U.S. patent application number 13/912297 was filed with the patent office on 2013-10-17 for storage system and operation method of storage system.
This patent application is currently assigned to HITACHI, LTD.. The applicant listed for this patent is Masataka Innan, Akira Murotani, Akinobu Shimada. Invention is credited to Masataka Innan, Akira Murotani, Akinobu Shimada.
Application Number | 20130275690 13/912297 |
Document ID | / |
Family ID | 36450094 |
Filed Date | 2013-10-17 |
United States Patent
Application |
20130275690 |
Kind Code |
A1 |
Innan; Masataka ; et
al. |
October 17, 2013 |
STORAGE SYSTEM AND OPERATION METHOD OF STORAGE SYSTEM
Abstract
The present invention is able to improve the processing
performance of a storage system by respectively virtualizing the
external volumes and enabling the shared use of such external
volumes by a plurality of available virtualization storage devices.
By virtualizing and incorporating the external volume of an
external storage device, a first virtualization storage device is
able to provide the volume to a host as though it is an internal
volume. When the load of the first virtualization storage device
increases, a second virtualization storage device 2 is newly
introduced, and connected to the storage system. When a transfer
direction is issued from a management terminal, the external volume
relating to the selected logical volume is transferred from the
first virtualization storage device to the second virtualization
storage device.
Inventors: |
Innan; Masataka; (Odawara,
JP) ; Murotani; Akira; (Odawara, JP) ;
Shimada; Akinobu; (Chigasaki, JP) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Innan; Masataka
Murotani; Akira
Shimada; Akinobu |
Odawara
Odawara
Chigasaki |
|
JP
JP
JP |
|
|
Assignee: |
HITACHI, LTD.
Tokyo
JP
|
Family ID: |
36450094 |
Appl. No.: |
13/912297 |
Filed: |
June 7, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13222569 |
Aug 31, 2011 |
8484425 |
|
|
13912297 |
|
|
|
|
12367706 |
Feb 9, 2009 |
8180979 |
|
|
13222569 |
|
|
|
|
11181877 |
Jul 15, 2005 |
|
|
|
12367706 |
|
|
|
|
Current U.S.
Class: |
711/154 |
Current CPC
Class: |
G06F 3/067 20130101;
G06F 3/061 20130101; G06F 3/0635 20130101; G06F 3/0637 20130101;
G06F 3/0605 20130101; G06F 3/0647 20130101; G06F 3/0665
20130101 |
Class at
Publication: |
711/154 |
International
Class: |
G06F 3/06 20060101
G06F003/06 |
Foreign Application Data
Date |
Code |
Application Number |
May 24, 2005 |
JP |
2005-150868 |
Claims
1. (canceled)
2. A storage system comprising: at least one physical storage
devices providing a logical volume; a first virtualization device
coupled to a computer and being configured to provide a first
logical volume with a first identification number to the computer,
wherein the first logical volume includes a first virtual storage
area mapped to all or a part of a storage area of the logical
volume, and a second virtualization device coupled to the computer
and a first virtualization device, wherein the second
virtualization device is configured to, in response to a request
from the host: obtain information of the first logical volume sent
from the first virtualization device, and provide a second logical
volume to the computer, wherein the second logical volume includes
a second virtual storage area mapped to the all or the part of the
storage area, which is also mapped to the first virtual storage
area in the first logical volume, of the logical volume, wherein
the second logical volume is provided to the computer with the
first identification number included in the information obtained
from the first virtualization device, but a port number of the
second logical volume provided to the computer is different from a
port number of the first logical volume provided to the
computer.
3. A storage system according to claim 2, wherein the first logical
volume and the second logical volume are volumes of access targets
by the computer.
4. A storage system according to claim 2, wherein each of the first
and second identification numbers is a logical unit number for
identifying volumes of access targets by the computer.
5. A storage system according to claim 2, wherein the computer
includes a plurality of host bus adapters, wherein the first
virtualization device includes the first port, wherein the second
virtualization device includes the second port, wherein the first
path is identified by the combination of a first host bus adapter
number of identifying at least one of the host bus adapters, the
first port number, and the first identification number for
identifying the first logical volume, and wherein the second path
is identified by the combination of a second host bus adapter
number for identifying at least one of the host bus adapters, the
second port number, and the first identification number for
identifying the first logical volume.
6. A storage system according to claim 2, wherein the computer is
configured to switch the first path to the second path in case that
a request to the first logical volume through the first path is not
processed by the first virtualization device.
7. A storage system according to claim 2, wherein information
regarding an owner right, which indicates authority to use the
logical volume, can be transferred by the first virtualization
device to the second virtualization device, while both the first
path and the second path are defined.
8. A storage system according to claim 2, wherein the storage
system further includes an external storage system coupled to the
first virtualization device and the second virtualization system
through a storage area network, and wherein the logical volume is
in the external storage device.
9. A storage system according to claim 2, wherein the first
virtualization device holds first mapping information indicating
mapping between the first logical volume and the all or the part of
the storage area in the logical volume, and wherein the second
virtualization device holds second mapping information indicating
mapping between the second logical volume and the all or the part
of the storage area in the logical volume.
