0

In the previous display, you can see that GK13 is now using CTC address 183A to GK03.

On GK62:

@SET MIM VCFPREF=CTCMIM0067I Command SETOPTION 6350052I SETOPTION MIM processing completeMIM0292I VCF switched to alternate VCF path 184A@D PATHMIM0067I Command DISPLAY 697MIM0176 MIM PATH DISPLAY:LAST RESTART AT 08:38:36 ON 2013.176 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK62 LOCAL 0 0 0 183A GK23 USABLE IDLE 0 0 0 154A GK23 USABLE IDLE 0 0 0 0B3A GK23 USABLE IDLE 0 0 0 187A GK13 USABLE IDLE 0 0 0 0BFA GK13 USABLE IDLE 0 0 0 0B7A GK13 USABLE IDLE 0 0 0>184A GK03 IN-USE IDLE 4,794 4,794 0.001852 4,794 180A GK03 USABLE IDLE 0 0 0 0B4A GK03 USABLE IDLE 0 0 0 XCF GK03 USABLE 60,408 60,408 0.003041 60,408 XCF GK13 USABLE 0 0 0

In the previous display, you can see that GK62 is now using CTC address 184A to GK03.

No commands are required for GK23 because it was already using CTCs. The command is not needed on the current VCFMASTER(GK03) because it detects the switch from the clients.

Add a System to the MIMplex when COMMUNICATION=CTCDASD

This procedure dynamically adds a system to an existing MIMplex when COMMUNICATION=CTCDASD or CTCONLY.


Follow these steps:

Set up your new system.


13-Jul-2016 20/86

1.

a.

b.

c.

2.

3.

Review the current configuration.In this example, we have a two system MIMplex with systems GK03 and GK13. We are going to add system GK62.The current INIT member contains the following statements:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED) GLOBALVALUE VCFMASTER=(GK03,GK13), MOSTP=YES, NOMASTER=WAIT, ANYELIG=YESCTCPATH FROMSYS=GK03 TOSYS=GK13 ADDR=(183A,154A,B3A)CTCPATH FROMSYS=GK13 TOSYS=GK03 ADDR=(183A,154A,B3A)

Confirm the current configuration on GK13 using the following DISPLAY commands:

@D SYSMIM0067I Command DISPLAY 355MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL MASTER zOS 2013.175 14:40:17.11 02 13 GK13 LOCAL ACTIVE zOS 2013.175 14:40:17.20 @D PATHMIM0067I Command DISPLAY 359MIM0176 MIM PATH DISPLAY: LAST RESTART AT 14:39:35 ON 2013.175 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK13 LOCAL 0 0 0>183A GK03 IN-USE IDLE 305 305 0.001085 305 154A GK03 USABLE IDLE 0 0 0 0B3A GK03 USABLE IDLE 0 0 0 @D GLOBALVMIM0067I Command DISPLAY 372MIM0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13 ) ELIGIBLE MASTER LIST: GK03, GK13

Confirm the current configuration on GK03 using the DISPLAY PATH command:

@D PATHMIM0067I Command DISPLAY 237MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:39:23 ON 2013.175 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK03 LOCAL 0 0 0.000024 624 183A GK13 USABLE IDLE 621 621 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0

Activate the new system GK62 using the DEFSYS command on one system only:

@DEFSYS (GK62,62,XE62)MIM0067I Command DEFSYS 319MIM0099I DEFSYS command is processingMIM0408I System Add Successful - GK62

Confirm that GK62 has been added successfully:

@D SYSMIM0067I Command DISPLAY 327MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS

01 03 GK03 LOCAL MASTER zOS 2013.175 14:44:03.55


13-Jul-2016 21/86

3.

4.

5.

6.

01 03 GK03 LOCAL MASTER zOS 2013.175 14:44:03.55 02 13 GK13 EXTERNAL ACTIVE zOS 2013.175 14:44:03.55 03 62 GK62 EXTERNAL FREED <- new system

(Optional) Update the VCFMASTER parameter on the GLOBALVALUE statement. This list shows the systems from busiest to least busy (in terms of the global ENQ activity).In this example, GK62 has the least amount of activity, so we want to put this system last in the list:

@GLOBALVALUE VCFMASTER(GK03,GK13,GK62) MIM0067I Command GLOBALVALUE 360MIM0386W System GK62 is ineligible to become VCF MASTERMIM0193I GLOBALVALUE command processing initiatedMIM0363I GLOBALVALUE command changes pending on external systems MIM0362I GLOBALVALUE command processing complete

Confirm the change:

@D GLOBALVMIM0067I Command DISPLAY 367MIM0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK62 ) ELIGIBLE MASTER LIST: GK03, GK13

GK62 becomes eligible when it is started.Test the connectivity of the devices using the PROCCTC utility before you use the CTCPATH command.

Issue the CTCPATH command to create the CTCPATH definitions on GK03 and GK13.

On GK03:

@CTCPATH FROMSYS=GK03 TOSYS=GK62 ADDR=(184A,180A,B4A)MIM0067I Command CTCPATH 500MIM0703I Path added for device 184AMIM0703I Path added for device 180AMIM0703I Path added for device 0B4A

Confirm the new paths:

@D PATHMIM0067I Command DISPLAY 520MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:39:23 ON 2013.175 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES 0B4A GK62 USABLE BUSY 0 0 0 WRITHDS 180A GK62 USABLE BUSY 0 0 0 WRITHDS 184A GK62 USABLE BUSY 0 0 0 WRITHDS N/A GK03 LOCAL 0 0 0.000055 4,980 183A GK13 USABLE IDLE 4,969 4,969 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0

On GK13:

@CTCPATH FROMSYS=GK13 TOSYS=GK62 ADDR=(187A,BFA,B7A)MIM0067I Command CTCPATH 685MIM0703I Path added for device 187AMIM0703I Path added for device 0BFAMIM0703I Path added for device 0B7A



13-Jul-2016 22/86

6.

7.

8.

9.


@D PATHMIM0067I Command DISPLAY 695MIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:39:35 ON 2013.175 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES 0B7A GK62 USABLE BUSY 0 0 0 WRITHDS 0BFA GK62 USABLE BUSY 0 0 0 WRITHDS 187A GK62 USABLE BUSY 0 0 0 WRITHDS N/A GK13 LOCAL 0 0 0 >183A GK03 IN-USE IDLE 4,859 4,859 0.001026 4,859 154A GK03 USABLE IDLE 0 0 0 0B3A GK03 USABLE IDLE 0 0 0

Update the INIT member with the contents of the previous commands and add the CTCPATHs from GK62:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED), (GK62,62,XE62,INIT=FREED) GLOBALVALUE VCFMASTER=(GK03,GK13,GK62), MOSTP=YES, NOMASTER=WAIT, CTCPATH FROMSYS=GK03 TOSYS=GK13 ADDR=(183A,154A,B3A)CTCPATH FROMSYS=GK03 TOSYS=GK62 ADDR=(184A,180A,B4A) CTCPATH FROMSYS=GK13 TOSYS=GK03 ADDR=(183A,154A,B3A)CTCPATH FROMSYS=GK13 TOSYS=GK62 ADDR=(187A,BFA,B7A) <-new CTCPATH FROMSYS=GK62 TOSYS=GK03 ADDR=(184A,180A,B4A)CTCPATH FROMSYS=GK62 TOSYS=GK13 ADDR=(187A,BFA,B7A) <-new

Note: If COMM=CTCDASD, ensure that the control files or list structures are allocated or cataloged on the new system.

Start MIM on GK62:

S MIM,FORMAT=CHKPT

The new system joins the existing MIMplex.

Note: COMMUNICATION=CTCDASD does not require a checkpoint file. However, if you have one allocated due to using ECMF REQUEUE, then format it.

Verify that the new system joined the MIMplex by entering some verification commands on the new system/GK62:

@D SYSMIM0023I system GK62 in file VCF synchronization COMPLETEMIM0067I Command DISPLAY 676MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS


13-Jul-2016 23/86

9.

1.

2.

1.

INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL MASTER zOS 2013.175 14:53:19.61 02 13 GK13 EXTERNAL ACTIVE zOS 2013.175 14:53:19.66 03 62 GK62 LOCAL ACTIVE zOS 2013.175 14:53:19.67 GK62 is now ACTIVE. @D GLOBALVALUE MIM0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, GK62 ) ELIGIBLE MASTER LIST: GK03, GK13, GK62 GK62 is now an eligible master. @D PATHMIM0176 MIM PATH DISPLAY:LAST RESTART AT 14:52:58 ON 2013.175 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES N/A GK62 LOCAL 0 0 0 187A GK13 USABLE IDLE 0 0 0 0BFA GK13 USABLE IDLE 0 0 0 0B7A GK13 USABLE IDLE 0 0 0>184A GK03 IN-USE IDLE 3 3 0.001479 3 180A GK03 USABLE IDLE 0 0 0 0B4A GK03 USABLE IDLE 0 0 0

All 6 CTC addresses have been allocated and 184A is the path currently IN-USE to the VCFMASTER(GK03).

Revert to the Original MIMplexWhen the new system does not join the MIMplex:

Shut down CA MIM on that system.

Enter the FREE command from one of the active systems.

Once the system has been added with the DEFSYS command, there is no way to remove it. You may need to restore the INIT member to the original state, and perform a global shutdown and control file/checkpoint file format.

Add a System when COMMUNICATION=DASDONLY

You want a system dynamically added to the established MIMplex using COMMUNICATION=DASDONLY. The CA MIM name and alias are honored as coded on the DEFSYS statement.

Follow these steps:

Ensure that no systems have been REMOVED from the MIMPLEX by reviewing the DISPLAY SYSTEMS command output:

MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS

01 03 GK03 EXTERNAL ACTIVE zOS 2012.013 12:25:45.38


13-Jul-2016 24/86

1.

2.

3.

4.

1.

2.

3.

4.

01 03 GK03 EXTERNAL ACTIVE zOS 2012.013 12:25:45.38 02 13 GK13 LOCAL ACTIVE zOS 2012.013 12:25:45.91 03 23 GK23 EXTERNAL FREED

Verify that there are no gaps between the index numbers. If there are no gaps, then it is safe to proceed with the next step. Otherwise, update the DEFSYS statement, shut down globally, then restart globally with a control file format.

Update the DEFSYS statement in the CA MIMINIT member:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED), (GK23,23,XE23,INIT=FREED), (GK62,62,XE62,INIT=FREED) /* new system

Start MIM on the new LPAR (GK62 in this example) with FORMAT=CHKPT or FORMAT=NONE if no checkpoints exist.The new system joins the established MIMplex and honors the CA MIM name and alias as coded on the DEFSYS statement.

Confirm the new system joined with the DISPLAY SYSTEMS command:

MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL ACTIVE zOS 2012.013 12:26:26.39 02 13 GK13 EXTERNAL ACTIVE zOS 2012.013 12:26:26.88 03 23 GK23 EXTERNAL FREED 04 62 GK62 LOCAL ACTIVE zOS 2012.013 12:26:27.36

The new system has been added to the MIMplex.

Add a System when COMMUNICATION=XCFYou want a system dynamically added to the established MIMplex using COMMUNICATION=XCF. The CA MIM name and alias are honored as coded on the DEFSYS statement.

Follow these steps:

Issue the DISPLAY INIT command to ensure that the MIMPLEX is running with COMMUNICATION=XCF and COMPATLEVEL= 11.9 or greater.

Issue the DISPLAY SYSTEMS command to verify that you have an active MIMPLEX. The new system cannot be the first active system in the MIMPLEX.

MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 MIMMC003 LOCAL MASTER zOS 2012.026 10:52:21.82 02 13 MIMMC013 EXTERNAL FREED

Issue the DEFSYS command with your preferred parameters.

@DEFSYS (MIMMC062,62,XE62)

Verify that the following message appears:

MIM0408I - System Add Successful - MIMMC062

Issue the DISPLAY SYSTEMS command to ensure that the new system was added correctly.


13-Jul-2016 25/86

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 MIMMC003 LOCAL MASTER zOS 2012.026 10:57:06.16 02 13 MIMMC013 EXTERNAL FREED 03 62 MIMMC062 EXTERNAL FREED

For a new system to become a VCFMASTER system, update the GLOBALVALUE parameter.

Issue a DISPLAY GLOBALVALUE command. Note the list of eligible master systems.

MIM0373I MIM GLOBALVALUE display:ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( MIMMC003, MIMMC013 )ELIGIBLE MASTER LIST: MIMMC003, MIMMC013

Issue the following command on one system only:

@GLOBALVALUE VCFMASTER={LIST}

Note: Copy the VCFMASTER list from the prior command and add the desired system.

@GLOBALVALUE VCFMASTER=( MIMMC003, MIMMC013 , MIMMC062)

Verify Output:

MIM0363I GLOBALVALUE command changes pending on external systemsGLOBALVALUE command processing complete

Issue a DISPLAY GLOBALVALUE command again to verify that the desired system was added to the VCFMASTER list.

MIM0373I MIM GLOBALVALUE display:ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAITVCFMASTER=( MIMMC003, MIMMC013, MIMMC062)ELIGIBLE MASTER LIST: MIMMC003, MIMMC013, MIMMC062

Update the GLOBALVALUE statement in the MIMINIT member to match the exact text of the operator command.

