HP StorageWorks XP Remote Web Console Software - Manual (Page 144)

144 Hitachi TrueCopy z/OS for HP XP12000/XP10000 and SVS 200 storage systems
4. When the MCU is fully powered on and ready to resume operations, resume the TC390A pairs at all
MCUs that were powered off (use the resume group option).
Planned outage of the RCU or remote copy connection
You must suspend all affected TC390 pairs prior to a planned outage of an RCU or of a remote copy
connection component (for example, switch, or extender). If you do not suspend the pairs first, the MCU(s)
will detect the communication failure, suspend all affected pairs, and generate SIMs and console
messages reporting the failures.
To perform a planned outage of a TC390 RCU or remote copy connection component:
1. Identify all TC390 M-VOLs that will be affected by the equipment outage. You need to know the MCU,
CU image, and LDEV ID for each of these M-VOLs.
• For RCU power-off, identify all M-VOLs that are paired with R-VOLs in the RCU to be powered off.
• For remote copy connection outage, identify all M-VOLs in all MCUs that use the path/component
to be powered off.
2. Connect to each MCU that contains affected M-VOLs and suspend all affected TC390 pairs. Make sure
to confirm the pair status changes (TC390 Pair Status window or CQUERY TSO command).
3. Perform the planned outage of the RCU or remote copy connection.
4. When the RCU is fully powered on and ready to resume operations, resume all TC390 pairs at each
MCU. Make sure to confirm the pair status changes.
Planned outage of the MCU and RCU
When you plan an outage of TC390 MCUs and RCUs at the same time, the MCUs must be powered off
before the RCUs and powered on after the RCUs.
To perform a planned outage of a TC390 MCU and RCU:
1. If RCU power-on will be difficult to control (for example, Power-Control-Interface setting), consider
increasing or disabling the RCU ready timeout group option (see ” Group options ” on page 33) for
each TC390A group with R-VOLs in the RCU(s) to be powered off.
2. Perform the planned outage of the MCU(s) as described in ” Planned outage of the MCU ” on
page 143. Do not power-on the MCU(s) yet.
3. If an RCU to be powered off is connected to an MCU that is not powered off, make sure to suspend
those TC390 pairs before powering off the RCU as described in ” Planned outage of the RCU or remote
copy connection ” on page 144.
4. Perform the planned outage of the RCU(s).
5. Power on the RCU(s). Verify that they are fully operational and ready to resume operations before
powering on the MCUs.
6. Power on the MCU(s) and verify that they are ready to resume operations. If you suspended any pairs
in step 3 , you can also resume those pairs now.
TrueCopy disaster recovery operations
Preparing for disaster recovery
The type of disaster and the status of the TC390 volume pairs will determine the best approach for disaster
recovery. For example, if all TC390 volume pairs are in the duplex state when a total system failure occurs
at a single point in time, the R-VOLs are current and recovery is straightforward. Unfortunately, some
disasters are not so “orderly” and involve intermittent or gradual failures occurring over a longer period of
time. The user should anticipate and plan for all types of failures and disasters.
The major steps in preparing for disaster recovery are:
1. Identify the volumes and volume groups that contain important files and data for disaster recovery, such
as DB2 log files, master catalog, key user catalogs, and system control data sets. In addition to
supporting TC390 remote copy operations as well as PPRC commands, the XP12000/XP10000
provides battery-backed nonvolatile duplexed cache, full hardware redundancy, dynamic sparing, and
an advanced RAID-5 implementation to ensure full data integrity in the event of a sudden power outage
or other failure.