IBM DS6000 Manual Do Utilizador

IBM System Storage DS6000 Series: Copy Services with IBM System z

The TPC for Replication server listens for incidents from the storage servers and takes action 
when being notified of a replication error from the concerned storage server. 

Figure 31-10 implies a replication error in an LSS that belongs to the session. The TPC 
server receives a corresponding SNMP trap message from the concerned storage server. 
The TPC for Replication server then issues a freeze command to all LSSs that are part of the 
concerned session. This implies a suspend of all PPRC pairs or Copy Sets that belong to this 
session. During this freeze process, write I/Os are held until the freeze process ends and the 
TPC server communicates to the storage server to continue processing write I/O requests to 
the concerned primary volumes in the session. 

After that, write I/O can continue to the suspended primary volumes. However, both sites are 
not in sync any longer but the data on the secondary site is consistent (power drop 
consistent). This is a so-called freeze-and-go policy.

Because the connectivity between the TPC for Replication server and the storage servers 
that the TPC server is managing can fail, the firmware in the storage server waist for a 
heartbeat signal from the TPC server. TPC for Replication can enable this heartbeat in the 
corresponding LSS for Metro Mirror sessions.

  Replication Manager Server  

 Session 

1

3

2