S/390 PartnerWorld for Developers, ITSO/EFS Project EFS Systems on a Linux Base: Additional Topics

124 EFS Systems on a Linux Base: Additional Topics

Once all that was done, we could use RTP to access any member of the sysplex from any

other system on our network that was defined to the Cross Domain Resource Manager. Only

system zos1 needed to be connected to the VTAM network and defined in the CDRM major

nodes. See “SNA over Ethernet” on page 153 for a complete description of the VTAM

parameters used for such a configuration. The following messages issued during an IPL of

the sysplex confirmed that the two SSCP’s were correctly configured to use RTP:

ISG011I SYSTEM ZOS2 -JOINING GRS COMPLEX

ISG004I GRS COMPLEX JOINED BY SYSTEM ZOS2

IST1086I APPN CONNECTION FOR FLEX.ZOS2SSCP IS ACTIVE -TGN =21

IST093I ISTP0102 ACTIVE

IST1488I ACTIVATION OF RTP CNR00002 AS PASSIVE TO FLEX.ZOS2SSCP

IST1488I ACTIVATION OF RTP CNR00001 AS ACTIVE TO FLEX.ZOS2SSCP

IST1096I CP-CP SESSIONS WITH FLEX.ZOS2SSCP ACTIVATED

TSO logon restrictions

The sysplex enforces session limits sysplex-wide; that is, a TSO user may only log on to one

system at a time.

8.14 Setting up OMVS

Setting up UNIX System Services for sysplex operation consists of several steps; we have

already shown how we used system variables to mount the sysplex root, system root, and

system-specific files systems in member BPXPRM00.

We created these files using DFSMS; we first cloned the ADCD volume Z4USS1 following the

recommended procedure for dumping and restoring HFS files; we then duplicated and

renamed the system-specific files. Finally, we created the new sysplex root (there is only

one); we created a unique system root for each member of the sysplex.

Cloning the HFS volume

The ADCD distribution ships all Hierarchical File System (HFS) files on one volume, Z4USS1.

To begin the process, we copied the files to a new volume, ZP4US1, using this job:

//COPYUS2 JOB 0,FLEX,CLASS=Z,MSGCLASS=A,TIME=720,REGION=30M

//DSS EXEC PGM=ADRDSSU

//SYSPRINT DD SYSOUT=*

//Z4USS1 DD VOL=SER=Z4USS1,UNIT=3390,DISP=SHR

//ZP4US1 DD VOL=SER=ZP4US1,UNIT=3390,DISP=SHR

//OUTDD1 DD DSN=&HFSDUMP,

// DCB=(BLKSIZE=32760),

// DISP=(NEW,DELETE,DELETE),

// SPACE=(CYL,(950,550)),

// UNIT=SYSALLDA

//SYSIN DD *

DUMP -

ALLDATA(*) -

ALLEXCP -

CANCELERROR -

COMPRESS -

DATASET( -

INCLUDE( HFS.*,HFS.** ) -

EXCLUDE( SYS1.**, SYS1.*, HFS.ADCPL.ROOT )) -

LOGINDDNAME(Z4USS1) -

OPTIMIZE(4) -

OUTDDNAME(OUTDD1) -

TOLERATE(ENQF) -

Chapter 8. z/OS base sysplex on z/VM 125

WAIT(2,2)

SERIAL

RESTORE -

DATASET( -

INCLUDE(**)) -

INDDNAME(OUTDD1) -

TOLERATE(ENQF)-

OUTDDNAME(ZP4US1) -

RECATALOG( CATALOG.ZP4.VZP4RS2 ) -

RENAMEU( HFS.P390.* ,HFS.ZOS1.* ) -

STORCLAS(DEFAULT)

/*

In the process, the system name portion of the system-specific file systems was changed to

ZOS1. Thus, HFS.P390.DEV became HFS.ZOS1.DEV, and so forth. We excluded the

sysplex root (ADCDPL.ROOT) because we intended to create a new one.

We used a similar job to create a new set of system-specific files systems by copying the

existing ones:

//CLONUSS JOB 0,FLEX,CLASS=Z,MSGCLASS=A,TIME=720,REGION=30M

//DSS EXEC PGM=ADRDSSU

//SYSPRINT DD SYSOUT=*

//ZP4US1 DD VOL=SER=ZP4US1,UNIT=3390,DISP=SHR

//OUTDD1 DD DSN=&HFSDUMP,

// DCB=(BLKSIZE=32760),

// DISP=(NEW,DELETE,DELETE),

// SPACE=(CYL,(950,550)),

// UNIT=SYSALLDA

//SYSIN DD *

DUMP -

ALLDATA(*) -

ALLEXCP -

CANCELERROR -

COMPRESS -

DATASET( -

INCLUDE( HFS.ZOS1.* )) -

LOGINDDNAME(ZP4US1) -

OPTIMIZE(4) -

OUTDDNAME(OUTDD1) -

TOLERATE(ENQF) -

WAIT(2,2)

SERIAL

RESTORE -

DATASET( -

INCLUDE(**)) -

INDDNAME(OUTDD1) -

TOLERATE(ENQF) -

OUTDDNAME(ZP4US1) -

RECATALOG(CATALOG.ZP4.VZP4RS2) -

RENAMEU( HFS.ZOS1.*, HFS.ZOS2.*) -

STORCLAS(DEFAULT)

/*

Creating the sysplex root file system and new system-specific root file systems proved to be a

little tricky. When we ran the jobs provided in SYS1.SAMPLIB, we got an error because the