Data Extraction utility (TSSAFR1)


This topic explains how to use the Fast Path Recovery Utility Data Extraction utility (TSSAFR1), including its control statement, JCL, memory, and DASD requirements. Additionally, how you select recovery parameters and required data sets.

Overview of Data Extraction utility (TSSAFR1)

TSSAFR1 is an OS/VS batch program that accesses the Checkpoint-ID Table in the first record of the restart data set (RDS) and selects the lowest of the following values:

  • Current checkpoint (BCPCCKID)
  • LCRE checkpoint (BCPLCKID)
  • Fast Path checkpoint (BCPFCKID)

This approach is similar to the logic used by IMS during emergency restart processing.

You can specify the option RDS=NO and provide a Checkpoint-ID in the same format as on an /ERE command. Typically, this option would be chosen if the RDS is unusable. The correct Checkpoint-ID can be found in the DFS994I message on the IMS hardcopy log or the JES log of the failed IMS session. If you are unsure about specifying a particular checkpoint, specify an earlier checkpoint.

RECON data selection

TSSAFR1 dynamically allocates the three RECON data sets if the data sets are not specified in the JCL stream. BMC Software recommends dynamic allocation.

TSSAFR1 searches the library referenced by the STEPLIB DD statement, to locate the DFSMDA members, and obtains the data set name associated with each ddname (RECON1, RECON2, RECON3). RECON3 is optional.

The RECON data sets are allocated and opened (TSSAFR1 does not issue a Reserve against the pack on which they reside during this process). Empty RECON data sets (spare data sets) are discarded. A VSAM open error 160(A0) is encountered for the empty RECON data sets, and an IEC161I message is printed in the job log.

TSSAFR1 selects the RECON data set with the highest use count. If all use counts are equal, TSSAFR1 selects the last RECON data set that is opened. Unselected RECON data sets are closed and ignored.

Log data set selection

TSSAFR1 dynamically allocates the required log data sets.

To determine the required log data sets, TSSAFR1 performs several tasks:

  • TSSAFR1 references, in order, the PRI OLDS, IPRIOLDS, SECOLDS, and ISECOLDS records in the selected RECON data set.
  • TSSAFR1 selects a log data set if its Checkpoint-ID date and time stamp is equal to or greater than the log open date and time stamp value and less than the log close date and time stamp value.
  • If a selected log data set has errors flagged in the RECON entry or a log close date and time stamp value of zero, the program replaces the data set with a log data set from a subsequent RECON record type in the search hierarchy.

If DBRC=NO is specified, you must specify the SLDS or OLDS data sets through JCL statements. SLDS are processed before OLDS. Missing log data sets or duplicate log data results in log sequence errors.

Log records

The table below lists the IMS log records that the product uses to determine DEDB recovery information:

DEDB recovery log records

Log records

07, 08, 47, 3730, 3801, 4001, 4006, 4030, 4084, 4086, 4087, 4089, 5607, 5637, 5638, 5921, 5922, 5923, 5937, 5938, 5947, 5950, 5953, 5957

The following table lists the log records that the product uses to determine which processing scheduling tables (PSTs) were active and which full-function PSBs need to be backed out.

Active PSTs

Log records

18, 41, 4C01, 4C02, 5050

This section contains the following topics:

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*