The Archive function compares your data set selection criteria to the output from an
IDCAMS DCOLLECT to evaluate the candidate L0 disk data sets. Data sets that meet
the criteria are moved to the archive.
About this task
This topic tells you how to archive L0 disk data sets. If you need to archive explicitly
identified ML2 data sets or use the Archive function to implement an ML3-like storage tier, your
task is documented elsewhere:
By default, the Archive function invokes an IDCAMS DCOLLECT and uses the
DCOLLECT output and your data set selection criteria to evaluate the candidate L0
disk data sets. However, you can specify that you want the Archive function to use the output from a
earlier execution of DCOLLECT for this purpose.
You can use console commands to monitor and control the Archive process as it executes. See step
9 for more information.
Note:
- Users at sites that use DFSMShsm ABARS should be aware that ABARS cannot back up directly from
the archive media or restore directly to the archive media the way it does with DFSMShsm-migrated
files. When automatic Restore is enabled, Advanced Archive for DFSMShsm becomes aware
of the ABARS attempt to access the data set and restore it to DFSMShsm. When automatic Restore is
not enabled, ABARS cannot back up the archived files. For these reasons, do not archive files that
are likely to be backed up by ABARS.
- When you archive VSAM L0 disk data sets, the entire VSAM sphere for the base cluster is
archived; therefore requesting that an individual component of a VSAM sphere be archived can cause
the request to be rejected.
Procedure
-
Open
SAXQSAMP library member AXQRCHIV for editing.
-
Provide appropriate job card information.
-
Replace all occurrences of
@HLQ with your data set prefix.
-
Provide the data set selection criteria that will classify candidate data sets as inactive and
the criteria that will exclude data sets from Archive processing. Use the information and examples
that are provided in Selection criteria keyword reference for Archive function, as a guide for criteria
keyword specification.
- You may code the data set selection criteria in the job itself.
- You may code the data set selection criteria in member CRITERIA of the
product parameter library.
-
Add the ARCHIVE_SOURCE=L0 data set selection criteria to the rest of
your data set selection criteria.
Do not include any blank spaces in your specification. For example,
ARCHIVE_SOURCE=L0 is valid, but
ARCHIVE_SOURCE= L0 and
ARCHIVE_SOURCE = L0 are not valid specifications.
Note: Your AXQRCHIV JCL can
include multiple ARCHIVE_SOURCE= criteria, but Archive processing honors only the
first occurrence. If you omit the ARCHIVE_SOURCE= criteria from the JCL, Archive
processing assumes that the candidate data sets are ML2 data sets.
-
If you want the Archive process to invoke an IDCAMS DCOLLECT and compare
your data set selection criteria against the
DCOLLECT output to evaluate the
candidate L0 disk data sets, go on to step 7. Otherwise, if you want the Archive process to use the
output from an earlier IDCAMS DCOLLECT process to compare against your data set
selection criteria, add a //FLATDCOL DD statement to the JCL.
Here is an example of a
//FLATDCOL DD statement:
//FLATDCOL DD DISP=SHR,DSN=ARCHMGR.DCOLLECT.OUTPUT
//FLATDCOL DD DISP=SHR,DSN=ADVARCV.DCOLLECT.OUTPUT
where
ADVARCV.DCOLLECT.OUTPUT is the name of the flat file that contains the output
from the earlier DCOLLECT process.
Note: When the //FLATDCOL DD
statement is present in the Archive JCL, Archive processing bypasses the function that determines
the catalog status of the candidate data sets and the function that determines the calculated
expiration date of the candidate data sets.
- Optional:
If you want to verify before archiving the data that the Archive process is going to produce
the expected result, run the Archive job in simulation mode:
-
Add PARM=SIM to the
EXEC control card in the JCL.
-
Run the simulation job.
-
Review the information that was written to the
ARCHRPT DD to determine which
data sets were selected for processing.
This is an example of the information written to an
ARCHRPT report:
VER 1.1 ***SIMULATE*** A D V A N C E D A R C H I V E FOR DFSMShsm ***SIMULATE*** AXQRCHIV JOB10487
DATE: 01/30/2018.030 TIME: 00:48:19 CANDIDATE RECORDS PASSED TO ARCHIVE TASK FOR RSPLEX0D PAGE 1
ORIG DATA SET NAME SIZE IN KB RESULTS
<VSER> <----------------------------------------> <--------> <----------------------------->
--------------------------------------------------------------------------------------------------------------------
START OF GROUP: RAMSE434
R1PD01 TSTAXQ.AXQ.AXQSE434.PS02 56 SIM: ARCHIVE FUNCTIONS BYPASSED
R1PD16 TSTAXQ.AXQ.AXQSE434.PS03 56 SIM: ARCHIVE FUNCTIONS BYPASSED
R1PD05 TSTAXQ.AXQ.AXQSE434.PS04 56 SIM: ARCHIVE FUNCTIONS BYPASSED
R1PD04 TSTAXQ.AXQ.AXQSE434.PS05 56 SIM: ARCHIVE FUNCTIONS BYPASSED
SIZE OF PROCESSED GROUP: 0.000GB #DATASETS IN GROUP: 4
********************************************************************************************************************
TOTAL ARCHIVED: 0.000GB #DATASETS: 4 #GROUPS PROCESSED: 1
******************************** BOTTOM OF DATA ********************************************************************
-
If the simulation results are not as you expected them to be, make one or both of these changes
and then repeat the simulation until the results are satisfactory.:
- Modify the data set selection criteria.
- Use the output from a different IDCAMS DCOLLECT process to compare against
your data set selection criteria.
-
When the results confirm that the DCOLLECT output and the selection criteria
are identifying all of the L0 disk data sets that you want to archive, remove the
SIM value from the EXEC control card and then go to step 8.
Note: Initially, it is likely that a large volume of data will be selected for archiving. It is
strongly recommended that you run the initial Archive job in simulation mode so that you can review
the data that would be selected for archiving before any of the data is actually archived. Note that
you can use the
ARCHIVE-SELECTION-MAXIMUM-FILES or
ARCHIVE-SELECTION-MAXIMUM-GB (or both) parameters to limit the amount of data that
is processed by any one Archive job. See
Parameters, for more
information.
-
Submit the job for execution.
Note: Should you need to shut down HSM, you must first stop the Archive job.
- Optional:
To monitor or stop the Archive process, use the appropriate console command.
| Console command |
Description |
| STATUS |
When you issue the STATUS command, the Archive process
issues a WTO message that shows the following information:
-
The current number of requests that are queued
-
The number of requests that completed successfully
-
The number of requests that failed to complete
Syntax:
F jobname,STATUS
where jobname is the
job name of the job currently executing.
|
| HALT |
When you issue the HALT command, the Archive process
terminates after the current group of queued requests has been processed. Any remaining candidate
data sets that are not yet queued for processing are ignored without any additional processing or
reporting. Syntax:
F jobname,HALT
where jobname is the job
name of the job currently executing.
|
| HALT,I |
When you issue the HALT,I command, the Archive process
terminates when the current data set in the group of queued requests has been processed. Any
remaining candidate data sets that are not yet queued for processing are ignored without any
additional processing or reporting. Syntax:
F jobname,HALT,I
where jobname is the
job name of the job currently executing.
|
What to do next
Follow the procedures in Verifying the contents of the archive to confirm that the correct
data sets were moved to the archive.If the Archive function issues error messages or verification
reveals that the Archive process did not produce the expected results, go on to Creating a candidate rejection report to create a report that provides
information about why candidate data sets were not processed.