10. In a storage system comprising at least one physical storage
devices providing a logical volume; a first virtualization device
coupled to a computer and being configured to provide a first
logical volume with a first identification number to the computer,
wherein the first logical volume includes a first virtual storage
area mapped to all or a part of a storage area of the logical
volume, and a second virtualization device coupled to the computer
and a first virtualization device, wherein the second
virtualization device is configured to perform a method, in
response to a request from the host, the method comprising:
obtaining information of the first logical volume sent from the
first virtualization device, providing a second logical volume to
the computer, wherein the second logical volume includes a second
virtual storage area mapped to the all or the part of the storage
area, which is also mapped to the first virtual storage area in the
first logical volume, of the logical volume, and providing the
second logical volume to the computer with the first identification
number included in the information obtained from the first
virtualization device, but a port number of the second logical
volume provided to the computer is different from a port number of
the first logical volume provided to the computer.
11. A method according to claim 10, wherein the first logical
volume and the second logical volume are volumes of access targets
by the computer.
12. A method according to claim 10, wherein each of the first and
second identification numbers is a logical unit number for
identifying volumes of access targets by the computer.
13. A method according to claim 10, wherein the computer includes a
plurality of host bus adapters, wherein the first virtualization
device includes the first port, wherein the second virtualization
device includes the second port, wherein the first path is
identified by the combination of a first host bus adapter number of
identifying at least one of the host bus adapters, the first port
number, and the first identification number for identifying the
first logical volume, and wherein the second path is identified by
the combination of a second host bus adapter number for identifying
at least one of the host bus adapters, the second port number, and
the first identification number for identifying the first logical
volume.
14. A method according to claim 10, wherein the computer is
configured to switch the first path to the second path in case that
a request to the first logical volume through the first path is not
processed by the first virtualization device.
15. A method according to claim 10, wherein information regarding
an owner right, which indicates authority to use the logical
volume, can be transferred by the first virtualization device to
the second virtualization device, while both the first path and the
second path are defined.
16. A method according to claim 10, wherein the storage system
further includes an external storage system coupled to the first
virtualization device and the second virtualization system through
a storage area network, and wherein the logical volume is in the
external storage device.
17. A method according to claim 10, wherein the first
virtualization device holds first mapping information indicating
mapping between the first logical volume and the all or the part of
the storage area in the logical volume, and wherein the second
virtualization device holds second mapping information indicating
mapping between the second logical volume and the all or the part
of the storage area in the logical volume.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation application of U.S. Ser.
No. 13/222,569, filed Aug. 31, 2011 which is a continuation
application of U.S. Ser. No. 12/367, 706, filed Feb. 9, 2009, which
is a continuation application of U.S. Ser. No. 11/181,877, filed
Jul. 15, 2005 (now abandoned), which relates to and claims priority
from Japanese Patent Application No. 2005-150868 filed on May 24,
2005, the entire disclosures of all of which are incorporated
herein by reference.
BACKGROUND OF THE INVENTION
[0002] The present invention relates to a storage system and an
operation method of a storage system.
[0003] For instance, government agencies, companies, educational
institutions and others manage data with a relative large storage
system for handling various data in large quantities. This storage
system, for example, is configured by including a storage device
such as a disk array device. For instance, a storage device is
configured by disposing a plurality of memory apparatuses in an
array to provide a memory area based on RAID (Redundant Array of
Inexpensive Disks). At least one or more logical volumes are formed
on a physical memory area provided by the memory apparatus group,
and this logical volume is provided to a host computer (hereinafter
abbreviated as "host"). By transmitting a write command or read
command, the host is able to write and read data into and from the
logical volume.
[0004] Data to be managed by companies and others is increasing
daily. Thus, companies and others, for example, equip the storage
system with a new storage device to expand the storage system. Two
methods can be considered for introducing a new storage device to
the storage system. One method is to replace the old storage device
with a new storage device. Another method is to make the old
storage device and new storage device coexist.
[0005] Nevertheless, when making a full transition from the old
storage device to a new storage device, the old storage device
cannot be utilized. Meanwhile, when making the old storage device
and new storage device coexist, the configuration of the storage
system will become complex, and the management and operation
thereof will become extremely troublesome.
[0006] Thus, the present applicant has proposed technology of
connecting a host and a first storage device and connecting the
first storage device and a second storage device so that the first
storage device will act over and process the access request from
the host (Japanese Patent Laid-Open Publication No. 2004-005370).
With this technology, the first storage device will also receive
and process commands targeting the second storage device. If
necessary, the first storage device issues a command to the second
storage device, receives the processing result thereof, and
transmits this to the host.
[0007] With the conventional technology described in the foregoing
document, the performance of the storage system is improved by
making the first storage device and second storage device coexist
without wasting any memory resource. Nevertheless, even with this
kind of reinforced storage system, the processing performance may
deteriorate during the prolonged operation thereof.
[0008] For example, if the number of hosts connected to the first
storage device increases, since numerous access requests will be
issued from the respective hosts, the processing performance of the
storage system will most likely deteriorate. Further, data to be
managed will increase daily, and the method of use and frequency of
use will differ diversely according to the nature of the respective
data.
[0009] Thus, further reinforcement of the storage system is
desired. In such a case, the first storage device may be replaced
with a different high-performance storage device, or a separate
first storage device may be added to the existing first storage
device. Nevertheless, the addition or replacement of the first
storage device cannot be conducted as with the addition of the
first storage device described in the foregoing document. This is
because the first storage device is serially connected to the
second storage device and uses the memory resource of the second
storage device, and the configuration of the storage system is
already complicated. The first storage device cannot be simply
added or replaced by only focusing attention on the first storage
device.
