AutoCollect Process › Setting Up the AutoCollect Environment › Selecting a MUF as a Source for Collection › AutoCollect and MUFPLEX (Including Shadow MUF)

AutoCollect and MUFPLEX (Including Shadow MUF)

In certain complex MUFPLEX environments, the MUFNAME parameter can be specified as an asterisk (*). This means that the MUFNAME is assigned at MUF startup from a list of values stored in the DBSIDPR module.

The Snapshot function still requires you to specify a MUFNAME parameter. In these cases, if the MUFNAME used by the MUF you want to target can change, submit a DBUTLTY Snapshot function for each of the possible MUFNAMEs. SNAPSHOTs that use a MUFNAME that is not active get a DBUTLTY return code error and do not affect the AutoCollect tables. Those that find an active MUFNAME produce the Snapshot rowset. You can use the DBUTLTY SET OPTION1=ON-ERROR-CONTINUE to bypass the snapshot functions that received the DBUTLTY error code.

The OUTNAME= parameter is provided in the Snapshot function to allow you to reassign the MUFNAME used in the Snapshot rowset when it is being generated. This allows you to select data from a MUF that may have multiple different MUFNAMEs, yet still build the rowsets with a consistent MUFNAME. For more information, see the OUTNAME parameter in the AUTOCOLL OPTION=SNAPSHOT function in the CA Datacom/DB DBUTLTY Reference Guide.

Example

Suppose you have a PRODMUF with a shadow MUF and you would typically migrate the workload from the active MUF to the shadow MUF once a week at midnight on Saturday.

After the shadow MUF is converted to the active MUF, you allow the new MUF to remain in place for the following week of processing. Then you restart a second MUF which becomes the shadow MUF for this active MUF. At the next weekly boundary, you repeat the process.

Now, you are interested in the collection of statistics from switchover on Sunday morning until 9 P.M. on Saturday.

The sequence of events is as follows:

8/05-8/12	Active MUF was PRODMF1A

8/12-8/19	Active MUF was PRODMF1B

8/19-8/26	Active MUF was PRODMF1A

8/26-9/02	Active MUF was PRODMF1B

To be able to capture these statistics as a single representation of the same source MUF within AutoCollect, the Snapshot function could be executed each Saturday for both possible MUF names with the OUTNAME being PRODMUF1.

The result would be as follows:

8/12 Snapshot from PRODMUF1A creates rowsets as PRODMUF1
8/12 Snapshot from PRODMUF1B fails with RC=16

8/19 Snapshot from PRODMUF1A fails with RC=16
8/19 Snapshot from PRODMUF1B creates rowsets as PRODMUF1

8/26 Snapshot from PRODMUF1A creates rowsets as PRODMUF1
8/26 Snapshot from PRODMUF1B fails with RC=16  

9/02 Snapshot from PRODMUF1A fails with RC=16  
9/02 Snapshot from PRODMUF1B creates rowsets as PRODMUF1  

The AutoCollect system would have 4 Snapshot rowsets, 1 for each week with the MUFNAME of PRODMUF1.

More complex MUFPLEX usage including full data sharing may require additional study and implementations using the OUTNAME parameter which is not included with this guide at this time.