Computer Drive User Manual

190 IBM System Storage DS6000 Series: Copy Services with IBM System z
recorded on the B volumes back to the A volumes. You must vary the A volumes offline, if
they are online at the primary site.
4. If you established a FlashCopy relationship (to the C volumes), you can now withdraw it.
5. Check that Metro Mirror paths exist between the A and B volumes. Reestablish, if
necessary.
6. Allow Metro Mirror pairs to reach full Duplex state (use the CQUERY command, or check
the syslog for IEA494I messages).
7. Issue Metro Mirror establish pair commands to A volumes specifying the
ACTION(FAILOVER) with the original PRIM and SEC parameters. This reestablishes the
A volumes as suspended primaries and begins change recording on the A volumes. It also
releases write inhibit on the A volumes. You can vary the A volumes online again at this
point.
8. You can now restart your applications at the primary site.
9. Now issue the Metro Mirror Establish pair command to the A volumes, specifying the
ACTION(FAILBACK) with the original PRIM and SEC parameters. This resynchronizes
the A and B volumes in the recovery configuration and restores the pairs to the Duplex
state.
CESTPAIR with FAILBACK action (step 3, 9)
For step 3, the command shown in Example 16-5 will reestablish the Metro Mirror pair and
copy any changes to the original primary volume (remember to keep the original PRIM and
SEC parameters reversed).
Example 16-5 CESTPAIR with Failback - B to A (Step 3)
CESTPAIR DEVN(X'6430') PRIM(X'0002' AAVCA X'30' X'00') -
SEC(X'2060' AAGXA X'30' X'00') -
ACTION(FAILBACK)
A similar command is issued in step 9; however, it has the original Metro Mirror direction as
shown in Example 16-6.
Example 16-6 CESTPAIR with Failback - A to B (Step 9)
CESTPAIR DEVN(X'6430') SEC(X'0002' AAVCA X'30' X'00') -
PRIM(X'2060' AAGXA X'30' X'00') -
ACTION(FAILBACK)
Syslog messages for failback process
Example 16-7 shows the syslog messages issued for the complete failback process. Note
that we had access to both devices from one LPAR and therefore could perform all operations
from there. You might have to split operations if you do not have access to both devices.
We issued CQUERY commands between each step of the process to check the status of our
volumes. You can see the status changes and the direction of mirroring change as we
progressed through the process.
Example 16-7 Syslog messages for failback process
ANTP8802I CQUERY DEVN(6030)
ANTP0090I CQUERY FORMATTED LVL 3 257
VOLUME REPORT
************** PPRC REMOTE COPY CQUERY - VOLUME ********************
* (PRIMARY) (SECONDARY) *
* SSID CCA LSS SSID CCA LSS*