SUMMARY OF THE INVENTION
[0010] The present invention was devised in view of the foregoing
problems, and an object of the present invention is to provide a
storage system and an operation method of a storage system
configured by hierarchizing a plurality of storage devices for
improving the processing performance thereof relatively easily.
Another object of the present invention is to provide a storage
system and an operation method of a storage system for improving
the processing performance by enabling the shared use of one or a
plurality of connection destination storage devices by a plurality
of connection source storage devices. Other objects of the present
invention will become clear from the detailed description of the
preferred embodiments described later.
[0011] In order to achieve the foregoing objects, the storage
system according to the present invention has a plurality of
connection source storage devices capable of respectively providing
a logical volume to a host device; a connection destination storage
device respectively connected to each of the connection source
storage devices and having a separate logical volume; and a
direction unit for directing the connection destination of the
separate logical volume. And each of the connection source storage
devices is configured by respectively having: a management
information memory unit for storing management information for
managing the separate logical volume; and a control unit for
connecting the logical volume and the separate logical volume via
an intermediate volume based on the management information stored
in the management information memory unit; wherein the connection
destination of the separate logical volume can be switched among
each of the connection source storage devices based on the
designation from the direction unit.
[0012] The logical volume of the connection source storage device
can be connected to a separate logical volume of the connection
destination storage device via an intermediate volume. This
connection may be made based on the management information stored
in the management information memory unit.
[0013] Here, when focusing on the connection source storage device,
the connection destination storage device is an external storage
device positioned outside the connection source storage device, and
the separate logical volume of the connection destination storage
device is an external volume positioned outside the connection
source storage device. Therefore, in the following explanation, for
ease of under the present invention, the connection destination
storage device may be referred to as an external storage device,
and the separate logical volume may be referred to as an external
volume, respectively.
[0014] The host device issues a read command, write command and so
on with the logical volume of the connection source storage device
as the access target. When the connection source storage device
receives an access request from the host device, it issues a
prescribed command to the external volume connected to the logical
volume of the access target, and reads and writes data from and
into the external volume. As described above, the logical volume of
the connection source storage device is an access destination
volume to become the access target from the host device, and the
external volume (separate logical volume) of the external storage
device is the data storage destination volume for actually storing
the data. The host device is not able to directly recognize the
external volume, and the external volume is transparent to the host
device.
[0015] The direction unit designates to which logical volume of the
connection source storage device the external volume should be
connected. Based on this designation, the connection designation of
the external volume is switched among the respective connection
source storage devices. In other words, when an external volume is
connected to a logical volume of one connection source storage
device via an intermediate volume, when the direction unit
designates the switch to the other connection source storage
device, the external volume is connected to a logical volume of the
other connection source storage device via an intermediate
volume.
[0016] Thereby, a plurality of connection source storage devices
may exclusively use one or a plurality of external volumes.
Accordingly, for example, when there are numerous access requests
to a specific external volume, such high-load external volume is
transferred to a separate connection source storage device in order
to disperse the load, and the processing performance of the overall
storage system can be improved thereby.
[0017] In the present embodiment, the connection destination of the
separate logical volume is switchable among each of the connection
source storage devices without stopping the access from the host
device to the logical volume.
[0018] In the present embodiment, the access destination of the
host device is switched among each of the connection source storage
devices according to the switching of the connection destination of
the separate logical volume. In other words, when the connection
destination of the external volume is switched from one connection
source storage device to the other connection source storage
device, the access destination of the host device will also be
switched from one connection source storage device to the other
connection source storage device.
[0019] In the present embodiment, the management information is
constituted by including first management information for
specifying the separate logical volume, and second management
information for managing the attribute of the separate logical
volume; the first management information is retained by each of the
connection source storage devices; and the second management
information is retained by the connection source storage device of
the switching destination selected as the connection destination of
the separate logical volume.
[0020] In other words, the management information for managing the
separate logical volume has first management information and second
management information, and the first management information is
stored in each of the connection source storage devices, and the
second management information is stored in the connection source
storage device requiring such second management information.
[0021] In the present embodiment, the first management information
contains volume identifying information for specifying the separate
logical volume in the storage system, usage authorization
information for specifying the connection source storage device
having usage authorization of the separate logical volume, and
switching status information for showing whether the connection
destination of the separate logical volume is being switched among
each of the connection source storage devices; and the second
management information contains a plurality of pieces of other
attribute information relating to the separate logical volume.
[0022] In the present embodiment, the usage authorization
information is set with the connection source storage device that
becomes the switching source among each of the connection source
storage devices, notified from the connection source storage device
that becomes the switching source to the connection source storage
device that becomes the switching destination, and the change of
the usage authorization information is determined by the connection
source storage device that becomes the switching source receiving
the setting completion report from the connection source storage
device that becomes the switching destination.
[0023] In the present embodiment, a switching status flag is set
while the connection destination of the separate logical volume is
being switched from the connection source storage device that
becomes the switching source to the connection source storage
device that becomes the switching destination, and the switching
status flag is reset when the connection destination of the
separate logical volume is switched; while the switching status
flag is being set, the connection source storage device that
becomes the switching source destages unwritten data relating to
the separate logical volume, and the connection source storage
device that becomes the switching destination processes write data
from the host device with an asynchronous system; and when the
switching status flag is reset, the switching destination storage
device destages the write data.