GLOBALVALUE VCFMASTER=(MIMMC003,MIMMC013, MIMMC062), ANYELIGIBLE=YES, NOMASTER=WAIT, MOSTPREFERRED=YES

Update the DEFSYS statement in the MIMINIT member. Be sure to code the new system last on the system list.

DEFSYS (MIMMC003,03,XE03,INITIAL=FREED), (MIMMC013,13,XE13,INITIAL=FREED), (MIMMC062,62,XE62)

Allocate checkpoint files for the new system. Defined them using the CHKPTDSN statement in the MIMINIT member.

MIMINIT CHKPTDSN=CHIMA14.MIMMC0&SYSCLONE..MIMCKP

Start the new system with FORMAT=CHKPT.

Confirm that the new system has been added by entering a DISPLAY SYSTEMS command.


13-Jul-2016 26/86

13. Confirm that the new system has been added by entering a DISPLAY SYSTEMS command.

MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 MIMMC003 EXTERNAL MASTER zOS 2012.026 11:05:55.14 02 13 MIMMC013 EXTERNAL FREED 03 62 MIMMC062 LOCAL ACTIVE zOS 2012.026 11:05:55.2

The new system has been added.


13-Jul-2016 27/86

How to Offload CA MIM Work to zIIP EnginesThis content illustrates the process z/OS systems programmers use to activate the CA MIM for z/OS zIIP enablement feature. This scenario describes how to verify your z/OS and CA MIM for z/OS environments before activating the feature, and how to verify results after activating it.

Prerequisites (see page 27)About zIIP Technology (see page 28)How You Offload CA MIM Work to zIIP Engines (see page 29)Verify CCS CAIRIM Component Release and Service Levels (see page 30)Verify MIMplex Systems Have Online zIIP Engines (see page 30)Verify CA MIM is Active (see page 32)Assess CA MIM Workload to be Offloaded to zIIP Engines (see page 33)Activate the zIIP Enablement Feature (see page 36)Verify zIIP Enablement Feature is Active (see page 37)Assess CA MIM Workload being Offloaded to zIIP Engines (see page 38)Results of zIIP Enablement Feature Offload (see page 42)Summary (see page 44)

PrerequisitesYou must have the following items in place before you activate the CA MIM zIIP Enablement feature.

CA MIM calls zIIP services that the CA Common Services CAIRIM component provides. Install one of these releases and PTFs that deliver zIIP services through the CCS CAIRIM component:

CCS r14.0 + PTF RO32488 + All published CCS zIIP Service PTFs

CCS r14.1 + All published CCS zIIP Service PTFs

CA MIM Release 12.0 or later installed.

At least one System z Integrated Information Processor (zIIP) engine online and operational.

Before You BeginReview the following considerations:

Do you have zIIP engine capacity available on one or more systems in your MIMplex?

Do you want to reduce CA MIM operational costs by offloading CA MIM address space work to the zIIP engines from the general processor engines?


13-Jul-2016 28/86

Review the following commands in Statement and Commands Reference (https://docops.ca.com/display/CMFZS125/Statement+and+Command+Reference):

DISPLAY CPUTIME

DISPLAY FACILITIES

DISPLAY OPTIONS

DISPLAY SYSTEMS

SETOPTION MIM ZIIP

Review the z/OS DISPLAY M=CPU command in the IBM z/OS MVS System Commands Guide.

Review the IEE174I message in the IBM z/OS MVS System Messages Guide.

Review information about stopping and restarting CA MIM address spaces, in CA MIM for z/OS Programming. (https://docops.ca.com/display/CMFZS125/CA+MIM+Programming)

About zIIP TechnologyIBM offers a special purpose mainframe engine that is known as the System z Integrated Information Processor (zIIP). The purpose of this processor is to offload specialized workloads to help reduce mainframe operating costs.

The type of work capable or allowed to execute on zIIP engines is limited to SRB-mode work executing within an Enclave. The software instructions that can run while in SRB-mode have special requirements. These requirements limit the type of software instructions that are eligible to run on zIIP engines. Because of these limitations, not all workloads can be offloaded to zIIP engines. However, we are happy to report that some percentage of CA MIM for z/OS address space work is now eligible to be offloaded to zIIP engines.

zIIP engines provide cost savings. The IBM monthly license fees to its mainframe customers are based on consumed CP engine service units, not zIIP engine service units consumed. Therefore, software products like CA MIM for z/OS that can have their work offloaded to zIIP engines can help mainframe customers:

Reduce their Total Cost of Ownership (TCO)

Defer costly upgrades to the higher capacity CPU hardware

For more information about CA products exploiting zIIP engines, see Leveraging Mainframe Specialty Engines: zIIP, zAAP, and IFL - CA Technologies (http://www.ca.com/us/collateral/technology-initiatives/na

./Leveraging-Mainframe-Specialty-Engines-zIIP-zAAP-and-IFL-Lead.aspx)

https://docops.ca.com/display/CMFZS125/Statement+and+Command+Reference



https://docops.ca.com/display/CMFZS125/CA+MIM+Programming


http://www.ca.com/us/collateral/technology-initiatives/na/Leveraging-Mainframe-Specialty-Engines-zIIP-zAAP-and-IFL-Lead.aspx





13-Jul-2016 29/86

1.

2.

3.

4.

5.

How You Offload CA MIM Work to zIIP EnginesThe following illustration shows the steps to offload CA MIM for z/OS work to zIIP engines:

Perform the following steps to offload work to a zIIP engine.

Verify CCS CAIRIM Component Release and Service Levels (see page 30)

Verify MIMplex Systems have Online zIIP Engines (see page 30)

Verify CA MIM is Active (see page 32)

Assess CA MIM Workload to be Offloaded to zIIP Engines (see page 33)

Activate the zIIP Enablement Feature (see page 36)


13-Jul-2016 30/86

6.

7.

8.

1.

Verify zIIP Enablement Feature is Active (see page 37)

Assess CA MIM Workload being Offloaded to zIIP Engines (see page 38)

Results of zIIP Enablement Feature Offload (see page 42)

Verify CCS CAIRIM Component Release and Service Levels

Verify that the CA Common Services CAIRIM component is installed at the proper release and service level. For more information, see . Complete these verification steps on all Prerequisites (see page 27)systems in the MIMplex.

Follow these steps:

In the CAS9 JOBLOG, the CAS9075I message from the CAIRIM component shows the release level at which the CAIRM component is running.

The CAS9075I message for the CCS 14.0 release:

CAS9075I - SERVICE(CA-RIM/BASE) VERS(1400) GENLVL(E000AS900)

The CAS9075I message for the CCS 14.1 release:

CAS9075I - SERVICE(CA-RIM/BASE) VERS(1400) GENLVL(E100AS900)

2. To determine whether the required CAIRIM prerequisite zIIP Enablement service PTFs are applied, contact your CCS administrator. The required CCS zIIP Enablement Service PTF numbers are shown in

. To ensure the CCS zIIP Enablement Service operates with the highest Prerequisites (see page 27)degree of availability, we also recommend applying all published PTFs associated with the service.

Verify MIMplex Systems Have Online zIIP EnginesVerify that zIIP engines are online and operational.

Follow these steps:

1. To determine the names of the systems in the MIMplex, issue the following command on one system:

F MIM,DISPLAY SYSTEMS

This command identifies the name of each system in the MIMplex system.

2. Review this example.

Example Display Analysis:

SYSA, SYSB, and SYSC are the names of the systems in this MIMplex.


13-Jul-2016 31/86

SYSA, SYSB, and SYSC are the names of the systems in this MIMplex.

Example Display Command:

F MIM,DISPLAY SYSTEMSMIM0108I SYSTEMS DISPLAYINDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS&MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 AA SYSA LOCAL MASTER ZOS 2013.165 07:58:05.16 02 BB SYSB EXTERNAL ACTIVE ZOS 2013.165 07:58:05.16 03 CC SYSC EXTERNAL ACTIVE ZOS 2013.165 07:58:05.10

3. To verify the availability of online zIIP engines, issue the following command on each system in the MIMplex:

DISPLAY M=CPU

This z/OS command identifies the status of general and specialty processors.


Example Display Command Analysis:

SYSA has two general processor engines online and one specialty zIIP engine online.

SYSB has four general processor engines online and two specialty zIIP engines online.

SYSC has six general processor engines online and four specialty zIIP engines online.

IEE174I Message CPU Field Symbol Definitions:

+

Identifies ONLINE general processor engine.

+I

Identifies ONLINE zIIP engine.

-

Identifies OFFLINE general processor engine.

-I

Identifies OFFLINE zIIP engine.

Example Display Commands:

SYSA:

D M=CPUIEE174I 08.03.50 DISPLAY MPROCESSOR STATUSID CPU SERIAL00 + 0AF68999901 + 0AF68999902 +I 0AF68999903 -


13-Jul-2016 32/86

03 -04 -05 -I

SYSB:

D M=CPUIEE174I 08.08.32 DISPLAY MPROCESSOR STATUSID CPU SERIAL00 + 01F68999901 + 01F68999902 + 01F68999903 + 01F68999904 +I 01F68999905 +I 01F68999906 -07 -08 -I09 -I

SYSC:

D M=CPUIEE174I 08.15.52 DISPLAY MPROCESSOR STATUSID CPU SERIAL00 + 0CF68999901 + 0CF68999902 + 0CF68999903 + 0CF68999904 + 0CF68999905 + 0CF68999906 +I 0CF68999907 +I 0CF68999908 +I 0CF69999909 +I 0CF6999990A -0B -

Verify CA MIM is ActiveYou want to verify and display the following information:

Verify that CA MIM for z/OS is running on each MIMplex system.

Display the release and facilities that are running in the CA MIM for z/OS address space.

Follow these steps:

1.To verify the release level and facilities of your CA MIM for z/OS address spaces, issue the following command on each system in the MIMplex:

F MIM,DISPLAY FACILITIES

2. Review the following command example.



13-Jul-2016 33/86

SYSA shows CA MIM for z/OS Release 12.0 running in the CAMIMGR address space with the GDIF and ECMF facilities activated.

SYSB shows CA MIM for z/OS Release 12.0 running in the CAMIMGR address space with the GDIF and ECMF facilities activated.

SYSC shows CA MIM for z/OS Release 12.0 running in the CAMIMGR address space with the GDIF and ECMF facilities activated.


SYSA:

F MIM,DISPLAY FACILITIESMIM0090I CAMIMGR FACILITIES (R12.0 0000) MIM - MULTI-IMAGE MANAGER GDIF - GLOBAL DATA INTEGRITY FACILITY ECMF - ENQ CONFLICT MANAGEMENT FACILITY

SYSB:


SYSC:


Assess CA MIM Workload to be Offloaded to zIIP Engines

You want to obtain information about the CA MIM for z/OS address space workload that could potentially be offloaded from general processor engines to zIIP specialty engines.

Follow these steps:

1.To view the potential CA MIM for z/OS address space workload that could be offloaded to zIIP engines, issue the following command on each system in the MIMplex:

F MIM,DISPLAY CPUTIME=DETAIL

The CPUTIME=DETAIL form of the command shows MIM address space TCBs and SRB work that can potentially be offloaded to zIIP engines. Other forms of the display provide more and less detail about the MIM address space CPU consumption. For example:

F MIM,DISPLAY CPUTIME=ALLF MIM,DISPLAY CPUTIME=SUMMARY


13-Jul-2016 34/86

2. Review the following DISPLAY CPUTIME=DETAIL command example.


In this example, the zIIP Enablement feature has not yet been activated. This means that the CA MIM for z/OS address spaces are running with the following settings:

SETOPTION MIM ZIIP=NO defined in the MIMCMNDS parmlib member

This display data represents CA MIM for z/OS running with none of its TCB mode work being converted to SRB mode work. Therefore, no work is running on zIIP engines.

MIM0660I Message Fields

WORKUNIT – Identifies the TCB names of tasks running in the MIM address space or general system SRB activity that is associated with the MIM address space.

TOTALCPU - Identifies, by TCB, total consumed CP and zIIP engine time.

TASKCP – Identifies, by TCB, the general CP engine time consumed.

ZIIPONCP – Identifies, by TCB, the consumed ZIIP-eligible work time that ran on a general CP engine because zIIP engines were busy.

ZIIP – Identifies, by TCB, the zIIP engine time consumed.

PCT – Identifies, by TCB, the percentages of ZIIPONCP and ZIIP time consumed relative to the TOTALCPU time consumed.

Total/Summary – Identifies the totals for all fields in the display. The values in the Total/Summary line always represent activity for the entire MIM address regardless of which form of the command is used.

Normalized - The values on this line always represent activity for the entire MIM address space regardless of which form of the command is used. A model-dependent scaling factor adjusts the zIIP value from the Total / Summary line. This method reflects the amount of CPU time that would have been used if that work had executed on general CP engines. The zIIP column is the actual amount of general CP engine time that is saved by using the MIM zIIP enablement feature.

DISPLAY CPUTIME=ALL DISPLAY CPUTIME=DETAIL DISPLAY CPUTIME=SUMMARY

MIM0660I Message Time Values

ss.ssssss = seconds.subseconds, up to 59.999999.

hhh:mm:ss = hours:minutes:seconds, up to 999:59:59.

hhhhhh:mm = hours:minutes, up to 999999:99.

