[ Bottom of Page | Previous Page | Next Page | Contents | Index ]

DUMP

Types of DUMP command keywords

There are three groups of DUMP command keywords:

• The I/O file keywords allow you to specify catalogs, back up files, and merge files.
• The Inclusion/Exclusion keywords allow you to include specific data sets on the backup file and to exclude specific data sets from the backup file.
• The Miscellaneous keywords allow you to specify parameters to customize your DUMP processing operations.

DUMP command syntax

Syntax

The syntax diagram below shows the DUMP command and its keywords. The use of lowercase in an operand indicates a variable for this operand. For example, "catalogname" is a variable that you would provide when you type the command.

>>-DUMP--,-----------------------------------------------------><
 
I/O file keywords:
 
|--+-TODD--=(--ddname1,ddname2--)-----+-------------------------|
   +-FROMCAT--=--catalogname----------+
   +-MERGEDD--=--ddname---------------+
   '-VERIFYDD--=(--ddname1,ddname2--)-'
 
Inclusion and exclusion keywords:
 
|--+---------------------------------------------+--------------|
   +-ALTINDX = {INCLUDE|SELECT|NONE}-------------+
   +-ANDVOL--------------------------------------+
   +-EXCDELETE--=--dsn---------------------------+
   +-EXCDSN--=--dsn------------------------------+
   +-EXCGEN--=--genericdsn-----------------------+
   +-EXCUNUPDATED--=--ccyy.ddd.hh.mm.ss----------+
   +-EXCVOL--=(--volser1,volser2,...,volser10--)-+
   +-INCDSN--=--dsn------------------------------+
   +-INCGEN--=--genericdsn-----------------------+
   +-INCRNG--=(--genericdsn1,genericdsn2--)------+
   +-INCUNUPDATED--=--ccyy.ddd.hh.mm.ss----------+
   +-INCVOL--=(--volser1,volser2,...,volser10--)-+
   '-NOANDVOL------------------------------------'
 
Miscellaneous keywords:
 
|--+----------------------------------------------+-------------|
   +-ABENDLIMIT--=--nn----------------------------+
   +-CATSTATS--=--(exit1[,exit2,...,exit8])-------+
   +-CONSOLE={ALL|ERROR|NONE}---------------------+
   +-DELETE---------------------------------------+
   +-DELEXIT--=--exitname-------------------------+
   +-DSSTATS--=--(exit1[,exit2,...,exit8])--------+
   +-DUMPOPEN-------------------------------------+
   +-ERASE----------------------------------------+
   +-FASTDELETE-----------------------------------+
   +-FORCE----------------------------------------+
   +-HSMPROCESS-----------------------------------+
   +-ICFDUMP--------------------------------------+
   +-IOOPTS={FASTPATH|NOFASTPATH}-----------------+
   +-MASSDELETE={SIMULATE|NOSIMULATE}-------------+
   +-MCDSNM1--=--mcdsname-------------------------+
   +-MCDSNM2--=--mcdsname-------------------------+
   +-MCDSNM3--=--mcdsname-------------------------+
   +-MCDSNM4--=--mcdsname-------------------------+
   +-MOUNT----------------------------------------+
   +-MULTICAT-------------------------------------+
   +-NODUMPOPEN-----------------------------------+
   +-NOERASE--------------------------------------+
   +-NOMOUNT--------------------------------------+
   +-NOMULTICAT-----------------------------------+
   +-NOPRHSM--------------------------------------+
   +-NOPURGE--------------------------------------+
   +-NOVER----------------------------------------+
   +-OLD------------------------------------------+
   +-PURGE----------------------------------------+
   +-RECEXIT--=--exitname-------------------------+
   +-SELECTEXIT--=--exitname----------------------+
   +-SHR------------------------------------------+
   +-STATEXIT--=--exitname------------------------+
   +-SUPPRESSDATA---------------------------------+
   +-SUPSTATS--=(--exit1--[,--exit2,...,exit8--])-+
   +-SYSTEMBACKUP---------------------------------+
   '-VERX-----------------------------------------'
 

Keywords for DUMP command

I/O File keywords

The table below provides input/output keyword definitions and short forms. Following this table, each of the keywords is described in more detail.