[0024] Here, an asynchronous transfer mode is a mode for, in the
case of writing data in a logical volume, reporting the completion
of writing to the host device before writing such data in a
physical memory apparatus. Contrarily, a synchronous transfer mode
is a mode for, in the case of writing data in a logical volume,
reporting the completion of writing to the host device after
confirming that such data has been written in a physical memory
apparatus.
[0025] In the present embodiment, the connection source storage
device that becomes the switching source among each of the
connection source storage devices rejects the processing of access
from the host device to the separate logical volume, and destages
unwritten data relating to the separate logical volume.
[0026] In other words, the connection source storage device that
becomes the switching source, among the access requests from the
host device, rejects the access request relating to the external
volume to be transferred to the connection source storage device
that becomes the switching destination. A rejection may be made
positively or negatively. And, the connection source storage device
that becomes the switching source destages unwritten data relating
to such external volume to be transferred. As a result, the
consistency of the data stored in such external volume can be
maintained.
[0027] In the present embodiment, when the destage is complete, the
connection source storage device that becomes the switching source
issues a destage completion report to the connection source storage
device that becomes the switching destination; and upon receiving
the destage completion report, the connection source storage device
that becomes the switching destination performs the processing of
access from the host device to the separate logical volume.
[0028] In other words, the dirty data before transfer (before
switching) is written in a physical memory apparatus configuring
the external volume of the transfer target to maintain the
consistency of data.
[0029] In the present embodiment, a monitoring unit is further
provided for monitoring the load status relating to at least the
connection source storage device that becomes the switching source
among each of the connection source storage devices.
[0030] And, the connection source storage device that becomes the
switching source and the connection source storage device that
becomes the switching destination among each of the connection
source storage devices are respectively selected based on the
monitoring result of the monitoring unit.
[0031] As the load status, for instance, input/output per second
(IOPS), CPU usage rate, cache memory usage rate, data traffic and
so on may be considered. For example, when there is a logical
volume where the load status became higher and a prescribed
threshold value, the external volume to which such logical volume
is connected is transferred to a separate connection source storage
device. Thereby, the load of the connection source storage device
of the switching source can be reduced.
[0032] In the present embodiment, a management terminal to be
connected to each of the connection source storage devices is
further provided, wherein the direction unit and the monitoring
unit are respectively provided to the management terminal.
[0033] The storage system according to another perspective of the
present invention has a plurality of connection source storage
devices to be used by at least one or more host devices, and at
least one or more connection destination storage devices to be
connected to each of the connection source storage devices, wherein
the host device and each of the connection source storage devices
are respectively connected via a first communication network, and
each of the connection source storage devices and the connection
destination storage device are connected via a second communication
network separated from the first communication network
[0034] Further, the connection destination storage device has a
separate logical volume to be logically connected to a logical
volume of each of the connection source storage devices. And, each
of the connection source storage devices has a control unit for
creating the logical volume and connecting the logical volume and
the separate logical volume via an intermediate volume based on
management information; and a memory used by the control unit and
for storing the management information.
[0035] Moreover, the management terminal to be connected to each of
the connection source storage devices has a monitoring unit for
respectively monitoring the load status of each of the connection
source storage devices, and a direction unit for respectively
selecting the connection source storage device that becomes the
switching source and the connection source storage device that
becomes the switching destination among each of the connection
source storage devices based on the monitoring result of the
monitoring unit.
[0036] In addition, the management terminal switches the connection
destination of the separate logical volume from the connection
source storage device selected as the switching source to the
connection source storage device selected as the switching
destination based on the designation from the direction unit;
[0037] Further, the management information is constituted by
including first management information for specifying the separate
logical volume, and second management information for managing the
attribute of the separate logical volume, and the first management
information is respectively stored in the connection source storage
device selected as the switching source and the connection source
storage device selected as the switching destination.
[0038] The entirety of the second management information is stored
in the connection source storage device selected as the switching
source, and only the second management information relating to the
separate logical volume in which the connection destination is
switched is transferred from the connection source storage device
selected as the switching source to the connection source storage
device selected as the switching destination.
[0039] The operation method of a storage system according to yet a
different perspective of the present invention is a method of
operating a storage system having a first connection source storage
device and a second connection source storage device capable of
respectively providing a logical volume to a host device via a
first communication network, and a connection destination storage
device connected to each of the first and second connection source
storage device via a second communication network, comprising the
following steps.
[0040] In the initial operation step, the plurality of separate
logical volumes are respectively connected to one or a plurality of
logical volumes of the first connection source storage device via
an intermediate volume of the first connection source storage
device based on the management information for respectively
connecting to a plurality of separate logical volumes of the
connection destination storage device, and the first connection
source storage device is made to process the access request from
the host device.
[0041] In the device addition step, the second connection source
storage device is connected to the host device via the first
communication network, to the connection destination storage device
via the second communication network, and to the first connection
source storage device via a third communication network.
[0042] In the first management information transfer step,
information for respectively specifying the plurality of separate
logical volumes among the management information of the first
connection source storage device is transferred from the first
connection source storage device to the second connection source
storage device via the third communication network.
[0043] In the transfer target selection step, a separate logical
volume is selected to be transferred to the second connection
source storage device among the plurality of separate logical
volumes used by the first connection source storage device.
