Configure the spool data set structure according to your projected processing needs and anticipated workload so that CA Spool can run at an optimal performance level. It is generally best to have fewer, larger spool data sets because the spool data sets must be allocated in cylinders, and in one extent. CA Spool uses only the first extent of these data sets for spool space; any additional extents are ignored.
For maximum efficiency, we recommend the following:
Business Value:
Larger data sets improve the efficiency of CA Spool. When a large spool buffer pool is specified the I/O activity on the spool data sets is reduced thereby keeping the CA Spool service times at a minimum.
Additional Considerations:
Setting BUFSIZE=27998 reduces the I/O on the spool data sets since a maximum of eight blocks per track is used. Use this value unless your spool files are less than a few hundred records. This allows for 2 physical blocks/records per 3390 track.
Unless you are concerned about spool data set usage, keep the default SPOOLCMP=NO. If YES is specified, all spool files with a record length less than 251 bytes are compressed in the spool data sets.
Specify XEQBUFS=200 or higher; the CA Spool transformer interface will have files open for both input and output while the transform is taking place so if you expect to have 100 open spool files, you should double that amount.
Note that CA Spool can support up to 255 spool data sets which can be added to a running CA Spool system.
More Information:
For more information about configuring spool data sets, see the "Initialization" chapter in the CA Spool Customization Guide.
|
Copyright © 2011 CA.
All rights reserved.
|
|