Capture the Spool ESFLOG into a dataset

Products

• Spool

Issue/Introduction

How can the Spool ESFLOG be put into a dataset?

Resolution

The ESFLOG is dynamically allocated by the main task as a JES SYSOUT dataset using specific CLASS and DEST attributes. Values are defined via the LOGCLAS and LOGDEST parameters in the product startup file. Since it contains all the Spool activity, it might be useful to have the ESFLOG kept on a dataset for future use. In the ESFPARM parameter file the minimal definitions are:

LOGCLAS=3
LOGDEST=ESFLOG

They can be adjusted to meet the actual environment and current settings. Follow these steps depending on your current configuration:

1.- If there is a dedicated class for ESFLOG use the following parameters:

  XFERSAPI=YES                                                             
  XFER=XFERCLAS                                                     
  XFERCLAS=3         <== where 3 is the dedicated JES output class for the ESFLOG

With this setting, all the SYSOUTs allocated in JES output class 3 will be processed by Spool using the XFER SAPI interface.

2. If the ESFLOG class is being shared with other JES output then use this set of definitions:

  XFERSAPI=YES                                                             
  XFERCLAS=3                                                               
  XFERDEST=DEST                                                            
  XFEROPT=6      <== Perform file selection on both destination and XFERCLAS at the same time.

In this case, *ALL* SYSOUTs in JES under output class 3 and DEST that is defined on a NODE statement will be processed by the Spool XFER SAPI interface.

3.- These are sample definitions for the NODE destination that uses a GDG dataset to store the ESFLOG via the DISK driver:

DEFNODE DISK1,DISK-01,CLASS=ALL,GROUP=1,
        ACQUIRE=WORK,RELEASE=NOWORK   
 
NODE ESFLOG,DISK1,PURGE=YES,CLASS=3,   
     TCPDRIV=DISK,SEP=0,              
     TCPHOST='DSNAME=your.GDG.base.dataset.name,MEMBER=+1,DISP=NEW'