[0044] In the second management information transfer step,
attribute information relating to the separate logical volume
selected as the transfer target among the management information of
the first connection source storage device is transferred from the
first connection source storage device to the second connection
source storage device via the third communication network.
[0045] In the additional operation step, the separate logical
volume selected as the transfer target is connected to the logical
volume of the second connection source storage device via an
intermediate volume of the second connection source storage device
based on the information acquired at the first management
information transfer step and the second management information
transfer step, the path information for the host device to access
the logical volume of the second connection source storage device
is set in the host device, and the second connection source storage
device is made to process the access request from the host
device.
[0046] Incidentally, the third communication network may also be
used in combination with either the first communication network or
second communication network.
[0047] The whole or a part of the means, functions and steps of the
present invention may sometimes be configured as a computer program
to be executed with a computer system. When the whole or a part of
the configuration of the present invention is configured as a
computer program, such computer program, for instance, may be fixed
in various storage mediums and distributed, or transmitted via a
communication network.
BRIEF DESCRIPTION OF THE DRAWINGS
[0048] FIG. 1 is an explanatory diagram showing the overall concept
of an embodiment of the present invention;
[0049] FIG. 2 is an explanatory diagram showing the schematic
configuration of the storage system;
[0050] FIG. 3 is a block diagram showing the hardware configuration
of the storage system;
[0051] FIG. 4 is an explanatory diagram showing the frame format of
the memory configuration of the storage system;
[0052] FIG. 5 is an explanatory diagram showing the respective
configurations of the management table and attribute table to be
used by a first virtualization storage device;
[0053] FIG. 6 is an explanatory diagram showing the respective
configurations of the management table and attribute table to be
used by a second virtualization storage device;
[0054] FIG. 7 is an explanatory diagram showing the configuration
of the path definition information and the situation of the host
path being switched based on this path definition information;
[0055] FIG. 8 is a flowchart showing the processing of the
virtualization storage devices acquiring information on the
external storage device and creating a management table and the
like;
[0056] FIG. 9 is an explanatory diagram showing the processing in
the case of operating in the asynchronous transfer mode;
[0057] FIG. 10 is an explanatory diagram showing the processing in
the case of operating in the synchronous transfer mode;
[0058] FIG. 11 is a flowchart showing the transfer designation
processing to be performed using the management terminal;
[0059] FIG. 12 is an explanatory diagram showing a screen display
example of the load status being monitored with the management
terminal;
[0060] FIG. 13 is a flowchart showing the outline of the processing
for newly adding the second virtualization storage device to the
storage system and transferring the volume from the first
virtualization storage device;
[0061] FIG. 14 is a flowchart showing the access processing to be
executed with the first virtualization storage device, which is the
transfer source;
[0062] FIG. 15 is a flowchart showing the access processing to be
executed with the second virtualization storage device, which is
the transfer destination;
[0063] FIG. 16 is a flowchart showing the outline of the processing
for transferring the volume between a plurality of virtualization
storage devices;
[0064] FIG. 17 is a flowchart showing the processing for the second
virtualization storage device, which is the transfer destination,
to connect with the external volume, which is the transfer
target;
[0065] FIG. 18 is an explanatory diagram showing the frame format
of the situation of operating the storage system with a plurality
of virtualization storage devices; and
[0066] FIG. 19 is a flowchart showing the transfer designation
processing to be executed with the storage system according to the
second embodiment.
DESCRIPTION OF THE SPECIFIC EMBODIMENTS
[0067] FIG. 1 is an explanatory diagram of the configuration
showing the overall schematic of an embodiment of the present
invention. As shown in FIG. 1, this storage system, for instance,
may be configured by having a plurality of virtualization storage
devices 1, 2, a plurality of external storage devices 3, a
plurality of host devices (hereinafter referred to as a "host") 4,
an upper level SAN (Storage Area Network) 5, a lower level SAN 6, a
management terminal 7, and a device-to-device LAN (Local Area
Network) 8.
[0068] Here, the virtualization storage device 1, 2 corresponds to
a "connection source storage device", and the external storage
device 3 corresponds to a "connection destination storage device".
The host 4 corresponds to a "host device", the upper level SAN 5
corresponds to a "first communication network", the lower level SAN
6 corresponds to a "second communication network", the management
terminal 7 corresponds to a "management terminal", and the
device-to-device LAN 8 corresponds to a "third communication
network".
[0069] Incidentally, the upper level SAN 5 and lower level SAN 6,
for example, may be configured as a FC_SAN (Fibre Channel_Storage
Area Network) or IP_SAN (Internet Protocol_SAN), but it is not
limited thereto, and, for instance, may also be configured as a LAN
or WAN (Wide Area Network). The upper level SAN 5 is used for
respectively connecting the respective hosts 4 and the respective
virtualization storage devices 1, 2. The lower level SAN 6 is used
for respectively connecting the respective virtualization storage
devices 1, 2 and the respective external storage device 3. The
upper level SAN 5 and lower level SAN 6 are separated, and the
traffic or failure of one communication network will not directly
influence the other communication network.
[0070] Attention is focused on the configuration of the first
virtualization storage device 1. The first virtualization storage
device 1 is used for virtualizing a volume 3A of the external
storage device 3 and providing this to the host 4. This first
virtualization storage device 1, for instance, has a control unit
1A, a first management table 1B, a second management table 1C, a
logical volume 1D, and an intermediate volume 1E.