Example MIM0660I Message Analysis


13-Jul-2016 35/86

The DISPLAY CPUTIME=DETAIL command shows only TCBs which have the potential to offload work to zIIP engines after the zIIP Enablement feature is activated. The time values shown in the TASKCP field represent the CP time that could be offloaded to zIIP engines once the feature is activated.

The ZIIPONCP, PCT, ZIIP, and PCT fields all contain zeros in the following example display command because the zIIP Enablement feature has not been activated. When you activate the feature, these fields begin to show true time amounts and percentages.

The Total/Summary line identifies the totals for all fields in the display.

Note: The values in the Total/Summary and the Normalized lines always represent activity for the entire MIM address. Not simply the totals for the TCBs which have the potential to offload work to zIIP engines. The DISPLAY CPUTIME=ALL form of the display shows activity for all MIM address space TCBs.

Example Display Command:

SYSA:

F MIM,DISPLAY CPUTIME=DETAILMIM0660I CPU TIME:LAST RESTART AT 20:32:16 ON 2013.16 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 00:24:15 00:24:15 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000013 0.000013 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000442 0.000442 0.000000 0.0 0.000000 0.0 MIMDRVFD 00:23:46 00:23:46 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.097600 0.097600 0.000000 0.0 0.000000 0.0 SRB 1.521509 -------- ---------- ---------- ---------- ---- ---------- ----- Total 00:50:03 00:50:03 0.000000 0.0 0.000000 0.0Normalized 00:50:03 00:50:03 0.000000 0.0 0.000000 0.0

SYSB:

F MIM,DISPLAY CPUTIME=DETAILMIM0660I CPU TIME:LAST RESTART AT 20:32:16 ON 2013.165 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 00:16:13 00:16:13 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000011 0.000011 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000641 0.000641 0.000000 0.0 0.000000 0.0 MIMDRVFD 00:13:46 00:13:46 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.097600 0.097600 0.000000 0.0 0.000000 0.0 SRB 1.512457 -------- ---------- ---------- ---------- ---- ---------- ----- Total 00:32:01 00:32:01 0.000000 0.0 0.000000 0.0Normalized 00:32:01 00:32:01 0.000000 0.0 0.000000 0.0

SYSC:

F MIM,DISPLAY CPUTIME=DETAILMIM0660I CPU TIME:LAST RESTART AT 20:32:16 ON 2013.165 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- -----

MIMDRDRV 00:04:15 00:04:15 0.000000 0.0 0.000000 0.0


13-Jul-2016 36/86

MIMDRDRV 00:04:15 00:04:15 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000017 0.000017 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000459 0.000459 0.000000 0.0 0.000000 0.0 MIMDRVFD 00:04:45 00:04:45 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.097600 0.097600 0.000000 0.0 0.000000 0.0 SRB 1.491234 -------- ---------- ---------- ---------- ---- ---------- ----- Total 00:11:02 00:11:02 0.000000 0.0 0.000000 0.0Normalized 00:11:02 00:11:02 0.000000 0.0 0.000000 0.0

Activate the zIIP Enablement FeatureDefine SETOPTION MIM ZIIP=YES and start the CA MIM for z/OS address spaces.

Running CA MIM for z/OS with SETOPTION MIM ZIIP=NO causes all CA MIM for z/OS TCB mode work to remain unchanged. That is, no eligible TCB mode work is converted to SRB mode work to allow that work to run on zIIP engines. The zIIP Enablement feature is disabled when running with SETOPTION MIM ZIIP=NO defined, which is the default operating mode.

This procedure defines the SETOPTION MIM ZIIP=YES command in the MIMCMNDS parmlib member and starts the MIM address spaces with this value in place.

In the example presented in this section, we illustrate how to predefine the SETOPTION MIM ZIIP=YES command in the MIMCMNDS parmlib member and restart your CA MIM for z/OS address spaces.

General zIIP Enablement Feature Operational Considerations

The zIIP Enablement feature can be activated one system at a time.

A global MIMplex shutdown and restart of MIM is not required to activate the feature.

A control file format restart is not required to activate the feature.

A REUSE=NO restart is not required to activate the feature.

If some systems in the MIMplex have zIIP engines, and others do not, the feature can be activated on the systems that do have zIIP engines, and the feature can be left deactivated on the systems that do not have zIIP engines. In this case, consider using IFSYS/ENDIF commands to nest SETOPTION ZIIP=YES and SETOPTION ZIIP=NO commands as appropriate.

During SETOPTION MIM ZIIP=YES command processing, when no zIIP engines are online, MIM automatically disables the feature and MIM continues running as though SETOPTION MIM ZIIP=NO had been specified. A MIM0658W message is issued in this case.

If a SETOPTION MIM ZIIP=YES command was specified for a given active instance of MIM and all available zIIP engines go offline at a later point in time, the MIM address space continues converting some of its TCB type work to SRB type work, but that work is dispatched by the system to CP engines rather than to zIIP engines. A MIM0658W message is issued in this case.

If you cannot stop and restart MIM and want to activate the feature dynamically, you can do so by issuing an F MIM, SETOPTION MIM ZIIP=YES command to each MIM address space from any authorized command source. Remember to then update the MIMCMNDS parmlib member to retain zIIP feature activation between MIM restarts.


13-Jul-2016 37/86

1.

a.

b.

2.

retain zIIP feature activation between MIM restarts.

Follow these steps:

To activate the zIIP Enablement feature:

Open the MIMCMNDS parmlib member.

Define a SETOPTION MIM ZIIP=YES command option.

Restart your MIM address space. During SETOPTION MIM ZIIP=YES command processing, the CA MIM for z/OS address space checks whether any zIIP engines are online to the system where CA MIM for z/OS is starting. When one or more zIIP engines are online, the CA MIM for z/OS address space begins offloading work to the zIIP engines.

Verify zIIP Enablement Feature is ActiveVerify that the zIIP Enablement feature is active on all designated MIMplex systems.

Follow these steps:

1.To verify the zIIP Enablement feature is active, issue the following command on each MIMplex system:

F MIM,DISPLAY MIM OPTIONS

This command shows CA MIM for z/OS operating values that are established using SETOPTION commands.



SYSA shows ZIIP=YES which means the zIIP Enablement feature is active.

SYSB shows ZIIP=YES which means the zIIP Enablement feature is active.

SYSC shows ZIIP=YES which means the zIIP Enablement feature is active.


SYSA:

F MIM,DISPLAY MIM OPTIONSMIM0038I MIM OPTION DISPLAY: AUTHCHECK=23 HOURS CELLTRACE=NO CFSIZEWARN=75 CMDPREFIX=& X'50' CMDRESPMAX=2048 CMDTIMEOUT=10 CTCVERIFY=2 MINS CYCLES=1 DOWNSYS=AUTOFREE HIBERNATE=3 MINS INTERVAL=1.000 LOCALSTOP=NOABEND LOCKOUT=15.00 MARGIN=30.00 MIHINTERVAL=60 MODE=DEMAND SHUTDOWN=GLOBAL STATCOLLECT=(SUBTYPE=(CF) NOSUBTYPE=(VF FC CP LS)) STATCYCLE=60.00 STATINTERVAL=15 TRACE=(OFF,JOBNAME=*,MAXQUEUED=1024) VCFDEBUG=OFF VCFFORCE=5 MINS VCFMAXDELAY=30.00 VCFMINDORM=.000 MS VCFRECOVERY=180 SETTRACE=NONE SETPRINT=NONE


13-Jul-2016 38/86

SETPRINT=NONE ZIIP=YES

SYSB:

F MIM,DISPLAY MIM OPTIONSMIM0038I MIM OPTION DISPLAY: AUTHCHECK=23 HOURS CELLTRACE=NO CFSIZEWARN=75 CMDPREFIX=& X'50' CMDRESPMAX=2048 CMDTIMEOUT=10 CTCVERIFY=2 MINS CYCLES=1 DOWNSYS=AUTOFREE HIBERNATE=3 MINS INTERVAL=1.000 LOCALSTOP=NOABEND LOCKOUT=15.00 MARGIN=30.00 MIHINTERVAL=60 MODE=DEMAND SHUTDOWN=GLOBAL STATCOLLECT=(SUBTYPE=(CF) NOSUBTYPE=(VF FC CP LS)) STATCYCLE=60.00 STATINTERVAL=15 TRACE=(OFF,JOBNAME=*,MAXQUEUED=1024) VCFDEBUG=OFF VCFFORCE=5 MINS VCFMAXDELAY=30.00 VCFMINDORM=.000 MS VCFRECOVERY=180 SETTRACE=NONE SETPRINT=NONE ZIIP=YES

SYSC:

F MIM,DISPLAY MIM OPTIONSMIM0038I MIM OPTION DISPLAY: AUTHCHECK=23 HOURS CELLTRACE=NO CFSIZEWARN=75 CMDPREFIX=& X'50' CMDRESPMAX=2048 CMDTIMEOUT=10 CTCVERIFY=2 MINS CYCLES=1 DOWNSYS=AUTOFREE HIBERNATE=3 MINS INTERVAL=1.000 LOCALSTOP=NOABEND LOCKOUT=15.00 MARGIN=30.00 MIHINTERVAL=60 MODE=DEMAND SHUTDOWN=GLOBAL STATCOLLECT=(SUBTYPE=(CF) NOSUBTYPE=(VF FC CP LS)) STATCYCLE=60.00 STATINTERVAL=15 TRACE=(OFF,JOBNAME=*,MAXQUEUED=1024) VCFDEBUG=OFF VCFFORCE=5 MINS VCFMAXDELAY=30.00 VCFMINDORM=.000 MS VCFRECOVERY=180 SETTRACE=NONE SETPRINT=NONE ZIIP=YES

Assess CA MIM Workload being Offloaded to zIIP Engines

Display the amount of CA MIM for z/OS address space workload that is being offloaded from general processor engines to zIIP processor engines.

Follow these steps:

1.To view the actual CA MIM for z/OS address space workload being offloaded to zIIP engines, issue the following command on each system in the MIMplex:

F MIM,DISPLAY CPUTIME=ALL

After the zIIP Enablement feature is activated, this command can identify CA MIM for z/OS address space workloads that are being offloaded to zIIP engines.

2. Review the following DISPLAY CPUTIME=ALL command example.



13-Jul-2016 39/86


In this example, the zIIP Enablement feature is activated on each system. The display commands represent CA MIM for z/OS running with certain MIM TCB mode work being converted to SRB mode work. This conversion allows that work to become eligible to run on zIIP engines.

MIM0660I Message Fields

WORKUNIT – Identifies TCB names of a tasks running in the MIM address space or general system SRB activity associated with the MIM address space.

TOTALCPU - Identifies, by TCB, total CP and zIIP engine time consumed.

TASKCP – Identifies, by TCB, the general CP engine time consumed.

ZIIPONCP – Identifies, by TCB, the ZIIP-eligible work time consumed that ran on a general CP engine because zIIP engines were busy.

ZIIP – Identifies, by TCB, the zIIP engine time consumed.

PCT – Identifies, by TCB, the percentages of ZIIPONCP and ZIIP time consumed relative to the TOTALCPU time consumed.

Total/Summary – Identifies the totals for all fields in the display. Note that the values in the Total/Summary line always represent activity for the entire MIM address regardless of which form of the command is used, DISPLAY CPUTIME=ALL, DISPLAY CPUTIME=DETAIL, or DISPLAY CPUTIME=SUMMARY.

Normalized - The values on this line always represent activity for the entire MIM address space regardless of which form of the command is used. A model-dependent scaling factor adjusts the zIIP value from the Total / Summary line. This method reflects the amount of CPU time that would have been used if that work had executed on general CP engines. The zIIP column is the actual amount of general CP engine time that is saved by using the MIM zIIP enablement feature.

MIM0660I Message Time Values

ss.ssssss = seconds.subseconds, up to 59.999999

hhh:mm:ss = hours:minutes:seconds, up to 999:59:59

hhhhhh:mm = hours:minutes, up to 999999:99

Example MIM0660I Message Analysis

The DISPLAY CPUTIME=ALL command shows:

TCBs that have the potential to offload work to zIIP engines after the zIIP Enablement feature is activated

TCBs whose work is not eligible for the zIIP offload

Combined, the display totals show accurate CP and zIIP engine usage for the entire MIM address space.


13-Jul-2016 40/86

The ZIIPONCP, PCT, ZIIP and PCT fields all contain significant digits which indicate the zIIP Enablement feature has been activated. These fields show true time amounts and percentages of zIIP-eligible work that is run on zIIP engines (ZIIP) or zIIP-eligible work that was run on CP engines due to the lack of zIIP processor availability (ZIIPONCP).

The TOTAL line identifies the totals for all fields in the display. The values in the TOTAL line always represent activity for the entire MIM address not merely the totals for the TCBs that have the potential to offload work to zIIP engines. Again, the DISPLAY CPUTIME=ALL form of the display shows activity for all MIM address space TCBs.

The zIIP Enablement feature is activated, and the following percentages of work is being run on zIIP engines.

For SYSA, TOTALCPU time is 02:27:09, and ZIIP time is 02:25:05, which means 98.5% of MIM address space work has been offloaded to zIIP engines. ZIIPONCP time is 51.588534 which means an additional 0.5% of work from the MIM address space was zIIP-eligible but was run on CP engines due to the lack of zIIP processor availability.

