IBM Flex System V7000 Storage Node 4939H49 ユーザーズマニュアル
製品コード
4939H49
8068ch09-Copy Servies.fm
Draft Document for Review January 29, 2013 12:52 pm
344
IBM Flex System V7000 Storage Node Introduction and Implementation Guide
volume. Additionally, the source and target volumes must belong to the same IBM Flex
System V7000 Storage Node, but they do not have to reside in the same storage pool.
System V7000 Storage Node, but they do not have to reside in the same storage pool.
Before you start a FlashCopy (regardless of the type and options specified) you must run
either
either
prestartfcmap
or
prestartfcconsistgrp
,
which puts the con troll canister Cache into
write-through mode, which results in the flushing of all I/O currently bound for your source
volume. After FlashCopy is started, an effective copy of a source volume to a target volume
has been created. The content of the source volume is immediately presented on the target
volume and the original content of the target volume is lost. This FlashCopy operation is also
referred to as a
volume. After FlashCopy is started, an effective copy of a source volume to a target volume
has been created. The content of the source volume is immediately presented on the target
volume and the original content of the target volume is lost. This FlashCopy operation is also
referred to as a
time-zero copy
(T
0
).
Immediately following the FlashCopy operation, both the source and target volumes are
available for use. The FlashCopy operation creates a bitmap that is referenced and
maintained to direct I/O requests within the source and target relationship. This bitmap is
updated to reflect the active block locations as data is copied in the background from the
source to target and updates are made to the source.
available for use. The FlashCopy operation creates a bitmap that is referenced and
maintained to direct I/O requests within the source and target relationship. This bitmap is
updated to reflect the active block locations as data is copied in the background from the
source to target and updates are made to the source.
Figure 9-1 shows the general process for how FlashCopy works while the full image copy is
being completed in the background. Also the handling of the redirection of the host I/O which
is being written to the source volume with respect to a T
being completed in the background. Also the handling of the redirection of the host I/O which
is being written to the source volume with respect to a T
0
point in time while the target volume
is held true to T
0
.
Figure 9-1 Flashcopy processing and redirection of host write I/O
When data is copied between volumes, it is copied in units of address space known as
grains
. Grains are units of data grouped together to optimize the use of the bitmap that track
changes to the data between the source and target volume. You have the option of using 64
KB or 256 KB grain sizes; 256 KB is the default. The FlashCopy bitmap contains 1 bit for each
grain and is used to track whether the source grain has been copied to the target. The 64 KB
grain size consumes bitmap space at a rate of four times the default 256 KB size.
KB or 256 KB grain sizes; 256 KB is the default. The FlashCopy bitmap contains 1 bit for each
grain and is used to track whether the source grain has been copied to the target. The 64 KB
grain size consumes bitmap space at a rate of four times the default 256 KB size.
The FlashCopy bitmap dictates read and write behavior for both the source and target
volumes as follows:
volumes as follows:
��
Read I/O request to source: Reads are performed from the source volume, which is the
same as for non-FlashCopy volumes.
same as for non-FlashCopy volumes.
��
Write I/O request to source: Writes to the source cause the grains to be copied to the
target if it has not already been copied, and then the write is performed to the source.
target if it has not already been copied, and then the write is performed to the source.