Release: All supported releases.
The set is CALC, but there is a record in the CALC set that should have targeted to a different page. The set is otherwise in tact from a next and prior pointer perspective.
This is an example of what IDMSDBAN output might look like in this situation. For each record occurrence that is on an incorrect page, there will be two pairs of 599702/599703 messages - four in all.
599703 - PRIOR LINK NOT FOUND SET=CALC START OF LINK 0/75163:18 OWNER IS AT 0/75159:0
CHAIN LENGTH 1 LINK POINTS TO 0/75163:29 PRIOR IN PFX 0/75163:4
599702 - NEXT LINK NOT FOUND SET=CALC START OF LINK 0/75163:18 OWNER IS AT 0/75159:0
CHAIN LENGTH 1 LINK POINTS TO 0/75163:29 PRIOR IN PFX 0/75163:4
599702 - NEXT LINK NOT FOUND SET=CALC START OF LINK 0/75163:4 OWNER IS AT 0/75163:0
CHAIN LENGTH 1 LINK POINTS TO 0/75163:18 PRIOR IN PFX 0/75163:3
599703 - PRIOR LINK NOT FOUND SET=CALC START OF LINK 0/75163:29 OWNER IS AT 0/75163:0
CHAIN LENGTH 1 LINK POINTS TO 0/75163:19 PRIOR IN PFX 0/75163:18
599702 - NEXT LINK NOT FOUND SET=CALC START OF LINK 0/75173:4 OWNER IS AT 0/75170:0
CHAIN LENGTH 1 LINK POINTS TO 0/75173:0 PRIOR IN PFX 0/75173:3
599703 - PRIOR LINK NOT FOUND SET=CALC START OF LINK 0/75173:4 OWNER IS AT 0/75170:0
CHAIN LENGTH 1 LINK POINTS TO 0/75173:0 PRIOR IN PFX 0/75173:3
599703 - PRIOR LINK NOT FOUND SET=CALC START OF LINK 0/75173:0 OWNER IS AT 0/75173:0
CHAIN LENGTH 1 LINK POINTS TO 0/75173:1 PRIOR IN PFX 0/75173:4
599702 - NEXT LINK NOT FOUND SET=CALC START OF LINK 0/75173:3 OWNER IS AT 0/75173:0
CHAIN LENGTH 1 LINK POINTS TO 0/75173:4 PRIOR IN PFX 0/75173:2
The above IDMSDBAN output is a result of the following conditions in the database.
Note that the incorrectly located record appears once in each of the group of four messages identifying it.
IDMSDBAN calculates the target page of CALC records, and if the result is not the same as the page on whose CALC chain the record resides, then it issues the above messages. Neither CALC set chain is really "broken".
In this case, if RHDCOPTF bit 93 is in use, an UNLOAD/RELOAD should be able to handle the corruption and fix the database as long as there are no correctly located versions of the bad records as described in point 3 above.