For SYSB, TOTALCPU time is 02:06:19, and ZIIP time is 01:58:35, which means 98.5% of MIM address space work has been offloaded to zIIP engines. ZIIPONCP time is 00:03:19 which means an additional 2.6% of work from the MIM address space was zIIP-eligible but was run on CP engines due to the lack of zIIP processor availability.

For SYSC, TOTALCPU time is 00:58:51, and ZIIP time is 00:56:24, which means 95.8% of MIM address space work has been offloaded to zIIP engines. ZIIPONCP time is 00:01:20 which means an additional 2.2% of work from the MIM address space was zIIP-eligible but was run on CP engines due to the lack of zIIP processor availability.


SYSA:

F MIM,DISPLAY CPUTIME=ALLMIM0660I CPU TIME: LAST RESTART AT 21:56:09 ON 2013.167 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMEQBLK 0.000008 0.000008 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000010 0.000010 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000010 0.000010 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000005 0.000005 0.000000 0.0 0.000000 0.0 MIMDRDRV 00:58:27 24.560154 50.261896 1.4 00:57:12 97.8 MIMDRHLP 0.000010 0.000010 0.000000 0.0 0.000000 0.0 MIMCMCNF 9.283093 9.283093 0.000000 0.0 0.000000 0.0 MIMCMREQ 4.556026 4.556026 0.000000 0.0 0.000000 0.0 MIMEQBST 0.000007 0.000007 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000005 0.000005 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000003 0.000003 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000002 0.000002 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000006 0.000006 0.000000 0.0 0.000000 0.0 MIMDRSTM 1.388640 1.388640 0.000000 0.0 0.000000 0.0 MIMDRACT 0.000007 0.000007 0.000000 0.0 0.000000 0.0 MIMDRMON 1.196736 1.196736 0.000000 0.0 0.000000 0.0 MIMZPXMP 00:20:24 0.000169 0.323856 0.0 00:20:23 99.9 MIMDRTRC 0.000362 0.000361 0.000000 0.0 0.000001 0.2 MIMCTPOL 0.267117 0.267117 0.000000 0.0 0.000000 0.0 MIMDRVFD 01:07:29 0.003030 0.988006 0.0 01:07:28 99.9 MIMDRCKP 0.010246 0.010246 0.000000 0.0 0.000000 0.0 MIMDRCFM 0.018199 0.018199 0.000000 0.0 0.000000 0.0 MIMDRCMD 25.094704 25.094704 0.000000 0.0 0.000000 0.0 MIMDRWTO 2.641808 2.641808 0.000000 0.0 0.000000 0.0

MIMDRLOG 0.205032 0.064629 0.014776 7.2 0.125627 61.2


13-Jul-2016 41/86

MIMDRLOG 0.205032 0.064629 0.014776 7.2 0.125627 61.2 SRB 3.380433 -------- ---------- ---------- ---------- ---- ---------- ----- Total 02:27:09 00:01:09 51.588534 0.5 02:25:05 98.5Normalized 03:51:01 00:01:09 51.588534 0.4 03:48:57 99.1

SYSB:

F MIM,DISPLAY CPUTIME=ALLMIM0660I CPU TIME: LAST RESTART AT 22:20:57 ON 2013.167 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMEQBLK 0.000006 0.000006 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000007 0.000007 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000004 0.000004 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000006 0.000006 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000009 0.000009 0.000000 0.0 0.000000 0.0 MIMDRDRV 01:22:07 00:02:04 00:02:46 3.3 01:17:16 94.0 MIMDRHLP 0.000012 0.000012 0.000000 0.0 0.000000 0.0 MIMCMCNF 43.196099 43.196099 0.000000 0.0 0.000000 0.0 MIMCMREQ 1.947101 1.947101 0.000000 0.0 0.000000 0.0 MIMEQBST 0.000008 0.000008 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000004 0.000004 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000002 0.000002 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000009 0.000009 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000009 0.000009 0.000000 0.0 0.000000 0.0 MIMDRSTM 1.586429 1.586429 0.000000 0.0 0.000000 0.0 MIMDRACT 0.000007 0.000007 0.000000 0.0 0.000000 0.0 MIMDRMON 4.593825 4.593825 0.000000 0.0 0.000000 0.0 MIMZPXMP 00:09:59 0.000230 9.249849 1.5 00:09:49 98.4 MIMDRTRC 0.001290 0.001289 0.000000 0.0 0.000001 0.0 MIMCTPOL 0.205120 0.205120 0.000000 0.0 0.000000 0.0 MIMDRVFD 00:31:52 0.000438 23.551103 1.2 00:31:29 98.7 MIMDRCKP 0.083769 0.083769 0.000000 0.0 0.000000 0.0 MIMDRCFM 0.018440 0.018440 0.000000 0.0 0.000000 0.0 MIMDRCMD 00:01:09 00:01:09 0.000000 0.0 0.000000 0.0 MIMDRWTO 12.633616 12.633616 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.273430 0.064914 0.021718 7.9 0.186798 68.3 SRB 6.653445 -------- ---------- ---------- ---------- ---- ---------- ----- Total 02:06:19 00:04:18 00:03:19 2.6 01:58:35 93.8Normalized 03:14:52 00:04:18 00:03:19 1.7 03:07:08 96.3

SYSC:

F MIM,DISPLAY CPUTIME=ALLMIM0660I CPU TIME: LAST RESTART AT 15:32:19 ON 2013.167 WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMEQBLK 0.000009 0.000009 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000008 0.000008 0.000000 0.0 0.000000 0.0 MIMDRDRV 00:38:22 28.290437 00:01:06 2.9 00:36:47 95.8 MIMDRHLP 0.000010 0.000010 0.000000 0.0 0.000000 0.0 MIMCMCNF 7.217557 7.217557 0.000000 0.0 0.000000 0.0 MIMCMREQ 0.166829 0.166829 0.000000 0.0 0.000000 0.0 MIMEQBST 0.000008 0.000008 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000004 0.000004 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000008 0.000008 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000003 0.000003 0.000000 0.0 0.000000 0.0 MIMEQBLK 0.000005 0.000005 0.000000 0.0 0.000000 0.0 MIMDRSTM 0.634003 0.634003 0.000000 0.0 0.000000 0.0 MIMDRACT 0.000010 0.000010 0.000000 0.0 0.000000 0.0 MIMDRMON 0.941714 0.941714 0.000000 0.0 0.000000 0.0 MIMZPXMP 00:04:58 0.000282 3.868266 1.2 00:04:54 98.7 MIMDRTRC 0.001098 0.001097 0.000000 0.0 0.000001 0.0 MIMCTPOL 0.082424 0.082424 0.000000 0.0 0.000000 0.0 MIMDRVFD 00:14:51 0.000571 9.238688 1.0 00:14:42 98.9 MIMDRCKP 0.011958 0.011958 0.000000 0.0 0.000000 0.0

MIMDRCFM 0.010271 0.010271 0.000000 0.0 0.000000 0.0


13-Jul-2016 42/86

MIMDRCFM 0.010271 0.010271 0.000000 0.0 0.000000 0.0 MIMDRCMD 22.345350 22.345350 0.000000 0.0 0.000000 0.0 MIMDRWTO 3.624448 3.624448 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.103861 0.025749 0.006174 5.9 0.071938 69.2 SRB 3.603698 -------- ---------- ---------- ---------- ---- ---------- ----- Total 00:58:51 00:01:03 00:01:20 2.2 00:56:24 95.8Normalized 01:31:27 00:01:03 00:01:20 1.4 00:56:24 97.3

Results of zIIP Enablement Feature OffloadThe zIIP offload scenario that is illustrated in this document involves a 3-system MIMplex running COMMUNICATION=CTCDASD with SYSA as the VCF MASTER system. All three systems were LPARs running on a single zEC12 CPU partitioned into many LPARs. The CA MII GDIF and ECMF facilities were managing ENQ requests at rate of about 230 ENQs per second on each LPAR for six days.

The results that were achieved in this MIMplex can differ from the results that you may achieve in your MIMplex. The following factors can impact the amount of CA MIM for z/OS address space work that gets offloaded to zIIP engines:

CPU model

WLM settings

CPU partitioning

Speed of CP and zIIP processors

Number of CP and zIIP processors

CA MIM for z/OS communication method used

CA MIM for z/OS facilities and features activated

Number of systems in the MIMplex, and so on

CA MIM for z/OS engineering conducted a number of contrived tests to determine which MIM communication methods produced the best zIIP offload percentages. These tests were conducted in a 2-system MIMplex on lightly loaded z/OS guest systems using various MIM cross-systems communication methods.

During each contrived test:

1,000,000 ENQ/DEQ pairs were created using an in-house ENQ generator utility.

CA MIM for z/OS DISPLAY CPUTIME=DETAIL commands measured the amount of consumed CPU time.


MIM DASD control files yielded the poorest zIIP offload results at 37.5 percent.

MIM CTC communication methods yielded the best zIIP offload results at nearly 99.8 percent.


13-Jul-2016 43/86

MIM CTC communication methods yielded the best zIIP offload results at nearly 99.8 percent.

The other MIM communication methods achieved results in between the two results above.

The variation in the results is due to the underlying IBM services used by each MIM communication method and their ability to run in SRB mode.

Given the observed results, some sites may want to consider switching from their current MIM communication method to one that yields a higher zIIP offload rate. This switch could help reduce mainframe operating costs or defer a CPU upgrade to a higher capacity model.

Example DISPLAY CPUTIME=DETAIL Commands:

COMMUNICATION=DASDONLY using a shared DASD Control File.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 00:03:37 00:02:22 0.023328 0.0 00:01:15 34.4 MIMZPXMP 13.716737 0.000000 0.001420 0.0 13.715317 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.007428 0.003383 0.000334 4.4 0.003711 49.9 SRB 4.857467 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 00:03:56 00:02:22 0.025082 0.0 00:01:28 37.5

COMMUNICATION=DASDONLY using a shared Coupling Facility List Structure Control File.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 00:01:53 0.004169 3.857184 3.4 00:01:49 96.5 MIMZPXMP 5.549710 0.000000 0.003706 0.0 5.546004 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.006835 0.002527 0.000590 8.6 0.003718 54.3 SRB 00:01:15 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 00:03:14 0.247066 3.861480 1.9 00:01:54 59.0

COMMUNICATION=XCF on the Virtual Control File Master System.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 34.358393 0.002453 0.008019 0.0 34.347921 99.9 MIMZPXMP 4.093996 0.000000 0.000047 0.0 4.093949 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 5.646007 0.000094 0.000070 0.0 5.645843 99.9 MIMDRLOG 0.004871 0.001511 0.000436 8.9 0.002924 60.0 SRB 0.001679 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 44.313972 0.213084 0.008572 0.0 44.090637 99.4

COMMUNICATION=XCF on the Virtual Control File Client System.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 40.076193 0.000445 0.041263 0.1 40.034485 99.8 MIMZPXMP 4.735944 0.000000 0.000230 0.0 4.735714 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 53.923254 0.000000 0.257082 0.4 53.666172 99.5 MIMDRLOG 0.009122 0.003778 0.000628 6.8 0.004716 51.6 SRB 22.852900 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 00:02:01 0.301478 0.299203 0.2 00:01:38 80.7


13-Jul-2016 44/86

COMMUNICATION=CTCONLY or CTCDASD on the Virtual Control File Master System.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 36.945028 0.000353 0.006654 0.0 36.938021 99.9 MIMZPXMP 4.976018 0.000000 0.000088 0.0 4.975930 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 5.584680 0.000000 0.000094 0.0 5.584586 99.9 MIMDRLOG 0.001883 0.000650 0.000211 11.2 0.001022 54.2 SRB 0.002142 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 47.574052 0.065304 0.007047 0.0 47.499559 99.8

COMMUNICATION=CTCONLY or CTCDASD on the Virtual Control File Client System.

MIM0660I CPU TIME: WORKUNIT TOTALCPU TASKCP ZIIPONCP PCT ZIIP PCT -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 41.856751 0.000421 0.010323 0.0 41.846007 99.9 MIMZPXMP 10.541308 0.000000 0.000073 0.0 10.541235 99.9 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 23.537456 0.000000 0.000167 0.0 23.537289 99.9 MIMDRLOG 0.006833 0.003163 0.000409 5.9 0.003261 47.7 SRB 0.008252 -------- ---------- ---------- ---------- ---- ---------- ----- TOTAL 00:01:16 0.168805 0.010972 0.0 00:01:15 99.7

SummaryzIIP specialty engines allow IBM mainframe customers to reduce their total cost of ownership or defer upgrades to higher capacity CPU models. The CA MIM for z/OS zIIP Enablement feature has yielded positive results for sites choosing to activate it. Some environments see significant cost savings by offloading CA MIM for z/OS address space work from general processors to the more cost efficient zIIP processors. Additionally, some environments find that offloading CA MIM for z/OS address space work to zIIP engines improves overall system throughput. These performance benefits are due to the feature:

Making general processor engine cycles, formerly used by CA MIM for z/OS, available to other workloads.

CA MIM for z/OS work completing faster on the 100 percent, full capacity zIIP engines.

Contact CA MIM for z/OS Technical Support with any questions about the operational characteristics of the CA MIM for z/OS product.


13-Jul-2016 45/86

CA MIM HYPERSTAR Feature - A Case StudyThis scenario describes a case study to the administrator using HYPERSTAR.

