Mainframe, MVS and zOS Discussion

Hi,

I am not sure if this is seen by everyone or it's only me. But why browse is substituted when I try to view or edit a dataset with a lot of or huge number of records? Is there a way to avoid it?

Your TSO session must have enough memory to keep all the dataset records for an edit; if there's not then browse will be substituted for edit. The solution is to increase the TSO session size; since this can impact various parts of the system it is not always possible for a programmer to make such a change. However, it is simply not possible to edit some data sets since they may be larger than can be read into memory.

However, it is simply not possible to edit some data sets since they may be larger than can be read into memory.

Then how do we edit huge datasets?

Then how do we edit huge datasets?

You don't. In fact, a lot of organizations have rules against modifying production data in an edit session. I worked with a customer in Canada, and they were not allowed to change any production data -- period.

This is what I have also observed that browse is substituted whenevr I tried to open a big data set. But now looking at this question again I think is browse like requires small storage than view or edit? I mean dataset is still same so how come they are treated differently?

I mean dataset is still same so how come they are treated differently?

Because edit requires you to be able to change records. Since browse is a read-only function, the system can load as many records into memory as it can and then set an internal marker for where to start reading the next set of records from the data set. When the system starts reading the next set of records, it can flush the entire memory space since nothing needs to be retained. When browsing a very large data set, you will probably notice that the time to move through the data set is extended; this is due to the system reading records into memory.