TODD=(ddname1[,ddname2])

(Required) Specifies the DD name of the JCL statement that identifies the backup file. The number of DD names you specify indicates the number of backup files AVM will produce. If you specify TODD more than once, the last statement overrides the previous TODD. You can also use the short form of TODD, which is TD.

FROMCAT=catalogname

Identifies the name of the catalog from which you want AVM to dump data sets. If you do not specify FROMCAT, or MULTICAT, Advanced VSAM Manager uses the high-level qualifier of the first data set name you specified in an INCDSN or INCGEN keyword as an alias to locate the correct catalog. You can also use the short form of FROMCAT, which is FC.

MERGEDD=ddname

Specifies the DD name of the JCL statement that identifies the previous backup file AVM is merging with the VSAM data sets it is dumping to the new backup file.

Note:

The MERGEDD and DELETE keywords are mutually exclusive.

VERIFYDD=(ddname1[,ddname2])

When you use this keyword, the DD name(s) must match the DD name(s) you specified in the TODD statement of this command. After AVM has dumped all selected data sets, it reopens and reads the backup file(s) to verify the readability of the contents.

Note:

If you use the VERIFYDD keyword, you must also include the DD statement: //VERLIST DD SYSOUT=*.

Inclusion and exclusion keywords

This section describes the inclusion and exclusion keywords that you can use for DUMP processing. Be aware that if you specify inclusion and exclusion keywords on the same DUMP operation, exclusion keywords prevail.

With the exception of ALTINDX, you can specify each exclusion or inclusion keyword more than once in the same DUMP command, with a cumulative effect. For example, if you specify EXCDSN (Exclude DSN) twice in the same DUMP command, AVM excludes both of the specified data sets from the backup file.

When you use more than one inclusion keyword in the same DUMP command, AVM dumps data sets meeting any one of the specified criteria, unless they are excluded. You can include the ANDVOL keyword if you want to select data sets from a specific volume that meet other inclusion criteria. When ANDVOL, INCVOL, and other inclusion keywords are specified together, data sets must match both the INCVOL specification and another inclusion specification.

To display which data sets will be selected based on your inclusion/exclusion specification, without creating a backup, you can specify the following command:

DUMP FROMCAT=xxxx,TODD=BACKUP,INCGEN=xxx,EXCDSN=xxx,
SUPPRESSDATA

Then, include a dummy DD statement for the TODD; for example:

//BACKUP DD DUMMY

The table below lists the inclusion and exclusion keyword definitions and short forms for the DUMP command. Following this table, each of the keywords is described in more detail.

The keywords are described below.

ALTINDX={INCLUDE|SELECT|NONE}

Lets you decide how you want AVM to treat alternate indexes during backups. If you do not specify ALTINDX, Advanced VSAM Manager does not back up alternate indexes unless it dumps data sets by volume (INCVOL keyword).

INCLUDE

Advanced VSAM Manager backs up alternate indexes related to clusters selected for backup. When executing DUMP with ALTINDX=INCLUDE, backups for all related alternate indexes and path entries will be created subsequent to the backup of the related base cluster.

• When ALTINDX=INCLUDE is specified, alternate index data set names are not evaluated with regard to DSN inclusion/exclusion criteria. Related alternate indexes and paths are treated as components of their related base cluster and are identified (for selection purposes by RESTORE) by the base cluster's data set name.
• If you are dumping by volume (using INCVOL), SELECT processing is forced in conjunction with this keyword. Internal activation of SELECT processing causes AVM to back up an alternate index twice if the alternate index and the particular alternate index's base cluster both occupy the same volume.

SELECT

AVM selects alternate indexes for DUMP processing based on data set names rather than by association with selected base clusters.