How To Read This Case Study (see page 45)About The HYPERSTAR Feature (see page 47)How HYPERSTAR Works (see page 47)Activating HYPERSTAR (see page 48)Test Results (see page 49)Average Cycle Times On The Four Client Systems (see page 57)

How To Read This Case StudyThe following instructions show you how to read this case study.


13-Jul-2016 46/86

1.

2.

3.

4.

5.

HYPERSTAR

Study the HYPERSTAR feature by reading and reviewing each of the following steps.

About The HYPERSTAR Feature (see page 47)

How HYPERSTAR Works (see page 47)

Activating HYPERSTAR (see page 48)

CA Test Environment

Test Results (see page 49)


13-Jul-2016 47/86

6. Average Cycle Times On The Four Client Systems (see page 57)

About The HYPERSTAR FeatureFor a MIMplex that has three or more active systems, performance of Virtual Control Files (VCF) may be improved by activating the HYPERSTAR feature. When this feature is active, a client system does not always transfer the VCF back to the master system after completing its processing. Instead, CA MIM for z/OS examines a ‘look ahead’ list of systems waiting for the VCF. When the list is not empty, the client transfers the VCF directly to that system instead of the master. These transfers assume that the client has a CTC or XCF path to the next system in the list. When the ‘look ahead’ list becomes empty, or when there is no path to the next system in the list, the client transfers the VCF back to the master.

The primary effect of HYPERSTAR is to reduce the amount of VCF I/O performed by the VCF master system, in many cases by more than 50 percent, which reduces the amount of CPU time that is used on the master system, and can potentially reduce the control file access times for all systems in MIMplex.

CA performed a benchmark on their test systems to study the impact of the HYPERSTAR feature.

How HYPERSTAR WorksThe following diagrams illustrate how CA MIM for z/OS transfers the Virtual Control File (VCF) with and without HYPERSTAR. This example assumes full connectivity between all systems. This connectivity is referred to as a symmetrical configuration.

GK03 is the VCFMASTER.

These diagrams assume that the VCF reserve requests from the three client systems are all queued simultaneously on the VCFMASTER/GK03.

Without HYPERSTARGK03 GK13 GK23 GK49

Request VCF

Request VCF

Request VCF

Send VCF to GK13

Send VCF back to GK03

Send VCF to GK23


Send VCF to GK49



13-Jul-2016 48/86

1.

2.

Six total I/O operations for the VCF transfers on GK03, that is three sends and three receives.

With HYPERSTARGK03 GK13 GK23 GK49

Request VCF

Request VCF

Request VCF

Send VCF to GK13

Send VCF to GK23

Send VCF to GK49


Two total I/O operations for the VCF Transfers on GK03, that is one send and one receive.

The total number of VCF transfers on GK03 was reduced from 6 to 2, or a 66.6 percent reduction. This reduction translates to reduced CPU usage on the VCFMASTER and improved ENQ response time.

The HYPERSTAR feature can be used with both XCF and CTC communications. In some cases one system may use a CTC path to communicate and others may use XCF signaling. The usage depends on the configuration and the value of SET MIM VCFPREFERENCE.

When any of the statements are true, HYPERSTAR has no effect:

A DASD control file or list structure is being used

The MIMplex has fewer than three systems.

Activating HYPERSTARThe following example shows how to activate the HYPERSTAR feature:

Determine if HYPERSTAR is inactive.

F MIM,DISPLAY FEATUREMIM0067I Command DISPLAY 061MIM0670I Features Display: Facility Feature Status MIM HYPERSTAR Inactive

HYPERSTAR is inactive.

Determine if any systems are in a FREED status:

F MIM,DISPLAY SYSTEMSMIM0067I Command DISPLAY 064MIM0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 LOCAL MASTER zOS 2014.129 08:47:48.37 02 13 GK13 EXTERNAL FREED


13-Jul-2016 49/86

2.

3.

02 13 GK13 EXTERNAL FREED 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL FREED 05 62 GK62 EXTERNAL FREED

The ACTIVATE FEATURE command works similar to ACTIVATE COMPATLEVEL. If there are systems in a FREED status, append the FORCE operand to the command.

Important! Read the CA MIM Programming (https://docops.ca.com/display/CMFZS125 /CA+MIM+Programming) and the Statement and Command Reference (https://docops.

for detailed ca.com/display/CMFZS125/Statement+and+Command+Reference)information about the ACTIVATE FEATURE command.

F MIM,ACTIVATE FEATURE=(HYPERSTAR,FORCE)MIM0067I Command ACTIVATE 068MIM0684I ACTIVATE FEATURE=HYPERSTAR has been scheduledMIM0619I ACTIVATE FEATURE=HYPERSTAR acceptedMIM0620I ACTIVATE FEATURE=HYPERSTAR in progressMIM0621I ACTIVATE FEATURE=HYPERSTAR complete

Confirm that the feature is now active:

F MIM,DISPLAY FEATUREMIM0067I Command DISPLAY 075MIM0670I Features Display: Facility Feature Status MIM HYPERSTAR Active

After activating HYPERSTAR, you may start seeing activity on the paths connecting two client systems. Without HYPERSTAR, you would typically only see activity on paths between the client and the VCFMASTER.

Test ResultsSee for details from this study. A Statistical Details (see page 50) General Summary (see page 49)follows:

General SummaryThe AVERAGE CYCLE time on the VCFMASTER was reduced from 4.3 ms to 3.6 ms.

The ENQ SERVICE time on the VCFMASTER was reduced from 4.2 ms to 3.7 ms.

The CPU usage of the MIMDRVFD task on the VCFMASTER was reduced from 9.1 CPU seconds to 6.3 CPU seconds.

The AVERAGE CYCLE time on the client systems was reduced from 5.6 ms to 4.7 ms.









13-Jul-2016 50/86

The number of “Saved I/Os” on the VCFMASTER averaged 400,000 which represents a 60 percent total reduction. The “Saved I/Os” represent I/O operations that would have occurred if HYPERSTAR were *not* active.

The average elapsed run time of the ENQ intensive batch job decreased from 8.1 minutes to 7.0 minutes.

Important! The performance gain depends on how many systems are queued for the VCF simultaneously. MIMplexes with more systems may get a larger benefit. MIMPLEXes with busier systems may get a larger benefit. Measure the benefit at your site by reviewing the DISPLAY HYPERSTAR command output and reviewing the number of saved I/O operations on the VCFMASTER system.

Statistical DetailsContents

Test 1 (see page 50)Test 2 (see page 52)Test 3 (see page 55)

The following tests show detail test results.

Test 1DISPLAY IO on GK03/VCFMASTER

WITHOUT HYPERSTAR

LAST RESET AT 11:15:10 ON 2014.126 COUNT: CYC= 102,140 BLOCKS READ= 118,435 WRITTEN= 108,490 XACT READ= 6.110M PROCESSED= 1.849M WRITTEN= 506,806 AVG: CYC= 0.004886 BLOCKS READ= 1.159 WRITTEN= 1.062 XACT READ= 59.827 PROCESSED= 18.111 WRITTEN= 4.961 RATE: CYC= 188.798 BLOCKS READ= 218.918 WRITTEN= 200.536 XACT READ= 1295.393 PROCESSED= 3419.336 WRITTEN= 936.794

WITH HYPERSTAR


DISPLAY SERVICE on GK03/VCFMASTER

WITHOUT HYPERSTAR


13-Jul-2016 51/86

GPK1021I GK03 GDIF SERVICE Display: Requests Time/Request Rate/Second Rate/Cycle Since 253,375 0.004528 388.016 2.487 11:12:46 2014.126 253,375 0.004528 491.036 2.487 11:15:10 2014.126

WITH HYPERSTAR

GPK0067I Command DISPLAY 249GPK1021I GK03 GDIF SERVICE Display: Requests Time/Request Rate/Second Rate/Cycle Since 216,515 0.004195 435.643 2.125 11:28:13 2014.126 216,515 0.004195 489.852 2.125 11:29:11 2014.126

DISPLAY CPU=DETAIL on GK03/VCFMASTER

WITHOUT HYPERSTAR

GPK0660I CPU Time:Last RESET at 11:15:10 on 2014.126 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 11.313037 11.313037 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000137 0.000137 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 9.633515 9.633515 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.000459 0.000459 0.000000 0.0 0.000000 0.0 SRB 2.002170 -------- ---------- ---------- ---------- ---- ---------- ----- Total 22.983071 20.980901 0.000000 0.0 0.000000 0.0Normalized 22.983071 20.980901 0.000000 0.0 0.000000 0.0

WITH HYPERSTAR

Last RESET at 12:00:05 on 2014.126 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 9.320014 9.320014 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000101 0.000101 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 6.176045 6.176045 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.000792 0.000792 0.000000 0.0 0.000000 0.0 SRB 0.995038 -------- ---------- ---------- ---------- ---- ---------- ----- Total 16.536231 15.541193 0.000000 0.0 0.000000 0.0Normalized 16.536231 15.541193 0.000000 0.0 0.000000 0.0

DISPLAY PATH on GK03/VCFMASTER

WITHOUT HYPERSTAR

LAST RESET AT 11:15:10 ON 2014.126 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 XCF GK13 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0.004618 102,140 184A GK62 USABLE IDLE 80,705 80,712 0 180A GK62 USABLE IDLE 0 0 0 0B4A GK62 USABLE IDLE 0 0 0 185A GK49 USABLE IDLE 80,008 80,031 0 181A GK49 USABLE IDLE 0 0 0 0B5A GK49 USABLE IDLE 0 0 0 186A GK23 USABLE IDLE 80,300 80,335 0 182A GK23 USABLE IDLE 0 0 0 0B6A GK23 USABLE IDLE 0 0 0 183A GK13 USABLE IDLE 79,926 79,931 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0


13-Jul-2016 52/86

WITH HYPERSTAR


DISPLAY HYPERSTAR on GK03/VCFMASTER

WITHOUT HYPERSTAR

GPK0067I Command DISPLAY 652GPK0951I HYPERSTAR Statistics for VCF Master System:HYPERSTAR is not activeLast RESET at 11:15:10 on 2014.126 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 319,069 319,067 641,949 0 0.00 1.00

WITH HYPERSTAR

GPK0951I HYPERSTAR Statistics for VCF Master System:Last RESET at 11:29:11 on 2014.126 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 326,705 129,498 264,914 403,288 60.36 2.52

ELAPSED WALL CLOCK TIME for sample ENQ job

WITHOUT HYPERSTAR

11.15.12 job start time11.24.11 job end time8M 59S elapsed job run time

WITH HYPERSTAR


///


WITHOUT HYPERSTAR

LAST RESET AT 11:50:53 ON 2014.126 COUNT: CYC= 102,184 BLOCKS READ= 124,813 WRITTEN= 109,794 XACT READ= 7.195M PROCESSED= 1.720M WRITTEN= 472,834 AVG: CYC= 0.003942 BLOCKS READ= 1.221 WRITTEN= 1.074 XACT READ= 70.418 PROCESSED= 16.832 WRITTEN= 4.627 RATE: CYC= 234.905 BLOCKS READ= 286.926 WRITTEN= 252.400


13-Jul-2016 53/86

RATE: CYC= 234.905 BLOCKS READ= 286.926 WRITTEN= 252.400 XACT READ= 6541.772 PROCESSED= 3954.156 WRITTEN= 1086.974

WITH HYPERSTAR



WITHOUT HYPERSTAR


WITH HYPERSTAR



WITHOUT HYPERSTAR

RESET at 11:50:53 on 2014.126 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 10.169810 10.169810 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000127 0.000127 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 8.823135 8.823135 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.001597 0.001597 0.000000 0.0 0.000000 0.0 SRB 1.642828 -------- ---------- ---------- ---------- ---- ---------- ----- Total 20.714091 19.071263 0.000000 0.0 0.000000 0.0 Normalized 20.714091 19.071263 0.000000 0.0 0.000000 0.0

WITH HYPERSTAR

RESET at 12:00:05 on 2014.126 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 9.320014 9.320014 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000101 0.000101 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 6.176045 6.176045 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.000792 0.000792 0.000000 0.0 0.000000 0.0 SRB 0.995038 -------- ---------- ---------- ---------- ---- ---------- ----- Total 16.536231 15.541193 0.000000 0.0 0.000000 0.0 Normalized 16.536231 15.541193 0.000000 0.0 0.000000 0.0


WITHOUT HYPERSTAR


13-Jul-2016 54/86

LAST RESET AT 11:50:53 ON 2014.126 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK13 USABLE 0 0 0 XCF GK62 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0.003734 102,184 184A GK62 USABLE BUSY 80,444 80,465 0 WRITVCF 180A GK62 USABLE IDLE 0 0 0 0B4A GK62 USABLE IDLE 0 0 0 185A GK49 USABLE IDLE 80,082 80,090 0 181A GK49 USABLE IDLE 0 0 0 0B5A GK49 USABLE IDLE 0 0 0 186A GK23 USABLE IDLE 79,883 79,948 0 182A GK23 USABLE IDLE 0 0 0 0B6A GK23 USABLE IDLE 0 0 0 183A GK13 USABLE IDLE 79,178 79,186 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0WITH HYPERSTARLAST RESET AT 12:00:05 ON 2014.126 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 XCF GK13 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0.002817 102,247 184A GK62 USABLE BUSY 36,771 34,986 0 WRITVCF 180A GK62 USABLE IDLE 0 0 0 0B4A GK62 USABLE IDLE 0 0 0 185A GK49 USABLE IDLE 32,364 34,452 0 181A GK49 USABLE IDLE 0 0 0 0B5A GK49 USABLE IDLE 0 0 0 186A GK23 USABLE IDLE 36,699 34,556 0 182A GK23 USABLE IDLE 0 0 0 0B6A GK23 USABLE IDLE 0 0 0 183A GK13 USABLE IDLE 31,342 33,227 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0


