Home Page of Oakwood Computing Ltdhome contact us

about Oakwood Computing Ltd
new recruits and graduate training schemes
programming languages, report and file manipulation utilities
z/OS, OS/390, MVS, CICS, DB2, IMS, ISPF courses
UNIX, Linux, VSE courses
Oakwood Computing Ltd
1 Dornton Road
South Croydon
Surrey CR2 7DR
UK
Tel: +44 (0) 20 8686 7266
courses@oakwoodcomputing.co.uk
Updated 11 November, 2003

 

VSAM Performance and Tuning

Duration: 1 day

Audience: Programmers and Operations Support staff who need to optimise VSAM peformance

Prequisites: At least 6 months experience of using VSAM datasets


This course is a practical workshop which uses examples of live datasets; the documentation is provided for reference and supports explanations and discussions.

Objectives: Top Contents

On successful completion of this course each student will be able to:

* describe how VSAM loads, deletes, and inserts records
* explain how VSAM constructs its index for a KSDS
* explain how freespace is used
* given a list of processing requirements, define a VSAM cluster which will perform efficiently
* explain the meaning of the LISTCAT information
* given a LISTCAT of datasets, identify the datasets which are not performing well at run time
* given a LISTCAT, identify the datasets which make poor use of disk space
* write commands which create, load, and use Alternate Indexes
* explain how the overhead of updating alternate indexes can be minimised if the index is rarely used
* explain the potential effect of composite keys, and state how to detect this problem
* identify circumstances which produce dead CIs
* describe how VSAM uses buffer space
* identify which insert strategy is used by ACCESS RANDOM and ACCESS DYNAMIC
* use batch Local Shared Resources
* explain how SHAREOPTIONS control access to datasets
* given a set of requirements, select a control interval size which performs well in batch and online systems
* describe the action which could be taken if a dataset has a large number of index levels


Contents: Top Objectives

VSAM dataset structure
Control intervals; control areas; RBA; how VSAM stores records in an ESDS, RRDS, KSDS; how VSAM retrieves, deletes, and inserts records; control interval splits, control area splits, freespace; index structure, index CISZ, key compression.

Defining VSAM datasets
DEFINE CLUSTER command and its parameters; choosing the control interval size; implications for disk space; VSAM defaults for parameters.

Listing Catalog Information
Interpreting LISTCAT output; identifying poor definitions; identifying inefficient disk utilisation; identifying poor I/O usage at run time; hardware considerations.

Alternate Indexes
Function of an alternate index; how VSAM accesses records using alternate indexes, performance considerations; defining and loading an alternate index; defining a path; JCL and COBOL requirements for alternate indexes.

Performance and data integrity
Functions and performance implications of bufferspace; NSR, BLSR, SMB buffering; insert strategies used by ACCESS RANDOM and ACCESS DYNAMIC; using multi-volume data sets; record layout, key compression and potential problems; SHAREOPTIONS; expiration dates; Record Level Sharing; using DFSMS Transactional VSAM Services.

Top Objectives Contents