Understanding message CAT2402E and chaining within TLMS.

Products

• TLMS Tape Management

Issue/Introduction

Chaining threshold is met, but there are enough chaining records allocated. Why?

Resolution

Explanation of TLMS Chaining/AUX record processing:

Here is an explanation of the processing of AUX records and the way that TLMS uses Auxiliary records. In the VMF there are two sources of AUX records. The first is a contiguous block that remains after the VMF is initialized or restored. There is also a chain of AUX records that have been used previously, but have been released. The count on the backup is the total of these two AUX record types. When TLMS needs an AUX record, the reusable chain is checked first and if there are any records available, one (or more) are used from the reusable chain. If there are no records on the reusable chain, a record is taken from the block of contiguous AUX records. When a record is taken from the block, we check the threshold values and determine if we are getting near the end of the block of records and if the record being used exceeds the threshold value, the CAT2402E message is issued.

The records on the reusable chain will come and go during normal TLMS processing, as records are added and removed from the reusable record chain. The warning messages will be issued until records are returned to the reusable chain.  The records are returned to the reusable chain whenever a chain is broken. This is normally at open output of a volume or chain of volumes that are using AUX records, but it will also occur if the UPV .....,CHV=BRKCHN command is issued against a scratch volume. Do not use EXPDT=60001 as any AUX records on that chain will not be recovered until the next BACKUP and RESTORE (or REORG if you are not changing ranges or AUX records).

Let me add a little more to this discussion and a warning. As long as AUX records remain on the reusable chain, you will not see the CAT2402E message until it is exhausted. This is where you need to be careful. We have had some sites ignore the CAT2402E until they had gotten down to less than a hundred unused AUX records, broke a bunch of chains to put records on the reusable chain, and processed for months. Eventually, the reusable chain was exhausted and a job ran that creates many multi datasets and all AUX records were exhausted and they received the CAT2401E message within minutes of the first CAT2402E they had seen in months.

The CAT2402E should be the first indication that the situation needs to be monitored. After the message is received, the BACKUP report will give you the total number of AUX records in the VMF and should be watched. Depending upon the application cycle and other factors the number can easily increase some days and decrease others. The time to worry is if you have a significant decrease suddenly or if it is just steadily declining.