WITHOUT HYPERSTAR

GPK0067I Command DISPLAY 681GPK0951I HYPERSTAR Statistics for VCF Master System:HYPERSTAR is not activeLast RESET at 11:50:53 on 2014.126 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 316,673 316,671 639,277 4 0.00 1.00

WITH HYPERSTAR

GPK0951I HYPERSTAR Statistics for VCF Master System:Last RESET at 12:00:05 on 2014.126 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 334,086 133,480 274,397 412,245 60.04 2.50


WITHOUT HYPERSTAR


WITH HYPERSTAR


13-Jul-2016 55/86

12.00.09 ob start time12.06.13 job end time6M 4S elapsed job run time

///


WITHOUT HYPERSTAR


WITH HYPERSTAR



WITHOUT HYPERSTAR


WITH HYPERSTAR



WITHOUT HYPERSTAR

Last RESET at 09:51:26 on 2014.127 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 10.352334 10.352334 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000115 0.000115 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 8.955743 8.955743 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.000643 0.000643 0.000000 0.0 0.000000 0.0 SRB 1.665057 -------- ---------- ---------- ---------- ---- ---------- ----- Total 21.013705 19.348648 0.000000 0.0 0.000000 0.0 Normalized 21.013705 19.348648 0.000000 0.0 0.000000 0.0

WITH HYPERSTAR


13-Jul-2016 56/86

WITH HYPERSTAR

Last RESET at 10:02:26 on 2014.127 WorkUnit TotalCPU TaskCP zIIPonCP Pct zIIP Pct -------- ---------- ---------- ---------- ---- ---------- ----- MIMDRDRV 10.037426 10.037426 0.000000 0.0 0.000000 0.0 MIMZPXMP 0.000116 0.000116 0.000000 0.0 0.000000 0.0 MIMDRTRC 0.000000 0.000000 0.000000 0.0 0.000000 0.0 MIMDRVFD 6.663939 6.663939 0.000000 0.0 0.000000 0.0 MIMDRLOG 0.001196 0.001196 0.000000 0.0 0.000000 0.0 SRB 1.011495 -------- ---------- ---------- ---------- ---- ---------- ----- Total 17.775529 16.764034 0.000000 0.0 0.000000 0.0Normalized 17.775529 16.764034 0.000000 0.0 0.000000 0.0


WITHOUT HYPERSTAR

LAST RESET AT 09:51:26 ON 2014.127 DEV SYSTEM STATE IO-STAT READS WRITES ERR TIME/RSV CYCLES XCF GK62 USABLE 0 0 0 XCF GK13 USABLE 0 0 0 N/A GK03 LOCAL 0 0 0.004076 102,171 184A GK62 USABLE IDLE 80,645 80,688 0 180A GK62 USABLE IDLE 0 0 0 0B4A GK62 USABLE IDLE 0 0 0 185A GK49 USABLE IDLE 80,315 80,339 0 181A GK49 USABLE IDLE 0 0 0 0B5A GK49 USABLE IDLE 0 0 0 186A GK23 USABLE BUSY 80,948 80,977 0 WRITVCF 182A GK23 USABLE IDLE 0 0 0 0B6A GK23 USABLE IDLE 0 0 0 183A GK13 USABLE IDLE 79,610 79,615 0 154A GK13 USABLE IDLE 0 0 0 0B3A GK13 USABLE IDLE 0 0 0

WITH HYPERSTAR



WITHOUT HYPERSTAR

GPK0951I HYPERSTAR Statistics for VCF Master System: HYPERSTAR is not active Last RESET at 09:51:26 on 2014.127 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 319,064 319,061 643,148 6 0.00 1.00

WITH HYPERSTAR


13-Jul-2016 57/86

WITH HYPERSTAR

Last RESET at 10:02:26 on 2014.127 #Reserve #Xfer #I/O #Saved %Saved AvgDepth 317,171 126,101 257,309 389,782 60.24 2.51


WITHOUT HYPERSTAR

09.51.26 job start time09.59.20 job end tmie7M 54S elapsed job run time

WITH HYPERSTAR


///

Average Cycle Times On The Four Client SystemsThis table shows the average cycle times on the four client systems that are used in this case study.

Without Hyperstar With Hyperstar

Test 1

GK13 AVG: CYC= 0.006395 AVG: CYC= 0.005377

GK23 AVG: CYC= 0.006347 AVG: CYC= 0.005210

GK49 AVG: CYC= 0.006347 AVG: CYC= 0.005180

GK62 AVG: CYC= 0.006391 AVG: CYC= 0.005154

Test 2

GK13 AVG: CYC= 0.005203 AVG: CYC= 0.004230

GK23 AVG: CYC= 0.005158 AVG: CYC= 0.004033

GK49 AVG: CYC= 0.005087 AVG: CYC= 0.003981

GK62 AVG: CYC= 0.005131 AVG: CYC= 0.004001

Test 3

GK13 AVG: CYC= 0.005633 AVG: CYC= 0.005230

GK23 AVG: CYC= 0.005535 AVG: CYC= 0.004994

GK49 AVG: CYC= 0.005535 AVG: CYC= 0.004984

GK62 AVG: CYC= 0.005581 AVG: CYC= 0.004990


13-Jul-2016 58/86

1.

2.

3.

How to Set Up Restart ManagerAs a System Administrator, you want to set up Restart Manager in your environment. Restart Manager lets you restart CA MIM automatically after a failure or to perform a planned restart on a single system. When used with CA MII, Restart Manager maintains the integrity of global resources throughout the restart process. Local ENQ/DEQ activity is suspended for managed QNAMEs from the instant a failure is detected or a LOADLIB REFRESHFROM command is issued. Activity remains suspended until the restart completes and normal global operation resumes.

Restart Manager also lets you activate maintenance dynamically and safely. You can use an alternate load library which provides automatic fall back to the previous load library. You can also commit the alternate load library to become the current library. This dynamic refresh helps you reduce system downtime.

Understanding How Restart Manager Works (see page 59).This overview describes the functionality and how it affects your environment.

Review the Restart Manager Prerequisites (see page 60).You review the setup required in your environment. You can also view these other

.considerations (https://docops.ca.com/display/CMFZS10/Restart+Manager+Usage)

https://docops.ca.com/display/CMFZS10/Restart+Manager+Usage

https://docops.ca.com/display/CMFZS10/Restart+Manager+Usage


13-Jul-2016 59/86

3.

4.

1.

Review the Restart Manager Set Option and Commands (see page 60).You review these commands to understand how they affect your environment.

Use the LOADLIB Command (see page 61).The LOADLIB command initiates a controlled restart of many tasks in the CA MIM address space.

Understanding How Restart Manager WorksThe CA MIM address space contains the following additional Task Control Blocks (TCBs):

MIMDRRM

Attaches and monitors the MIMDRDRV task

Reattaches MIMDRDRV when a failure occurs

Recognizes recursion and initialization problems

MIMRMJC

Manages the connection to JES

MIMRMBEQ

Acts as a placeholder for the GDIFs ‘blocking’ ENQs

GRS Exit (MIMEQXNQ)

Suspends access to global resources

The following process describes what occurs during a restart:

The primary ESTAE exit analyzes the completion code.


13-Jul-2016 60/86

2.

3.

4.

5.

6.

1.

a.

b.

2.

CA MIM activates the MIMEQXNQ exit.The Requesters of global resources are suspended.

The MIMDRDRV task and all subtasks terminate.

MIMDRRM analyzes the completion code, determines which LOADLIB to use, and attaches MIMDRBGN/MIMDRDRV.

MIMDRDRV reads MIMPARMS, attaches subtasks, and so on.

During synchronization, the suspended requesters are awakened, and CA MIM deactivates the MIMEQXNQ exit.

Review the Restart Manager PrerequisitesRestart Manager lets you restart CA MIM automatically after a failure, or to perform a planned restart on a single system. Dynamic Refresh lets you dynamically switch to an alternate load library on demand.

For example, you apply maintenance to an alternate library, and you want to load this alternate library without causing any CA MIM outages. In this example, your CA MIM command prefix is an "at" sign (@).

Follow these steps:

Verify the following requirements:

Started with SUB=MSTR.

Startup PROC has the following defined and are PDSs (not PDSEs):STEPLIB and MIMPARMSMIMMSGS DD (optional)

Change the CA MIM PROC to PGM=MIMDRRM instead of PGM=MIMDRBGN.

Restart Manager will be active the next time that you start CA MIM (normally at IPL time).

Note: You can set PGM=MIMDRRM and restart CA MIM on one system at a time. For example, you want to roll out Restart Manager on one system at a time. You first enable Restart Manager on a test system to complete your tests before you move on to other systems.


13-Jul-2016 61/86

Review the Restart Manager Set Option and Commands

Restart Manager provides the following commands:

Set option to prevent crash and restart looping:

@ SETOPTION RestartLimit=n

Display the current values for RestartLimit and RestartCount:

@ DISPLAY OPTIONS

Display the current values for LOADLIB:

@ DISPLAY LOADLIB

Command to initiate a restart of CA MIM dynamically:

@ LOADLIB

Review the LOADLIB CommandThe command initiates a controlled restart of the CA MIM address space. LOADLIB REFRESHFROMDuring the restart, CA MIM suspends access to resources that it manages on the local system. CA MIM resumes this access only after the restart completes. To avoid disruption of production workload, use this command only during periods of (relatively) low system activity.

When you issue the command on the VCF master system, global operations LOADLIB REFRESHFROMare suspended on all systems until the master system completes the refresh and restart. During the restart, some messages normally associated with recovery from a failure in VCF communication are issued on one or more systems.

Note: This behavior is normal, and no operator intervention is required.

DISPLAY LOADLIBDisplays the current setting for the following Load Library details:

Original DSN, VOL, DDName

Current DSN, VOL, DDName

Primary DSN, VOL, DDName


13-Jul-2016 62/86


Alternate DSN, VOL, DDName

This command has the following format:

@ DISPLAY LOADLIB

LOADLIBLets you define your alternate Load Library and restart CA MIM.


@ LOADLIB {[SETALTERNATE=(dsname[,volser])] | [REFRESHFROM=[{PRIMARY | CURRENT | ALTERNATE}]] | [COMMIT]}

SETALTERNATE

(Optional) Defines the specified dsname and optional volser of the alternate load library. SETALTERNATE verifies the following information:

dataset is usable

dataset exists

dataset is PDS and not PDSE

dataset is APF authorized

dataset actually contains CA MIM programs

REFRESHFROM

(Optional) CA MIM begins using the specified load library dynamically. REFRESHFROM initiates a controlled restart of CA MIM. It uses a specified load library on only one system at a time.

Note: LOADLIB REFRESHFROM replaces the SHUTDOWN RESTART command.

PRIMARY – Refresh using the PRIMARY load library.

CURRENT – Refresh using the CURRENT load library.

ALTERNATE – Refresh using the ALTERNATE load library.

Default: CURRENT

COMMIT


13-Jul-2016 63/86

COMMIT

(Optional) Sets the PRIMARY load library as the CURRENT LOADLIB.


13-Jul-2016 64/86

1.

How to Use Restart Manager to Apply Maintenance Dynamically

As a System Administrator, you want to apply CA MIM maintenance safely without requiring a CA MIM outage or address space shutdown then restart. This maintenance includes PTFs and APARs that contains fixes and new features. You use the Restart Manager and its LOADLIB command capabilities to achieve this goal safely. We recommend that you first apply maintenance to a different library than the primary, running library that CA MIM uses.

For example, you set the alternate library to make the maintenance available for the actual CA MIM refresh. However, you want the alternate library to replace your primary library until you do notknow that the maintenance works properly to avoid any problems. If problems occur, CA MIM falls back to the prior, working library. You can populate the alternate library with the maintenance as a test, complete your verification tasks, and then you can roll out the maintenance to other systems in your CA MIM complex.

Note: We recommend this approach as a safety measure on your first CA MIM address space. After you validate that the maintenance or feature runs successfully without issues, you can roll out the maintenance PTF to other CA MIM address spaces. Then, you can commit the changes to your primary load library, and CA MIM will refresh from the new primary library going forward.

Review the LOADLIB Command (see page 65)


13-Jul-2016 65/86

1.

2.

3.

Review the LOADLIB Command (see page 65)The LOADLIB command initiates a controlled restart of many tasks in the CA MIM address space.

Populate the Alternate Library with the Maintenance PTF (see page 67).You completed the prerequisites and you applied the PTF or APAR.

Commit the Alternate Library as the Primary Library (see page 67).You determined that the current library is stable, and you want to set the library as a primary.

Review the LOADLIB CommandThe command initiates a controlled restart the CA MIM address space. LOADLIB REFRESHFROMDuring the restart, CA MIM suspends access to resources that it manages on the local system. CA MIM resumes this access only after the restart completes. To avoid disruption of production workload, use this command only during periods of low system activity, such as an overnight batch window on a weekend.