Note:

This is also true if you specify ALTINDX=INCLUDE and AVM is dumping data sets by volume (INCVOL)

• When ALTINDX=SELECT is specified, alternate index names are interrogated for DSN inclusion/exclusion and will be selected for DUMP processing based on DSN selection criteria. If selected for processing (when ALTINDX=SELECT), DUMP will create a backup of the alternate index, independent of the related base cluster, in the order in which the data set name is retrieved from the catalog. AVM will also back up path entries related to selected alternate indexes.
• When restoring from a backup created with ALTINDX=SELECT, you must verify that you restored an alternate index's base cluster before you restore the alternate index itself. To do this, restore only clusters (by specifying ALTINDX=NONE) and then restore alternate indexes with SELECT and do not specify REPLACE.
• SELECT and FASTDELETE are mutually exclusive.

NONE

AVM does not back up alternate indexes, even during volume dumps.

ANDVOL

Allows you to specify AND logic in selecting data sets for DUMP processing by volume when you specify INCVOL. Only data sets that match both the INCVOL specification and another inclusion specification (FROMCAT, INCDSN, INCGEN, INCRNG, or INCUNUPDATED) are selected for backup.

EXCDELETE=dsn

AVM does not delete the designated data set after backup, even if you specify DELETE. Advanced VSAM Manager pads the data set name with blanks to 44 characters for comparison with the names of data sets that were backed up.

EXCDSN=dsn

Indicates that AVM should not copy a data set to the backup file. AVM pads the data set name with blanks to 44 characters for comparison with data set names in the catalog. Use of generic (partial) data set names is supported.

EXCGEN=genericdsn

Specifies a generic (partial) data set name you want AVM to exclude from the backup file. Advanced VSAM Manager ignores data sets whose names match the generic specification. The comparison with data set names in the catalog is for the length of the generic data set name specified.

EXCUNUPDATED=ccyy.ddd.hh.mm.ss

Excludes from the backup file any data sets that have not been opened since the specified date and time (ccyy.ddd.hh.mm.ss). Advanced VSAM Manager sends only modified data sets to the backup file.

You can specify the following special characters in the date and time; however, you must still place periods (.) in their proper positions.

• You can specify asterisks (*) to insert the current date and time. For example, ****.***.**.**.** indicates the current date and time and ****.***.10.00.00 indicates the current date at 10:00 AM.
• You can specify pound signs (#) in the day of the year positions (ddd) to insert the previous Julian day. For example, ****.###.10.00.00 indicates yesterday at 10:00 AM.

EXCVOL=(volser1[,volser2,...,volser10])

Specifies up to ten volume serial numbers. If a data set occupies one or more of the specified volumes, AVM does not back it up. If only one component or a part of a component resides on a specified volume, AVM does not back up the data set. The total number of volumes allowed is determined by the default table size established by the TABSIZE installation option and by the total number of volumes specified by EXCVOL and INCVOL.

INCDSN=dsn

Indicates the name of a data set you want AVM to copy to the backup file. AVM pads the data set name with blanks to 44 characters for comparison with data set names in the catalog. Use of generic (partial) data set names is supported.

INCGEN=genericdsn

Copies to the backup file any data set whose name matches the generic specification to the backup file. The comparison with data set names in the catalog is for the length of the generic data set name specified.

INCRNG=([genericdsn1],[genericdsn2])

Dumps data sets whose names begin with a character string between genericdsn1 and genericdsn2, inclusive, to the backup file. The default value for genericdsn1 (low-range boundary) is 44C' ' and the default value for genericdsn2 (high-range boundary) is 44X'FF'.

Note:

If you specify INCRNG, you must also specify the FROMCAT keyword.

INCUNUPDATED=ccyy.ddd.hh.mm.ss

Backs up data sets that have not been updated since the date and time you specify through this keyword. This allows you to select inactive data sets for backup and deletion.

You can specify the following special characters in the date and time; however, you must still place periods (.) in their proper positions.

