The following information can be used as a guideline in determining
the size of this data set.
Table 1
contains the size and number of records written when a checkpoint
is taken. The number of tracks or the amount of tape occupied by
the checkpoint data set can be determined by applying the number of
records and their sizes against either the track capacities of the
direct access device or the recording density and type for the magnetic
tape device.
Table 1. Determining the Size
of a Checkpoint Data SetDescription of Record |
Size (in bytes) |
Number Required |
---|
Checkpoint header record (CHR) |
400 |
1 |
DD name table (DDNT) |
400 |
(((D+20)/21)+1)/2 |
Data set descriptor record (DSDR) |
400 |
(A+1)/2 |
Problem program image record (PPIR) |
4096 |
B |
Supervisor record (SUR) |
4096 |
C |
Subsystem checkpoint record (SSCR) |
|
|
|
4096 |
Number of SYSIN or SYSOUT plus number of DDs with SUBSYS= |
|
20/4096 |
E/20 |
|
4096 |
1 |
Records Code:
A
= the number of data sets defined in the job step
B = (Storage
allocated - freespace + (20 * number of gaps in allocated storage))/4076.
-
- Example of calculation of B:
- Storage allocated = 9000 to 4FFFF and 1A000000 to 2FFFFFFF
- Freespace=
|
|
|
|
|
|
|
Size |
Number of Gaps |
|
200 |
1 |
|
3700 |
1 |
|
8000 |
1 |
|
0 |
1 (not allocated; counts as a gap) |
|
17F9000 |
1 |
|
1900000 |
1 |
|
3104900 |
6 |
- Therefore:
- B = ((47000 + 16000000) - 3104900 + (X'14' * 6)) /X'FEC'
- B = 130C0
Note: All values in the above example are in hexadecimal.
- C =
- A variable number of records for system and protected data management
control blocks. This will be approximately three records plus one
record for every three DASD data sets and one record for every ten
non-DASD data sets.
- D =
- Number
of JCL-specified data sets dynamically unallocated at checkpoint.
- E =
- One for every 90 ESPIEs established before checkpoint.
|
|
|
|
|
|
Checkpoint/restart's
dynamic storage requirement varies with the resources used by the
caller. The largest impact on the dynamic storage requirement results
from the amount of storage allocated to the job and the amount of
storage fragmentation. Checkpoint/restart's dynamic storage is allocated
out of subpools 229 and 230.