Switching from cluster to noncluster mode (or vice versa) is possible only after normal termination. A starting nucleus checks in the PPT whether the previous session ended abnormally with a pending autorestart. If this is the case and the previous nucleus ran in the same mode as the starting nucleus, the session autorestart logic will be executed. If the previous nucleus ran in a different mode than the starting nucleus, the session start will terminate with an error.
The following sections illustrate a few scenarios where a cluster nucleus starts after the normal termination of a noncluster nucleus. PLOGRQ is not set to FORCE. These scenarios apply to two PLOGs as well as up to eight PLOGs.
The previous session was noncluster mode, there are remaining PLOGs to be copied, there is no UEX2/12 in use, and the PLOG datasets are different from what was used in the previous session. The results of this scenario are as follows:
The information in the PPT entry of the noncluster nucleus remains, and the new entry of the cluster nucleus is written.
Initialization continues.
The previous session was noncluster mode, there are remaining PLOGs to be copied, there is no UEX2/12 in use, and the PLOG datasets are the same as what was used in the previous session of a noncluster nucleus. The results of this scenario are as follows:
A warning that the PLOG is being overwritten will occur and the PLOG flag in the previously used PPT block will be reset or the PPT entry will be overwritten (whichever is appropriate).
Initialization continues.
The previous session was noncluster mode, there are remaining PLOGs to be copied, UEX2/12 is in use, and the PLOG datasets are different from what was used in the previous session of a noncluster nucleus. The results of this scenario are as follows:
UEX2/12 is called to submit a PLCOPY job that will copy and merge the PLOGs. The information in the PPT entry of the noncluster nucleus remains, and the new entry of the cluster nucleus is written.
Initialization continues.
Note:
If ADARES detects that there is data to be copied both from a cluster
nuclei and from a noncluster nucleus (different PLOGs), it will copy the oldest
data first.