• You can specify asterisks (*) to insert the current date and time. For example, ****.***.**.**.** indicates the current date and time and ****.***.10.00.00 indicates the current date at 10:00 AM.
• You can specify pound signs (#) in the day of the year positions (ddd) to insert the previous Julian day. For example, ****.###.10.00.00 indicates yesterday at 10:00 AM.

INCVOL=(volser1[,volser2,...,volser10])

Permits you to specify up to ten volume serial numbers for each INCVOL specification. If you need to specify more than ten volumes, you can specify the INCVOL keyword multiple times. The total number of volumes allowed is determined by the default table size established by the TABSIZE installation option and by the total number of volumes specified by INCVOL and EXCVOL.

If any data set components occupy any specified volumes, AVM dumps the entire data set when it processes the catalog. Additional guidelines for using INCVOL are listed below:

• If you specify INCVOL, you must also specify either the ICFDUMP keyword or one of the INCDSN, INCGEN, or FROMCAT keywords. These keywords are used to select the data sets to be dumped from the specified volume(s).
• If you specify INCVOL, AVM assumes ALTINDX=SELECT unless you specify ALTINDX=NONE. ALTINDX=SELECT causes AVM to back up all alternate indexes on the volume under their own names. If you specify ALTINDX=INCLUDE, both INCLUDE and SELECT are in effect. Activating both INCLUDE and SELECT causes AVM to back up an alternate index twice if the alternate index and the particular alternate index's base cluster both occupy the same volume.
• If you did not specify ALTINDX=NONE, INCVOL and FASTDELETE are mutually exclusive.
• You can include the ANDVOL keyword if you want to select data sets from a specific volume that meet other inclusion criteria. When you specify INCVOL, ANDVOL, and other inclusion keywords together, data sets must match both the INCVOL specification and another inclusion specification to qualify for the DUMP operation.
• You can specify the operands for INCVOL in any order.

NOANDVOL

Overrides the installation default of ANDVOL=Y. If you specify NOANDVOL, then OR logic is used when you specify the INCVOL keyword in the DUMP command. AVM selects data sets that match any inclusion specification (FROMCAT, INCDSN, INCGEN, INCRNG, and INCUNUPDATED) for backup.

Miscellaneous keywords

The table below provides a list of miscellaneous keywords. Following this table, each of the keywords is described in more detail.

The keywords are described below.

ABENDLIMIT=nn

Allows you to specify a value for the step condition code. If AVM processing returns a condition code that exceeds this value, AVM issues a user 2009 ABEND when the processing ends. You can use this facility to signal a greater warning indicator to the user monitoring the execution of the backup job.

CATSTATS=(exit1[,exit2,...,exit8])

Generates the HVMCATR1 statistics report (described in Introduction to statistics reports). You can specify up to eight user exits that you want AVM to invoke immediately after it opens each user catalog or master catalog. The specified exit is passed data for the catalog being processed.

CONSOLE={ALL|ERROR|NONE}

Allows you to control the AVM messages displayed on the operator's console.

ALL

Indicates that in addition to displaying all error messages, you want AVM to display a message as it backs up each data set.

ERROR

Indicates that you want AVM to display only error messages.

NONE

Blocks all messages to the operator's console.

DELETE

Deletes from the catalog those data sets it is currently backing up. These deletions occur after it closes the backup file and only if it has successfully backed up all selected data sets. DELETE and MERGEDD are mutually exclusive keywords.

Note:

DELETE, FASTDELETE, and MASSDELETE are mutually exclusive. These keywords cannot be specified in the same DUMP command.

DELEXIT=exitname

Specifies a user exit that you want AVM to invoke immediately before it deletes selected data sets. You can allow or suppress deletion through this exit.

DSSTATS=(exit1[,exit2,...,exit8])

Generates the data set statistics reports, HVMDSRP1, HVMDSRP2, and HVMDSRP3 (described in Introduction to statistics reports. You can specify up to eight user exits that AVM will give control to immediately after backing up selected data sets.

