HP XP7 Storage - Manual (Page 122)

When WWNN is specified, CESTPATH parameters are the following.
Description Parameter
MCU SAID (x’00 and port number) aabb
RCU SAID (x’00 and port number) ccdd
WWN is the unique number for the controller and is indicated as follows.
I/O control after failure—with CGROUP FREEZE/RUN
When a failure occurs in the P-VOL, you can stop host I/O and copy operations, and then restart
them when conditions improve using the TSO CGROUP command with the FREEZE and RUN options.
•
CGROUP FREEZE :
Stops host I/O to specified P-VOLs
Presents state change pending (SCP) with extended long busy status to host I/O requests.
This causes the host to queue I/Os for the P-VOLs. SCP is indicated until the CGROUP/RUN ◦
command is issued or until the SCP time expires. (Pairs registered in a consistency group
are not in the SCP state, even after the CGROUP (FREEZE) command is run.)
◦ Stops update copy operations to associated S-VOLs
Remote paths are blocked between the specified MCU and RCU to stop update copy
operations to the S-VOLs.
◦ Splits pairs with P-VOLs on the specified MCU CU. (Pairs registered in a consistency
group are not in the SCP state, even after the CGROUP (FREEZE) command is run.)
•
CGROUP RUN :
◦ Starts host I/O to P-VOLs again
- Presents a state change interrupt (SCI) to the hosts, so that they reissue the I/Os that
were waiting while the P-VOLs were in the SCP state.
- Changes the P-VOL fence level to Never (PPRC CRIT=NO), so that the split P-VOLs
accept host write I/O operations.
The CGROUP FREEZE/RUN command is supported for XP7 when it functions as the MCU. XP7
provides all required host reporting for CGROUP operations (for example, IEA494I with extended
long busy (ELB) state), which is a key component of operations in IBM’s Geographically Dispersed
Parallel Sysplex (GDPS) environment, which also uses this command.
122 Pair operations using PPRC