Not Using System Logger Service - HP XP P9500 Software User's Manual

To merge Business Continuity Manager log data that was generated by multiple operating systems
in a SIMPLEX environment, you must output the log for each operating system to an external dataset,
transfer the files to a computer, and then put the multiple files in chronological order.

Not Using System Logger Service

To obtain logs, use a DD statement to allocate two sequential log datasets to each job.
Secure capacity for log datasets are as follows:
The amount of capacity for log datasets is not fixed, and can be estimated based on the
frequencies at which CLI commands are issued or errors occur. At the very least, HP
recommends securing enough capacity for output until the time that the volume of a log dataset
output to one sequential dataset can be backed up. For details about the amount of log data,
see the HP StorageWorks P9000 for Business Continuity Manager Software Installation Guide.
For the log dataset, specify only the amount allocated at initialization. Any specified additional
allocation will be disregarded. As a result, if you need to specify additional allocation, HP
recommends that you use the system logger service.
The log dataset formats are as follows:
RECFM: VB
LRECL: 8196
BLKSIZE: 8200
Allocating a Log Dataset
Specify log datasets as follows:
For TSO/E, after logging on, use the ALLOC command to specify YKLOG01 DD and YKLOG02
DD statements. Alternatively, use a command procedure to specify YKLOG01 DD and YKLOG02
DD statements.
For a script, specify YKLOG01 DD and YKLOG02 DD statements for the JCL during script
execution.
The following is an example specification for a script or logon procedure:
//YKLOG01 DD DSN='USER01.YKLOG01',DISP=OLD,VOL=SER=WKVL01,UNIT=DASD
//YKLOG02 DD DSN='USER01.YKLOG02',DISP=OLD,VOL=SER=WKVL01,UNIT=DASD
The following is an example of TSO/E input that allocates the log dataset in advance:
ALLOC DS('USER01.YKLOG01') VOL(WKVL01) NEW SP(100) CYL
ALLOC DS('USER01.YKLOG02') VOL(WKVL01) NEW SP(100) CYL
Allocate a log dataset for each job. After the jobs terminates, the log dataset can be reused by
another job. In such a case, log output is appended to the end of the previous output.
In cases such as the following, reallocate a new log dataset instead of using the existing one:
If an I/O error occurs for the log dataset
If the OS crashes
Writing to a Log Dataset
If both YKLOG01 DD and YKLOG02 DD statements are specified, processing is performed as
follows:
1. Logs are written to YKLOG01. If the log dataset is reused, log output is appended to the
previous output.
2. If YKLOG01 becomes full, logs are written to YKLOG02.
3. If YKLOG02 becomes full, logs are written to YKLOG01. In this case, the previously existing
data in YKLOG01 is overwritten.
4. Steps 2 and 3 are repeated.
562 Troubleshooting