
Table 32 State table for Continuous Access Synchronous vs. Continuous Access Asynchronous (continued)
Transfer data via
ESCON
Writing control on Cnt Ac-S async volumeDescriptionStorage system
internal state
RAID
Manager
state
ResponseWriting dataContinuous Access
Asynchronous
Continuous
Access
Synchronous
SidefileWait until under
threshold
Via SidefileOver
Threshold
SidefileUsualNormalDeleting from PAIR
using [pairsplit -S]
N/ADeleting
SidefileUsualVia BitmapUsing
[pairsplit ]
Suspending
from PAIR
N/ASuspending
Timeout of
over
threshold
NoneUsualVia BitmapSamePSUSPSUSPSUS
NoneUsualVia BitmapTimeout Over ThresholdNonePFUS
NoneUsualVia BitmapSame (link down, etc)PSUEPSUEPSUE
NoneUsualVia BitmapSamePDUBPDUBPDUB
* If the host has more write I/Os in COPY state, then host write I/Os are delayed until there is enough space in the sidefile.
Legend:
Bitmap: Host write data is managed via a cylinder BITMAP of delta data.
Normal: Host write data is not managed by BITMAP or sidefile.
Usual: Host side writing response is not delayed.
HWM (High Water Mark): Sidefile quantity is over 30% of cache storage.
Continuous Access Asynchronous/Continuous Access Journal error state
In the case of an ESCON or fibre-channel (FC) failure, the S-VOL FIFO queue is missing a data
block that was transferred from the P-VOL FIFO queue. The RCU waits to store the next sequenced
data block in the S-VOL FIFO queue until the Continuous Access Asynchronous copy pending
timeout occurs (defined using Continuous Access Synchronous). If the timeout occurs during this
waiting state, the pair status changes from PAIR to PSUE, and non-sequenced data blocks are
managed by the S-VOL bitmap. The missing data block can be recovered using the pairresync
command, which merges the S-VOL bitmap with the P-VOL bitmap. Figure 56 (page 142) shows
this situation on the secondary side.
Figure 56 Continuous Access Asynchronous suspension condition
142 Data replication operations with RAID Manager
Kommentare zu diesen Handbüchern