When you issue the command on the VCF master system, global operations LOADLIB REFRESHFROMare suspended on all systems until the master system completes the refresh and restart. During the restart, some messages normally associated with recovery from a failure in VCF communication are issued on one or more systems. In this example, your CA MIM command prefix is an ampersand (&).

Note: This behavior is normal, and no operator intervention is required.

DISPLAY LOADLIBDisplays the current setting for the following Load Library details:

Original DSN, VOL, DDName

Current DSN, VOL, DDName


Alternate DSN, VOL, DDName


@ DISPLAY LOADLIB


13-Jul-2016 66/86

LOADLIBLets you define your alternate Load Library and restart CA MIM.This command has the following format:

@ LOADLIB {[SETALTERNATE=(dsname[,volser])] | [REFRESHFROM=[{PRIMARY | CURRENT | ALTERNATE}]] | [COMMIT]}

SETALTERNATE

(Optional) Defines the specified dsname and optional volster of the alternate load library. SETALTERNATE verifies the following information:

dataset is useful

dataset exists

dataset is PDS and not PDSE

dataset is APF authorized

dataset actually contains CA MIM programs

REFRESHFROM

(Optional) CA MIM begins using the specified load library dynamically. REFRESHFROM initiates a controlled restart of CA MIM. It uses a specified load library on only one system at a time.

Note: LOADLIB REFRESHFROM replaces the SHUTDOWN RESTART command.

PRIMARY – Refresh using the PRIMARY load library.

CURRENT – Refresh using the CURRENT load library.

ALTERNATE – Refresh using the ALTERNATE load library.

Default: CURRENT

COMMIT

(Optional) Sets the PRIMARY load library as the CURRENT LOADLIB.


13-Jul-2016 67/86

1.

2.

3.

4.

5.

6.

Populate the Alternate Library with the Maintenance PTF

You already set up Restart Manager correctly, and you want to apply maintenance safely without causing a CA MIM outage or shutdown. For example, you want to activate a PTF dynamically on a test system (SYSA) first. A successful activation and verification lets you apply the maintenance on your other CA MIM systems after you validate the maintenance initially.

The dynamic refresh capability of Restart Manager lets you switch the load library without stopping and restarting the CA MIM address space or causing a system outage. If an unexpected abend occurs, CA MIM falls back to the previous load library. In this example, you assume that CA MIM on SYSA is running with a STEPLIB load library named MIM.LOAD.LIB.

Follow these steps:

Create an alternate load library named on volume VOL001.MIM.LOAD.ALT.LIB

Populate MIM.LOAD.ALT.LIB with a load library that contains the new maintenance PTF or APAR.

Execute the following command to set alternate load library as MIM.LOAD.ALT.LIB on VOL001:

@ LOADLIB SETALTERNATE=(MIM.LOAD.ALT.LIB,VOL001)

Execute the following command to verify that CA MIM knows the correct alternate library.:

@ DISPLAY LOADLIB

Wait for a low activity period on the system and execute the following command:

@ LOADLIB REFRESHFROM=ALTERNATE

CA MIM suspends serialization activity on the system for a short period while it refreshes to the alternate load library.

After CA MIM resynchronizes, execute the DISPLAY command again to verify that CURRENT library displays MIM.LOAD.ALT.LIB.

@ DISPLAY LOADLIB

CA MIM is now operating on the alternate load library.

Note: If CA MIM encounters errors while operating on the alternate load library, it reverts to the primary load library and clears all information regarding the alternate load library.


13-Jul-2016 68/86

1.

2.

3.

Commit the Alternate Library as the Primary Library

When you determine that MIM.LOAD.ALT.LIB is stable and safe, execute the LOADLIB COMMIT command to set MIM.LOAD.ALT.LIB as the PRIMARY library. After you commit the alternate library, any unplanned CA MIM restarts utilize MIM.LOAD.ALT.LIB.

Follow these steps:

Execute the following command to verify that the CURRENT is the load library that you want to commit:

@ DISPLAY LOADLIB

Execute the following command to accept the current in-use library as the primary load library for CA MIM:

@ LOADLIB COMMIT

Execute the following command to verify that the PRIMARY load library now matches the CURRENT load library:

@ DISPLAY LOADLIB

Notes:

The committed library is retained across IPLs or CA MIM shutdowns. If you want to notutilize this new library at initial startup, update the STEPLIB DD in the CA MIM JCL PROC.

You can apply most maintenance dynamically using this process. If exceptions arise, the ++HOLDATA portion of the APAR/PTF specifies that you cannot apply the maintenance dynamically.

You applied a maintenance update without restarting your CA MIM.


13-Jul-2016 69/86

How to Activate MIMQUERY in the MIMplex Dynamically

As a System Administrator, you want to implement the MIMQUERY feature in the MIMplex dynamically. By dynamically applying this maintenance for the MIMQUERY feature, you minimize system downtime by using Restart Manager and Dynamic Refresh to enable the MIMQUERY feature on all systems in the MIMplex. MIMQUERY lets you collect global information about the current state of resources in the MIMplex.

To enable the MIMQUERY feature in your environment dynamically, you must create a load library with all required maintenance applied and copy sample PROC and PARMLIB members to your run time libraries. Then, you can dynamically refresh systems in the MIMplex one system at a time to an alternate library until all systems are running on the alternative load library.

If the alternate load library is successful, you can commit this alternate load library as the primary in your MIMplex.

Note: You can apply the concepts in this detailed example to implement any PTF or APAR without restarting your MIMplex.

The following diagram shows how you activate MIMQUERY without restarting your MIMplex:


13-Jul-2016 70/86

1.

2.

3.

4.

Review the MIMQUERY Prerequisites (see page 71).These prerequisites help ensure that you set up your environment correctly for MIMQUERY to operate in your MIMplex.

Review the Environment Setup (see page 73).Visualize your environment before and after you install and activate MIMQUERY on all systems.

Verify that You Installed the Maintenance and Activate MIMQUERY Dynamically (see page 74).Verify that you installed the required maintenance successfully and that you activated MIMQUERY in your MIMplex by loading the alternate load library dynamically. You can commit this library for future resets.

Use MIMQUERY Request Resource Information (see page 76).Request global information for a resource, and all systems recognize and process the request successfully.


13-Jul-2016 71/86

Review the MIMQUERY PrerequisitesMIMQUERY provides an interface similar to the IBM GQSCAN or ISGQUERY macros. When you activate this feature, programs on any system in the MIIplex can use the MIMQUERY macro to obtain global ENQ resource information. Before you invoke the MIMQUERY macro, you must launch certain CA Common Services FMIDs and address spaces. For example, you must have a CA CCI address space active on all systems in the MIMPLEX. MIMQUERY uses CA CCI to communicate with other systems. Before you can use MIMQUERY, verify that you meet the requirements:

Verify that you configured CA CCI (Common Communications Interface) and that CCI is available between all systems in the MIMplex.

Verify that you applied all necessary PTFs on all systems.

Verify that CA MIM is running Release 12.1 or later with compat level 12.0 or higher

CA Common Services and CCS for z/OS Component Requirements

The following CCS for z/OS components are required for CA MIM when you use FEATURE=MIMQUERY:

CAW4E10 - Specifies the CAICCI component.CAICCI provides CA enterprise applications with a common communications software layer that insulates the applications from dealing with protocol specifics, error recovery, and system connection establishment.

Note: If you use FEATURE=MIMQUERY, you install CAICCI on all systems in the mustMIMplex and provide full connectivity on all systems. For example, each system have mustthe ability to send messages to any other system.

CEI0E10 - Specifies the CAMASTER component.CAMASTER is a non-cancelable started task that provides system services used by various CA products and CA Common Services, such as the CA Health Checker address space and the CA Common Address Space Shell.

Note: If you use FEATURE=MIMQUERY, or if you use the MIM health checks, then you mustinstall CAMASTER on all systems in the MIMplex.


13-Jul-2016 72/86

CAS9E10 - Specifies the CAIRIM component.Prepares your operating system environment for all CA applications and starts them. The common driver for a collection of dynamic initialization routines eliminates the need for user SVCs, SMF exits, subsystems, and other installation requirements commonly encountered when installing systems applications.Integral parts of CAIRIM are CAISSF, CA LMP, and CA zIIP Enablement Services.

Note: These modules must reside in LINKLIST or as part of the CA MIM STEPLIB concatenation.

CA MIM Setup RequirementsCopy .CBTDPROC(MIMQUERY) to any JCL PROCLIB that is automatically searched as part of z/OS rthlqSTART command processing, such as SYS2.PROCLIB.

Copy .CBTDPARM(MIMQUERY) and .CBTDPARM(MIMQMSGS) to the PARMLIB DD dataset rthlq rthlqthat is pointed to by your CA MIM startup PROC.

Required PTFsDeploy the following CA Common Service (CCS) PTFs before activating the MIMQUERY feature:

RO52583

RO53583

RO57236

RO37237

RO72853

RO77749

RO80830

RO82009


13-Jul-2016 73/86

Review the Environment SetupIn this example, you have three systems in your MIMplex that all use CTC communication. SYSB and SYSC are in a SYSPLEX, but SYSA (master) is not in the SYSPLEX. SYSA has the MIMQUERY feature installed first. You want all systems to use a common load library, and then you want to refresh those systems to use that library. The following diagram shows your environment before you complete the installation and activation on all systems:

The following diagram describes your environment after you load the alternate load library and activate MIMQUERY in your MIMplex.


13-Jul-2016 74/86

1.

2.

a.

b.

c.

d.

e.

After you activate MIMQUERY, an auxiliary address space is created on SYSA, SYSB, and SYSC for the MIMQUERY server. The servers communicate with the CCI.

Important! If any system cannot join due to a prerequisite problem, the activate feature command fails.

The auxiliary address spaces impact the operation of the MIMplex in any way. In the do notprevious diagram, the borders around each MIMQUERY address space and system indicate a separate LPAR.

After your MIMplex is synchronized, a job on any system can make a request for resource information.

JOBA on SYSA requests global information for Resource1.

MIMQUERY on SYSA accepts the request and validates that it is acceptable.

MIMQUERY then passes the request to itself, SYSB, and SYSC.

All three systems create a response based on their own knowledge of Resource1 and all systems send the response back to SYSA.

SYSA receives the responses, formulates an answer, and passes the answer to the requesting job.


13-Jul-2016 75/86

1.

2.

3.

4.

5.

6.

7.

8.

Verify that You Installed the Maintenance and Activate MIMQUERY Dynamically

As the System Administrator, you want to verify that SYSA can use and understand the MIMQUERY functionality. Then, you want to activate MIMQUERY dynamically. Ultimately, you want SYSA, SYSB, and SYSC to use a common load library that has MIMQUERY installed and then you can activate MIMQUERY on SYSA, SYSB, and SYSC.

Follow these steps:

Create an alternate load library named on volume VOL001.MIM.LOAD.ALT.LIB

Populate MIM.LOAD.ALT.LIB with a load library that has MIMQUERY installed.

If you have not already done so, perform the following copies;a. Copy .CBTDPROC(MIMQUERY) to any JCL PROCLIB that is automatically searched as rthlqpart of z/OS START command processing, such as SYS2.PROCLIB.b. Copy .CBTDPARM(MIMQUERY) and .CBTDPARM(MIMQMSGS) to the PARMLIB DD rthlq rthlqdataset that is pointed to by your CA MIM startup PROC.

Execute the following command on SYSA to set the alternate load library as MIM.LOAD.ALT.LIB on VOL001:

@ LOADLIB SETALTERNATE=(MIM.LOAD.ALT.LIB,VOL001)

Execute the following command on SYSA to load the ALT library:

@ LOADLIB REFRESHFROM=ALTERNATE

CA MIM goes through a restart process, but the address space shut down because does notCA MIM rebuilds the entire structure dynamically.

Execute the following command to verify that you installed the MIMQUERY feature in your load library:

@ DISPLAY FEATURE

As of CA MIM Release 12.1, the feature list displays both HYPERSTAR and MIMQUERY. If MIM.LOAD.ALT.LIB contains an installed MIMQUERY, the result displays that MIMQUERY is installed, but inactive.

DISPLAY FEATURE shows all installed features. If only HYPERSTAR appears as installed, then your load library does not have MIMQUERY installed.

Repeat Steps 3 through 5 on SYSB and SYSC.The alternate load library with MIMQUERY is installed on SYSA, SYSB, and SYSC. If you want, you can issue @ DISPLAY FEATURE again, and the MIMQUERY service appears as available but not active.

Execute the following command on any system to activate MIMQUERY on SYSA, SYSB, and SYSC:


13-Jul-2016 76/86

8.

9.

10.

11.

1.

2.

3.

4.

5.

@ ACTIVATE FEATURE(MIMQUERY)or@ ACTIVATE FEATURE=MIMQUERY

Execute the following command on any system to verify that MIMQUERY activated successfully:

@ DISPLAY FEATURE

The result indicates that MIMQUERY is installed and activated.

When you determine that the MIM.LOAD.ALT.LIB with MIMQUERY installed is stable and safe, execute the following command to set MIM.LOAD.ALT.LIB as the PRIMARY library.

@ LOADLIB COMMIT

Going forward, any unplanned CA MIM restarts utilize MIM.LOAD.ALT.LIB.