DUMPOPEN

When you specify this keyword along with the SHR keyword, AVM attempts to dump data sets that are open for update elsewhere on the system. If the data set being backed up is modified during the backup, the data of that particular data set could be unusable or inaccurate. If the data set that is opened for update is defined with the highest SHAREOPTIONS (1,1), the backup for this data set does not take place and AVM resumes processing with the next selected data set.

ERASE

Causes AVM to erase the data set data components using the IDCAMS DELETE command. You can specify the ERASE keyword only when you also specify the DELETE, FASTDELETE, or MASSDELETE keyword.

FASTDELETE

Deletes each data set after backing it up from the catalog. This is similar to an AMS permanent export for each data set being dumped. FASTDELETE and ALTINDX=SELECT are mutually exclusive.

Note:

DELETE, FASTDELETE, and MASSDELETE are mutually exclusive. These keywords cannot be specified in the same DUMP command.

FORCE

Forces Advanced VSAM Manager to ignore VSAM and AMS errors and continue DUMP processing. This is useful in certain data salvage efforts, but it can also be dangerous since the results are unpredictable. Be sure that you understand the operating principles of AVM before specifying this keyword.

HSMPROCESS

Indicates that AVM should recall and then back up HSM-migrated data sets. You must provide the name of the HSM Migration Control Data set (MCDS) for HSMPROCESS processing. You can do this at installation time or at run time using the MCDSNM1, MCDSNM2, MCDSNM3, and MCDSNM4 keywords.

ICFDUMP

Back up the volumes defined through INCVOL keywords as ICF volumes. In this mode, AVM accesses the VVDS on each volume to determine which ICF catalogs have data on the volume. AVM then scans each ICF catalog for data sets on the volume it is processing.

Note:

If you specify multiple volumes for backup, the data sets on the backup file may not be in collating sequence. This requires that you specify FULLSEARCH when restoring the data.

IOOPTS={FASTPATH|NOFASTPATH}

Allows you to specify the processing option to be invoked.

• FASTPATH indicates that AVM should use its fastest internal routines to perform backups.
• NOFASTPATH indicates that AVM should use VSAM level logic to perform backups.

Notes to users with vendor compression:

The IOOPTS=FASTPATH option backs up vendor-compressed data in compressed format. Specify IOOPTS=NOFASTPATH if you want vendor-compressed data to be backed up in decompressed format.

MASSDELETE={SIMULATE|NOSIMULATE}

Allows you to select data sets that you want AVM to list (SIMULATE) or delete (NOSIMULATE). AVM does not produce backups in either case. MASSDELETE is mutually exclusive to the TODD keyword.

Note:

DELETE, FASTDELETE, and MASSDELETE are mutually exclusive. These keywords cannot be specified in the same DUMP command.

MCDSNM1=mcdsname

Specifies the name of the HSM Migration Control Data set (MCDS). This specification is required for the HSMPROCESS function. If you have divided your MCDS into several members, you can specify up to three additional MCDS names using the MCDSNM2, MCDSNM3, and MCDSNM4 keywords.

MCDSNM2=mcdsname

If you divide the MCDS into several members, specifies the name of an additional HSM Migration Control Data set (MCDS).

MCDSNM3=mcdsname

If you divide the MCDS into several members, specifies the name of an additional HSM Migration Control Data set (MCDS).

MCDSNM4=mcdsname

If you divide the MCDS into several members, specifies the name of an additional HSM Migration Control Data set (MCDS).

MOUNT

When allocating data sets and catalogs, specifies mount volumes to dynamic allocation.

MULTICAT

Specifies a dump of data sets from multiple catalogs when used with INCGEN or INCDSN. MULTICAT and FROMCAT are mutually exclusive keywords.

NODUMPOPEN

Does not dump data sets that are open for update elsewhere on the system.

NOERASE

Does not erase the data set data components in data sets it deletes. This is the default setting.

NOMOUNT

Does not specify mount volumes to dynamic allocation when allocating catalogs and data sets.

