IBM Mainframe Forum

Hi!,

I am reading VSAM Demistified redbook, topic 'The loss of useful space in Data CA' (4.1.2 ); page no (204); In this, it is explained that:

In the Index set, each CI has only one record. Each record has a pointer to
the highest possible key in an index record in the next lower level, and a
pointer to the beginning of that index record.

reference figure 4-1

My doubt is how many levels will be present in the INDEX Set and on what it depends?

What it means is that the index portion of your KSDS has only one reference record for each CI withing the data portion. This reference is the value of the highest key contained within the data portion CI.

The actual CI within the index portion will contain many records, each record pointing to only one CI within the data portion.

You will need to read again the section on how an index works

Read pages 11 and 12 of VSAM Demystified which explain the sequence set and index set concepts. The number of levels in the index set depend upon how many record keys (that is, how many data component CI) are in the file. VSAM always keeps the top level of the index set in a single CI. If the number of data CI in the file grows to the point that a second high level index set CI is needed, VSAM creates it and then creates a new higher level index set CI that contains pointers to the two CI in what used to be the high level. So the number of levels will depend upon how many data CI there are, the key length, how good the compression is for the key, the index component CI size, and so forth.

I referred to the pages mentioned. It was of great help. Thank you Robert Sample and expat for the time!