Update your STEPLIB statement in the CA MIM start up PROC to use the MIM.LOAD.ALT.LIB for future system IPLs.

Use MIMQUERY to Request Resource InformationAfter your MIMplex is synchronized, you can use MIMQUERY to request resource information from all systems as follows:

Execute the following command on SYSA:

@ DISPLAY CONFLICTS

MIMQUERY on SYSA accepts the request and validates that it is acceptable.

MIMQUERY passes the request to itself, SYSB, and SYSC.

All systems create responses that are based on their own knowledge of the resource, and SYSB and SYSC send their responses back to SYSA.

SYSA receives the responses, formulates an answer, and passes the answer to the requestor.

The following diagram shows the request lifecycle after you complete the MIMQUERY activation in your MIMplex:


13-Jul-2016 77/86

Notes:

The MIMQUERY member of the CAI.CBTDMAC data set contains a DSECT mapping for this standard parameter list. The macro contains detailed information about how to use the macro and incorporate it into your assembler program. For example, all calls must follow the format in the MIMQUERY macro.

The MIMQSAMP member of the CAI.CBTDSAMP data is a sample program that helps you make a sample MIMQUERY request


13-Jul-2016 78/86

How to Use the Restart Manager and Dynamic Refresh Enhancements

An additional DD statement should be added to the cataloged procedure used to start CA MIM:

//MIMRMFIL DD DISP=SHR,DSNAME=dsname[,UNIT=devtype,VOL=SER=volume]

The following is the complete syntax for all of the LOADLIB commands:

LOADLIB { SETALTERNATE=(dsname[,volser]) | REFRESHFROM[={INITIAL | PRIMARY | | ALTERNATE}] |CURRENT COMMIT | RESETALTERNATE | SAVE } [STARTFROM={INITIAL | | CURRENT | ALTERNATE} ]PRIMARY [SETAPF={ | NO}]YES

This command specifies which of the saved load libraries should be used at the initial startup. If the initial Restart Manager file is somehow corrupted or not present, CA MIM will always start from the initial. This command also always saves the current LOADLIB environment to the MIMRMFIL dataset.

To use these new features, you must be using MIM 12.5 or higher.

A full shutdown and restart of the MIM address space is required.

Additionally, your MIM address space must be configured to use the MIM Restart Manager and any load libraries must be defined as PDS.


13-Jul-2016 79/86

How to Change a System Name Without a Control File Format

As a system administrator, you want to change the system name on the DEFSYS statement.

Prior COMPATLEVEL values required a global outage and control file format to change the name of a system on the DEFSYS statement as long as the COMPATLEVEL=12.0, you can make this change dynamically without a control file format.

In this example, there is a system named GK99 and an alias of 99 that we wish to rename to GK62 and an alias of 62.

The current INIT member has:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED), (GK23,23,XE23,INIT=FREED), (GK49,49,XE49,INIT=FREED), (GK99,99,XE99,INIT=FREED) GLOBALVALUE VCFMASTER=(GK03,GK13,GK23,GK49,GK99), MOSTP=YES MIMINIT COMMUNICATION=XCFALLOCATE DDNAME=MIMCKP00,DSNAME=KERGL01.&SYSNAME..CKPT00

The following diagram shows you how to change a system name without a control file format:


13-Jul-2016 80/86

1. Confirm the COMPATLEVEL is 12.0

MIM0067I Command DISPLAY 308 MIM0037I MIM INIT display: BATCHJOB=NO BLKSIZE=6144 CANCEL=YES CHKPTDSN=NONE CMDPREFIX=@ X'7C' COMMANDS=MIMCMNDS COMMUNICATION=XCF COMPATLEVEL=12.00 DEVCLASS=NONE

2. Confirm that the system you wish to rename is currently FREED or DISABLED:

@D SYS


13-Jul-2016 81/86

@D SYS GPK0067I Command DISPLAY 579 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 LOCAL MASTER zOS 2016.113 10:37:53.28 02 13 GK13 EXTERNAL ACTIVE zOS 2016.113 10:37:52.31 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL FREED 05 99 GK99 EXTERNAL FREED

3. Remove the target system from the VCFMASTER list:

GPK0067I Command DISPLAY 879 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK23, *GK49, *GK99 ) ELIGIBLE MASTER LIST: GK03, GK13 @GLOBALV VCFMASTER=(GK03,GK13,GK23,GK49) GPK0067I Command GLOBALVALUE 952 GPK0386W System GK23 is ineligible to become VCF MASTER GPK0386W System GK49 is ineligible to become VCF MASTER GPK0193I GLOBALVALUE command processing initiated GPK0363I GLOBALVALUE command changes pending on external systemsGPK0362I GLOBALVALUE command processing complete @D GLOBALV GPK0067I Command DISPLAY 981 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK23, *GK49 )

4. Enter the ALTERSYS command on one system only, as follows:

@ALTERSYS GK99,NAME=(GK62,62),STATUS=ENABLED GPK0067I Command ALTERSYS 066 GPK0712I Processing ALTERSYS command... GPK0706I System GK62 attribute altered: ENABLED GPK0711I ALTERSYS Command Succeeded

5. Confirm the name has been changed:

@D SYS GPK0067I Command DISPLAY 071 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 LOCAL MASTER zOS 2016.113 10:47:11.78 02 13 GK13 EXTERNAL ACTIVE zOS 2016.113 10:47:11.69 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL FREED 05 62 GK62 EXTERNAL FREED

6. Enter GLOBALVALUE with new system name:

@GLOBALV VCFMASTER=(GK03,GK13,GK23,GK49,GK62) GPK0067I Command GLOBALVALUE 226 GPK0386W System GK23 is ineligible to become VCF MASTER GPK0386W System GK49 is ineligible to become VCF MASTER GPK0386W System GK62 is ineligible to become VCF MASTER GPK0193I GLOBALVALUE command processing initiated GPK0363I GLOBALVALUE command changes pending on external systemsGPK0362I GLOBALVALUE command processing complete @D GLOBALV GPK0067I Command DISPLAY 269 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK23, *GK49, *GK62 )


13-Jul-2016 82/86

7. Update the init member as needed:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED), (GK23,23,XE23,INIT=FREED), (GK49,49,XE49,INIT=FREED), (GK62,62,XE62,INIT=FREED) GLOBALVALUE VCFMASTER=(GK03,GK13,GK23,GK49,GK62), MOSTP=YES ALLOCATE DDNAME=MIMCKP00,DSNAME=KERGL01.&SYSNAME..CKPT00

8. Start MIM on GK62 with FORMAT=CHKPT (if a checkpoint file is being used)

GPK0067I Command DISPLAY 386 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL MASTER zOS 2016.113 11:01:35.78 02 13 GK13 EXTERNAL ACTIVE zOS 2016.113 11:01:35.72 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL FREED 05 62 GK62 LOCAL ACTIVE zOS 2016.113 11:01:35.79 @D GLOBALV GPK0067I Command DISPLAY 389 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK23, *GK49, GK62 ) ELIGIBLE MASTER LIST: GK03, GK13, GK62


13-Jul-2016 83/86

How to Remove a System from the MIMPLEX when COMMUNICATION=DASDONLY

The REMOVE Command is no longer supported beginning with release 12.0. This procedure shows you how to hide an old system using the ALTERSYS command.

Following is the current DEFSYS:

DEFSYS (GK03,03,XE03,INIT=FREED), (GK13,13,XE13,INIT=FREED), (GK23,23,XE23,INIT=FREED), (GK49,49,XE49,INIT=FREED), (GK62,62,XE62,INIT=FREED)

In this example, your goal is to remove the GK49 from the MIMPLEX:

1.Ensure that the GK49 is currently FREED:

@D SYSTEMS GPK0067I Command DISPLAY 026 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL ACTIVE zOS 2016.116 09:44:01.92 02 13 GK13 LOCAL ACTIVE zOS 2016.116 09:44:01.93 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL FREED 05 62 GK62 EXTERNAL FREED

2.Enter the ALTERSYS command to change the status to DISABLED

@ALTERSYS GK49,STATUS=DISABLED GPK0067I Command ALTERSYS 036 GPK0712I Processing ALTERSYS command... GPK0706I System GK49 attribute altered: DISABLEDGPK0711I ALTERSYS Command Succeeded

3.Next, verify that the system is now hidden from the DISPLAY SYSTEMS command output:

GPK0067I Command DISPLAY 249 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL ACTIVE zOS 2016.116 09:46:24.89 02 13 GK13 LOCAL ACTIVE zOS 2016.116 09:46:24.94 03 23 GK23 EXTERNAL FREED 05 62 GK62 EXTERNAL FREED

Note that system GK49 is now missing from the display. GK49 is still in the control file, but beginning with 12.1 GA, systems that are DISABLED are hidden from the DISPLAY SYSTEMS command.

4. If you really want to see all systems, including DISABLED, enter the following command:

@D SYSTEMS=STATUS=BOTH GPK0067I Command DISPLAY 227

GPK0108I SYSTEMS DISPLAY


13-Jul-2016 84/86

GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 EXTERNAL ACTIVE zOS 2016.116 09:45:57.89 02 13 GK13 LOCAL ACTIVE zOS 2016.116 09:45:57.90 03 23 GK23 EXTERNAL FREED 04 49 GK49 EXTERNAL DISABLED 05 62 GK62 EXTERNAL FREED

5. Next, update necessary MIMPARMS, including:

Remove the checkpoint file allocation(if used)

>Remove any IFSYS statements(if used)

>Do *not* remove the old system from the DEFSYS statement as this may cause inconsistencies between the control file and the checkpoint file.

This procedure hides the old system from the DISPLAY SYSTEMS command. If you want to completely remove the system from the control file, you will need to update the DEFSYS statement and restart MIM globally with a control file format.


13-Jul-2016 85/86

How to Remove a System from the MIMPLEX when COMMUNICATION=CTCONLY

In this example, we are going to remove the system GK23 from the MIMPLEX.

1.First, enter the DISPLAY SYSTEMS command and verify that GK23 is currently FREED:

@D SYSTEMS GPK0067I Command DISPLAY 101 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 LOCAL MASTER zOS 2016.117 08:53:49.44 02 13 GK13 EXTERNAL ACTIVE zOS 2016.117 08:53:49.04 03 23 GK23 EXTERNAL FREED zOS 2016.117 08:53:46.22 04 49 GK49 EXTERNAL ACTIVE zOS 2016.117 08:53:49.24 05 62 GK62 EXTERNAL ACTIVE zOS 2016.117 08:53:49.36

2.Next, enter the DISPLAY GLOBALVALUE command and determine if GK23 is present in the VCFMASTER list:

@D GLOBALVALUE GPK0067I Command DISPLAY 203 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, *GK23, GK49, GK62 ) ELIGIBLE MASTER LIST: GK03, GK13, GK49, GK62

3.Since GK23 was in the list, it needs to be removed. This can be done by entering the GLOBALVALUE command without GK23:

@GLOBALVALUE VCFMASTER=(GK03,GK13,GK49,GK62) GPK0067I Command GLOBALVALUE 210 GPK0193I GLOBALVALUE command processing initiated GPK0363I GLOBALVALUE command changes pending on external systems GPK0362I GLOBALVALUE command processing complete @D GLOBALVALUE GPK0067I Command DISPLAY 223 GPK0373I MIM GLOBALVALUE display: ANYELIGIBLE=YES MOSTPREFERRED=YES NOMASTER=WAIT VCFMASTER=( GK03, GK13, GK49, GK62 ) ELIGIBLE MASTER LIST: GK03, GK13, GK49, GK62

4.Next, Change GK23 to a DISABLED status:

@ALTERSYS GK23,STATUS=DISABLED GPK0067I Command ALTERSYS 257 GPK0712I Processing ALTERSYS command... GPK0706I System GK23 attribute altered: DISABLED GPK0711I ALTERSYS Command Succeeded

5.Next, confirm that GK23 is now missing from the DISPLAY SYSTEMS command output:

@D SYSTEMS GPK0067I Command DISPLAY 262 GPK0108I SYSTEMS DISPLAY INDEX ALIAS SYSTEM RELATION STATUS OPSYS LAST ACCESS 01 03 GK03 LOCAL MASTER zOS 2016.117 08:56:22.54 02 13 GK13 EXTERNAL ACTIVE zOS 2016.117 08:56:22.03 04 49 GK49 EXTERNAL ACTIVE zOS 2016.117 08:56:21.57 05 62 GK62 EXTERNAL ACTIVE zOS 2016.117 08:56:22.16


13-Jul-2016 86/86

6.Next, update necessary MIMPARMS, including:

>Remove the checkpoint file allocation, unless system symbolics are used as part of the dataset name.

>Remove any IFSYS statements(if used)

>Remove the system from the GLOBALVALUE VCFMASTER list.

>Comment out or delete any CTCPATH statements containing the system being removed.

>Do *not* remove the old system from the DEFSYS statement as this may cause inconsistencies between the control file and the checkpoint file.

This procedure hides the old system from the DISPLAY SYSTEMS command. If you really want to completely remove the system from the control file, you will need to update the DEFSYS statement and restart MIM globally with a control file format.

ca mim for z/os - 12.5 scenarios · ca mim for z/os - 12.5 13-jul-2016 3/86 ... how to change a...

Documents