NOMULTICAT

Turns off the MULTICAT function. A dump of data sets from multiple catalogs is not specified.

NOPRHSM

Indicates that AVM should not recall HSM-migrated data sets for backup. This keyword overrides the installation default option of HSMPROCESS=YES.

NOPURGE

For the DELETE function, date-protected data sets will not be allowed to be deleted if the retention period is still valid.

NOVER

Suppresses automatic data set verification (IDCAMS Verify). Before exporting a data set, AVM checks to see if it was properly closed. If it was not, AVM issues a VERIFY for the data set. If you specify NOVER, AVM does not issue the verify command.

OLD

Allocates selected data sets with DISP=OLD rather than DISP=SHR.

PURGE

For DELETE function, processes date-protected data sets.

RECEXIT=exitname

Specifies a user exit that you want AVM to enter for each record written to the backup file.

Note:

Refer to Introduction to User exits for more information on user exits.

SELECTEXIT=exitname

Specifies a user exit that AVM invokes when it selects a data set for backup. This exit can allow you to suppress the dumping of the data set.

Note:

Refer to Introduction to User exits for more information on user exits.

SHR

Allocates selected data sets with DISP=SHR (rather than DISP=OLD), allowing you to back up data sets allocated elsewhere in the system (provided they have not been allocated DISP=OLD by another task).

Note:

This specification is required for the DUMPOPEN keyword.

STATEXIT=exitname

Specifies a user exit that you want AVM to invoke after it dumps each selected data set.

Note:

Refer to Introduction to User exits for more information on user exits.

SUPPRESSDATA

Does not write data records to the backup file. Only the control information is written to the backup file and then used to DEFINE the data set during the restore.

SUPSTATS=(exit1[,exit2,...,exit8])

Allows you to specify up to eight catalog or data set statistics exits that will not be given control, even if you specified them in CATSTATS or DSSTATS keywords. This keyword is useful in suppressing reports that you specified as defaults when you installed AVM.

SYSTEMBACKUP

Backs up VSAM data sets in all connected user catalogs to one backup file. The data sets will be in ascending collating sequence within each catalog, requiring that you specify FULLSEARCH when you restore data sets from the backup file produced.

You can exclude catalogs from the backup by specifying their fully qualified names in an EXCDSN keyword. You can include and exclude data sets in the catalogs to be backed up by using the INCxxx and EXCxxx keywords.

You cannot specify the VERIFYDD keyword in conjunction with SYSTEMBACKUP.

VERX

Performs an IDCAMS VERIFY on any data sets that were not properly closed. The VERIFY fails against data sets opened on the system where AVM operates. AVM, however, verifies data sets open on other systems (in a non-GRS or non-MSX environment).

Specifying the DUMP command

You can execute DUMP by specifying as few as two keywords: TODD (which indicates the DD name of the JCL statement that identifies the backup file), and at least one of the following keywords:

• FROMCAT, which specifies the catalog whose data sets are being backed up
• INCDSN (include data set), which identifies a specific or generic data set to back up
• INCGEN (include generic), which specifies a generic (partial) data set name to back up

Example 1

Use the following instruction to instruct Advanced VSAM Manager (AVM) to back up all VSAM data sets in the catalog USER.CAT to the backup file specified by the DD statement named BACKUP:

DUMP FROMCAT=USER.CAT,TODD=BACKUP

Example 2

Use the IG (include generic) keyword to dump all data sets whose names start with TEST.VSAM from the catalog pointed to by alias TEST.

DUMP TODD=BACKUP,IG=TEST.VSAM,ALTINDX=INCLUDE

Example 3

Use the following instruction to back up all data sets from catalog USER.CAT that reside on a specific volume (VSAM01), except those with names starting with TEST.VSAM.

DUMP TODD=BACKUP,EG=TEST.VSAM,FROMCAT=USER.CAT,
IV=VSAM01

Example 4

Back up all data sets with names beginning with PAYR, containing any middle qualifier ('*'), and having a third node that begins with TECH. Use AVM's proprietary I/O routine (IO=FASTPATH) for faster backup.