[0071] Here, the control unit 1A corresponds to a "control unit",
the first management table 1B corresponds to "first management
information", the second management table 1C corresponds to "second
management information, the logical volume 1D corresponds to a
"logical volume", and the intermediate volume 1E corresponds to an
"intermediate volume".
[0072] The control unit 1A controls the overall operation of the
first virtualization storage device 1. The control unit 1A, for
instance, creates a logical volume 1D and provides this to the host
4. Further, the control unit 1A connects the logical volume 1D and
external volume 3A via the intermediate volume 1E by using the
first management table 1B and second management table 1C. Moreover,
the control unit 1A transfers the whole or a part of the external
volume 3A under its own control to the second virtualization
storage device 2 based on the designation from the management
terminal 7.
[0073] The first management table 1B is used for identifying the
respective external volumes 3A in the storage system and connecting
a desired external volume 3A to the logical volume 1D. The second
management table 1C is used for managing other attribute
information such as the copy status or difference management
information (difference bitmap) of the respective external volumes
3A.
[0074] The second virtualization storage device 2 may be configured
the same as the first virtualization storage device 1. The second
virtualization storage device 2, as with the first virtualization
storage device 1, is able to connect the whole or a part of the
respective external volumes 3A to the logical volume 2D via the
intermediate volume 2E. And, the second virtualization storage
device 2, as with the first virtualization storage device 1, is
able to provide the external volume 3A to the host 4 as though it
is one's own internal volume.
[0075] The second virtualization storage device 2, for instance,
may be configured by having a control unit 2A, a first management
table 2B, a second management table 2C, a logical volume 2D and an
intermediate volume 2E. Each of these components 2A to 2E have the
same configuration as each of the components 1A to 1E described
with reference to the first virtualization storage device 1, and
the detailed description thereof is omitted.
[0076] Nevertheless, it should be noted that the size of the second
management table 2C is smaller than the size of the second
management table 1C of the first virtualization storage device 1.
In the present embodiment, only the attribute information relating
to the external volume 3A transferred from the first virtualization
storage device 1 to the second virtualization storage device 2 is
copied from the second management table 1C of the first
virtualization storage device 1 to the second management table 2C
of the second virtualization storage device 2. Accordingly, the
table size of the second management table 2C is smaller than that
of the second management table 1C.
[0077] When the first virtualization storage device 1 is already
being used prior to the second virtualization storage device 2
being added to the storage system; that is, when the first
virtualization storage device 1 is virtualizing and using all
external volumes 3A, the first virtualization storage device 1 has
already obtained the attribute information of all external volumes
3A. Under these circumstances, when the second virtualization
storage device 2 is added to the storage system, and a part of the
external volume 3A is transferred from the first virtualization
storage device 1 to the second virtualization storage device 2,
only the attribute information relating to such transferred
external volume 3A is copied from the second management table 1C of
the first virtualization storage device 1 to the second management
table 2C of the second virtualization storage device 2.
[0078] Each external storage device 3 has at least one or more
external volumes 3A. An external volume is a volume existing
outside the respective virtualization storage devices 1, 2. Each
external volume 3A, for example, is provided on a physical memory
area of one or a plurality of memory apparatuses. As such memory
apparatus, for instance, a hard disk drive, optical disk drive,
semiconductor memory drive, tape drive and so on may be considered.
Further, as the hard disk drive, for example, various disks such as
a FC (Fibre Channel) disk, SAS (Serial Attached SCSI) disk and SATA
(Serial AT Attachment) disk may be used. Each external volume 3A is
connected to one of the logical volumes 1D, 2D via the intermediate
volume 1E, 2E, and provides a memory area to the virtualization
storage devices 1, 2.
[0079] The management terminal 7 is connected to both of the
virtualization storage devices 1, 2 via the device-to-device LAN 8.
The management terminal 7, for example, is configured as a personal
computer, portable information terminal (including portable phones)
or the like, and has a monitoring unit 7A. The monitoring unit 7A
respectively monitors the load status of the respective
virtualization storage devices 1, 2, and is able to display the
monitoring result on a terminal screen.
[0080] As the load status, for instance, input/output per second
(IOPS), CPU usage rate, cache memory usage rate and so on may be
considered. A user such as a system administrator is able to
comprehend the load status of the respective virtualization storage
devices 1, 2 based on the monitoring result of the monitoring unit
7A, and thereby determine the disposition of the volumes.
[0081] Incidentally, at least a part of the judgment process by the
user may be realized by a computer program, and the volume
disposition may also be automatically conducted based on the load
status of the respective virtualization storage devices 1, 2. The
user's decision to transfer the volume is notified to the
respective virtualization storage devices 1, 2 via the management
terminal 7.
[0082] Next, the operation method of the storage system according
to the present embodiment is explained. In the most initial state,
only the respective external storage devices 3 exist in the storage
system. Thereafter, the user introduces the first virtualization
storage device 1 to the storage system, virtualizes the external
volume 3A of the respective external storage devices 3 with the
first virtualization storage device 1, and provides this to the
respective hosts 4. Thereafter, for instance, four more hosts 4 are
additionally increased, and, when the processing performance of the
first virtualization storage device 1 is used up to its upper
limit, the user decides the introduction of the second
virtualization storage device 2. The user is able to decide the
introduction of the second virtualization storage device 2 based on
the monitoring result of the monitoring unit 7A (S0).
