SYSMOD entries are merged when either CONTENT or FORFMID is specified
on the GZONEMERGE command. If CONTENT is specified, SMP/E will consider
all SYSMOD entries in the original global zone as candidates for merging.
If FORFMID is specified, SMP/E will consider only the SYSMOD entries
associated with the selected FMIDs for merging.
SMP/E reads through the original zone, gathering data on the candidate
SYSMOD and HOLDDATA entries. For each candidate entry in the originating
zone, SMP/E checks to see whether that entry exists in the destination
zone.
- If a SYSMOD entry is not found, then the entry from the originating
global zone is stored in the destination global zone.
- If a SYSMOD entry is found, processing continues
to determine the higher level SYSMOD. If the originating global zone
contains a higher or equal level of the SYSMOD than the destination
global zone, then the entry from the originating global zone replaces
the existing SYSMOD entry in the destination global zone. Any existing
HOLDFIXCAT (or HOLDERR, HOLDSYS, and HOLDUSER) subentries from the
destination global zone will be retained. If the destination global
zone contains a higher level of the SYSMOD than the originating global
zone, then processing continues to the next entry.
SMP/E uses
the REWORK subentry of the global zone SYSMOD entry to determine if
a SYSMOD in the originating global zone is at a higher level than
the same-named SYSMOD in the destination zone. If the REWORK date
of the SYSMOD in the originating global zone is more recent than the
REWORK date of the SYSMOD in the destination global zone, then that
SYSMOD is considered to be at a higher level in the originating global
zone and would be merged.
- If a SYSMOD entry in the originating global zone has a SOURCEID
subentry that matches the name of the ORDER entry from which it originated,
and that ORDER entry is being renamed because a like-named entry already
exists in the destination global zone, then SMP/E ensures that the
SOURCEID value matches the new name for the ORDER entry. To accomplish
this it is necessary to process DEFINITION entries before CONTENT
entries.