DUMP TODD=BACKUP,IG=(PAYR.*.TECH/),IO=FASTPATH,
ALTINDX=INCLUDE

AVM also backs up any alternate indexes associated with these data sets.

Example 5

Back up all data sets with names beginning with TSOUSER. Use Advanced VSAM Manager's proprietary I/O routine (IO=FASTPATH) for faster backup. Delete data sets after successfully backing them up.

DUMP TODD=BACKUP,IG=(TSOUSER.**),IO=FASTPATH,DELETE,
VERIFYDD=BACKUP

AVM will also verify (VERIFYDD) the readability of all data sets backed up.

Special notes on DUMP processing

This section contains additional information about DUMP processing. It provides tips and explanations to get the most out of DUMP processing.

Issuing multiple DUMP commands

You can issue multiple DUMP commands in a single step. This can be done to produce duplicate backups for disaster recovery purposes, to do a FROMCAT DUMP of more than one catalog, or to process different dump selection criteria to distinct backups. An example is shown below.

DUMP IG=TEST.VSAM/,TODD=BACKUP
DUMP ID=PAYR.*.CUST,TODD=BACKUP2

To perform multiple DUMP commands successfully using a single output file, you will need to specify JCL DISP=MOD instead of “NEW”, “OLD”, or “SHR.” Additionally, you can create two AVM backups with a single DUMP command using TODD=(ddname1,ddname2).

DB2® data sets

Because of the non-standard format of the data set, attempting to back up a non-linear DB2 data set results in a HVM1117E message with a return code of 114. DB2 must be down while you back up DB2 system data sets (for example, a catalog or directory). We also strongly recommend that you take down DB2 when you back up DB2 application table spaces. When processing DB2 data sets, AVM is subject to the same restrictions as DFDSS.

Compressed data sets

To dump Compressed Extended Format Data sets in compressed format, you must use Advanced VSAM Manager's IOOPTS=FASTPATH processing. If AVM encounters an error dumping a Compressed Extended Format Data set during a backup where IOOPTS=FASTPATH is in effect, it will dynamically change to IOOPTS=NOFASTPATH and retry the dump again. In this case, the data set will be dumped in uncompressed format.

DFSMS/MVS

Data compressed by third-party vendor software is backed up in compressed format when IOOPTS=FASTPATH is in effect. Conversely, if IOOPTS=NOFASTPATH is in effect, vendor-compressed data is backed up in decompressed format.

DFSMShsm

HSM-migrated data sets are recalled and dumped by AVM if the HSMPROCESS option is in effect. HSMPROCESS requires that the MCDS name also be specified. For more information, refer to Example 2: Back up an HSM-migrated data set.

Length of data sets

You can use either IOOPTS processing option (FASTPATH or NOFASTPATH) to dump data sets greater than 32,756 in length.

Backup file format

The backup file that DUMP processing creates is a variable-blocked, spanned data set. This backup file has an LRECL of X, which allows Advanced VSAM Manager (AVM) to process data set records greater than 32,760 in length. You cannot alter the LRECL of the AVM backup file.

The backup file uses the optimum block size generated by the system as a default. You can override the default block size by including a block size specification (DCB=BLKSIZE) on the DD statement that references the backup file.

The DUMP backup file consists of the following for each data set:

• A header record
• A series of portable data set records containing catalog information needed to define the data set
• Data records from the data set
• A trailer record

The backup file has the following format for each data set:

A DSECT mapping the header record is provided in member HVMDHDR in the SHVMSAMP data set.

The table below shows the format and description of the header and trailer records.

The individual data sets on the backup file are identical to IDCAMS portable data sets. Each logical data record on the file is a VBS record.

AVM can extract and convert the data sets on the backup file to REPRO format or EXPORT format files (the latter being a standard IDCAMS portable data set).