[0083] Then, the second virtualization storage device 2 is added to
the storage system (S1). The user or a corporate engineer selling
the second virtualization storage device 2 respectively connects
the second virtualization storage device 2 to the upper level SAN
and lower level SAN 6 (S2A, S2B). Further, the second
virtualization storage device 2 is connected to the first
virtualization storage device 1 via the device-to-device LAN 8
(S3).
[0084] Next, contents of the first management table 1B of the first
virtualization storage device 1 are copied to the second
virtualization storage device 2 (S4). Thereby, the first management
table 2B is created in the second virtualization storage device
2.
[0085] The user selects the external volume 3A to be transferred
from the first virtualization storage device 1 to the second
virtualization storage device 2 based on the monitoring result of
the monitoring unit 7A, and designates the transfer of the volume
(S5).
[0086] Based on the designation from the management terminal 7,
only the attribute information relating to the external volume 3A
transferred to the second virtualization storage device 2 among the
attribute information stored in the second management table 1C of
the first virtualization storage device 1 is transferred from the
first virtualization storage device 1 to the second virtualization
storage device 2 (S6).
[0087] The second virtualization storage device 2 connects the
external volume 3A designated by the management terminal 7 and the
logical volume 2D by using the first management table 2B and second
management table 2C (S7). And, the second virtualization storage
device 2 sets information for making the host 4 recognize the
logical volume 2D, and the host 4 sets a path for accessing this
logical volume 2D (S8).
[0088] The data used by the host 4, in reality, is stored in a
prescribed external volume 3A. Before the transfer of the volume,
the host 4 is accessing a prescribed external volume 3A from the
logical volume 1D of the first virtualization storage device 1 via
the intermediate volume 1E. The host 4 is totally unaware that such
data is stored in a prescribed external volume 3A.
[0089] When transferring such prescribed external volume 3A from
the first virtualization storage device 1 to the second
virtualization storage device 2, the second virtualization storage
device 2 connects such prescribed external volume 3A to the logical
volume 2D via the intermediate volume 2E. The host 4 is able to
access this logical volume 2D by correcting the path information,
and is thereby able to read and write desired data.
[0090] As described above, in the present embodiment, a plurality
of virtualization storage devices 1, 2 may be used to virtualize
and utilize the external volume 3A. And, the external volume 3A may
be transferred between the respective virtualization storage
devices 1, 2. Accordingly, the first virtualization storage device
1 and second virtualization storage device 2 can be used to
disperse the processing load, and the processing performance of the
storage system can be improved thereby. Thus, even when the demand
of storage services increases, by appropriately adding
virtualization storage devices, it will be possible to deal with
such increased demand, and the usability can be improved.
[0091] Incidentally, it is not necessary to make the respective
virtualization storage devices 1, 2 coexist, and, for instance,
after transferring all external volumes 3A from the first
virtualization storage device 1 to the second virtualization
storage device 2, the first virtualization storage device 1 may be
removed from the storage system. Embodiments of the present
invention are now described in detail below.
1. First Embodiment
[0092] FIG. 2 is an explanatory diagram showing the overall
schematic of the storage system according to the present
embodiment. To foremost explain the correspondence with FIG. 1, the
first virtualization storage device 100A illustrated in FIG. 2
corresponds to the first virtualization storage device 1 of FIG. 1,
and the second virtualization storage device 100B corresponds to
the second virtualization storage device 2 of FIG. 1. Similarly,
the external storage device 200 illustrated in FIG. 2 corresponds
to the external storage device 3 of FIG. 1, the host 10 of FIG. 2
corresponds to the host 4 of FIG. 1, and the management terminal 20
of FIG. 2 corresponds to the management terminal 7 of FIG. 1. The
communication network CN1 of FIG. 2 corresponds to the upper level
SAN 5 of FIG. 1, the communication network CN2 of FIG. 2
corresponds to the lower level SAN 6 of FIG. 1, and the
communication network CN3 of FIG. 2 corresponds to the
device-to-device LAN 8 of FIG. 1.
[0093] To foremost explain the network configuration of the storage
system, the respective hosts 10 are respectively connected to the
respective virtualization storage devices 100A, 100B via the upper
level network CN1. The respective virtualization storage devices
100A, 100B are respectively connected to the respective external
storage device 200 via the lower level network CN2. And, the
respective virtualization storage devices 100A, 100B and the
management terminal 20 are connected via the management network
CN3. For example, the communication network CN1, CN2 may be
configured as an IP_SAN or FC_SAN. Further, for instance, the
communication network CN3 may be configured as a LAN. Nevertheless,
the management communication network CN3 may be abolished, and
either or both the upper level network CN1 and lower level network
CN2 may be used to transfer information for managing the storage
system.
[0094] The schematic configuration of the storage system is now
explained. The host 10, for example, may be configured by having an
HBA (Host Bus Adapter) 11, a volume management unit 12, and an
application program 13 (abbreviated as "application" in the
diagrams). When the upper level network CN1 is configured as an
IP_SAN, in substitute for the HBA 11, for instance, a LAN card
equipped with a TCP/IP offload engine may be used. The volume
management unit 12 manages the path information and the like to the
volume to be accessed.
