Previous Topic: Determine if System Logger is Active

Next Topic: Logger Extract Utility TBELOGRX

External Logger Setup

The External Logger requires the creation of a System Logger Log Stream. IBM's Administrative Data Utility program IXCMIAPU is used for this purpose. Sample JCL is provided in member BESCLS of HLQ.CTAPJCL.

Note: For information about all the options available for this utility, see the IBM reference manual, Setting up a Sysplex.

Example: JCL for Creating a System Logger Data Set

This example shows JCL for creating a System Logger data set. In this example, the Log Stream will be called BES.sysid.LOG and is the name that must be supplied in the LogStream attribute in parmlib member DYNAMIC. This example instructs the system to create a DASD-only Log Stream which does not require the coupling facility.

//DEFINE   EXEC PGM=IXCMIAPU                                 
//SYSPRINT DD   SYSOUT=*                                     
//SYSIN    DD   *                                            
 DATA TYPE (LOGR)                                            
    DELETE LOGSTREAM NAME(BES.sysid.LOG)                         
    DEFINE LOGSTREAM NAME(BES.sysid.LOG) DASDONLY(YES) RETPD(4)  
           STG_SIZE(640) LS_SIZE(25600) HLQ(BES) AUTODELETE(YES)

Keywords and options used in this example are explained as follows:

RETPD(4)

Instructs the System Logger to retain the log data for 4 days.

AUTODELETE(YES)

Instructs the System Logger to automatically delete the log data after the number of days specified by the RETPD keyword.

The System Logger deletes the data on a log data set basis when the log data set is full, is no longer in use as the current log data set for the log stream, and is not allocated to any system in the sysplex the next time a data set switch occurs.

STG_SIZE(n)

Instructs the system to create a VSAM linear data set with a size of n*4096 bytes for staging logger records.

The example shows the stage size is 640. This means a staging data set will be allocated with a size of 640 * 4096 or 2,621,440 bytes.

DASDONLY

For DASDONLY Log Streams, staging data sets are required as part of the System Logger configuration. The System Logger automatically duplexes data to the staging data set for the system at the same time it writes the data to local storage buffers.

LS_SIZE(n)

Instructs the system to automatically create VSAM linear data sets of size n * 4096 when necessary to hold logger records as they are off-loaded from the staging library.

IBM recommends that you size the data sets as large as your installation can afford to make them. This will minimize the number of log data sets required to represent a Log Stream. It will also minimize the number of times the System Logger must reallocate and switch to using a new log data set when an old one becomes full. Because allocating and switching to a new log data set incurs overhead, it should be done as little as possible.

By default, each Log Stream is limited to a maximum of 168 log data sets unless you define data set directory extent records in the LOGR couple data set and make it the active primary LOGR couple data set. By defining data set directory extent records, a Log Stream is no longer limited to 168 log data sets.

The example shows the size is 25600. This means each logger data set will be a size of 25600 * 4096 or 100MB. This further implies the maximum amount of data that can be held concurrently in the Log Stream will be 100MB * 168 or 16.8GB.

HLQ(nnnnnnnn)

Instructs the system to define the staging and logger VSAM linear data sets using the nnnnnnnn high level qualifier in the data set name. We recommend that you specify the CA Tape Encryption PREFIX in this parameter.