|
Notes |
|
|
|
|
There is no possibility to determine the size of a member in a PDSE accept to read the whole member twice. This is too expensive to solve this issue. Currently it looks like we must close the issue as not fixable. |
|
|
|
|
In a PDS the TTR can be used to determine the member size, but for this the whole dictory must be read and sort based on the TTR to determine the differences.
This will also very expensive and based on the fact that this solution will not work for PDSE a general solution of this problem is not possible. |
|
|
|
|
|
The DFSORT has the same problem and no solution for it. The overspace issue is describt in the documantation. |
|
|
|
|
|
The only possible solution would be to reduce the space if the dynalloc function fails with the VTOC error and a member is written as PS. |
|
|
|
|
|
With version 9.1.10 at data set allocation the primary and secondary space are reduced by factor 2 if the dynamic allocation function failed down to the default values. |
|
|
|
|
|
There is no cabebility to determine the used space of a member except to read it twice. The only solution for this problem is, to reduce the primary and secondary space if the dynalloc() function failed, down to the minimum defaults. If a member of a big PDS is written as PS dataset, then the overallocated space are release at close. If such a member written to a PDS(E) the size of the PDS(E) will reduce to the maximum which is possible to allocate. |
|