[0095] The first virtualization storage device 100A, for example,
may be configured by having a host connection interface
(abbreviated as "I/F" in the drawings) 111T, a controller 101A, and
an external storage connection interface 111E. Incidentally,
although the first virtualization storage device 100A has a logical
volume 164 as described later, the hierarchical memory
configuration will be described later together with FIG. 4.
[0096] The host connection interface 111T is used for connecting to
the respective hosts 10 via the upper level communication network
CN1. The external storage connection interface 111E is used for
connecting to the respective external storage devices 200 via the
lower level communication network CN2.
[0097] The controller 101A is used for controlling the operation of
the first virtualization storage device 100A. Although details of
the controller 101A will be described in detail later, the
controller 101A, for instance, may be configured by having one or a
plurality of microprocessors, memory, data processing circuit and
the like. A management table T1A and attribute table T2A are
respectively stored in the control memory 140 used by the
controller 101A. The management table T1A corresponds to the first
management table 1B of FIG. 1, and the attribute table T2A
corresponds to the second management table 1C of FIG. 1. These
management tables T1A, T2A will be described in detail later. Write
data and the like written from the host 10 is stored in the cache
memory 130 used by the controller 101A.
[0098] The second virtualization storage device 100B, as with the
first virtualization storage device 100A, may be configured by
having a host connection interface 111T, a controller 101B, and an
external storage connection interface 111E. And, a management table
T1B and attribute table T2B are stored in the control memory 140
used by the controller 101B.
[0099] The respective external storage devices 200, for example,
may be configured by respectively having a controller 210, a
communication port 211, and a logical volume 240. Since the logical
volume 240 is a volume existing outside the respective
virtualization storage devices 100A, 100B, this is sometimes
referred to as an external volume in the present specification.
[0100] The management terminal 20, for instance, is configured as a
personal computer, workstation, portable information terminal or
the like, and has a monitoring unit 21. The monitoring unit 21
respectively acquires the load status of the respective
virtualization storage devices 100A, 100B, and displays the
acquired load status on a terminal screen.
[0101] Incidentally, reference numeral 30 in FIG. 2 represents a
switch. In FIG. 2, although the switch 30 is only shown in the
upper level network, one or a plurality of such switches may also
be provided to the lower level network CN2.
[0102] FIG. 3 is an explanatory diagram showing the detailed
hardware configuration of the respective virtualization storage
devices 100A, 100B. To explain the first virtualization storage
device 100, the first virtualization storage device 100A, for
instance, may be configured by having a plurality of channel
adapters (hereinafter referred to as a "CHA") 110, a plurality of
disk adapters (hereinafter referred to as a "DKA") 120, a cache
memory 130, a control memory 140, a connection control unit 150, a
memory unit 160, and a service processor (hereinafter abbreviated
as "SVP") 170.
[0103] Each CHA 110 performs data communication with the host 10.
Each CHA 110 may have at least one or more communication interfaces
111T for communicating with the host 10. Each CHA 110 may be
configured as a microcomputer system equipped with a CPU, memory
and so on. Each CHA 110 interprets and executes the various
commands such as a read command or write command received from the
host 10.
[0104] Each CHA 110 is assigned a network address (e.g., IP address
or WWN) for identifying the respective CHAs 110, and each CHA 110
may also individually function as a NAS (Network Attached Storage).
When there is a plurality of hosts 10, each CHA 110 receives and
processes the request from each host 10 individually. Among the
respective CHAs 110, a prescribed CHA 110 is provided with an
interface (target port) 111T for communicating with the host 10,
and the other CHAs 110 are provided with an interface (externally
connected port) 111E for communicating with the external storage
device 200.
[0105] Each DKA 120 is used for transferring data to and from the
disk drive 161 of the memory unit 160. Each DKA 120, as with the
CHA 110, is configured as a microcomputer system equipped with a
CPU, memory and so on. Each DKA 120, for example, is able to write
data that the CHA 110 received from the host 10 or data read from
the external storage device 200 into a prescribed disk drive 161.
Further, each DKA 120 is also able to read data from a prescribed
disk drive 161 and transmit this to the host 10 or external storage
device 200. When inputting and outputting data to and from the disk
drive 161, each DKA 120 converts a logical address into a physical
address.
[0106] When the disk drive 161 is managed according to RAID, each
DKA 120 performs data access according to such RAID configuration.
For example, each DKA 120 respectively writes the same data in
separate disk drive groups (RAID groups) (RAID 1, etc.), or
executes a parity account and writes the data and parity in the
disk drive group (RAID 5, etc.). Incidentally, in the present
embodiment, the respective virtualization storage devices 100A,
100B virtualize and incorporate the external volume 240 of the
external storage device 200, and provides this to the host 10 as
though it is one's own internal volume.
[0107] Therefore, the respective virtualization storage devices
100A, 100B do not necessarily have to have a memory unit 160. The
respective virtualization storage devices 100A, 100B are used to
virtualize and utilize the external volume 240. When the respective
virtualization storage devices 100A, 100B do not have a memory unit
160, the DKA 120 will not be required. Incidentally, the
configuration may also be such that one virtualization storage
device has a memory unit 160, and the other virtualization storage
device does not have a memory unit 160.
[0108] The cache memory 130 stores the data received from the host
10 or external storage device 200. Further, the cache memory 130
stores data read from the disk drive 161. As described later, the
memory space of the cache memory 130 is used to create a virtual,
intermediate memory apparatus (V-VOL).
[0109] The control memory 140 stores various types of control
information to be used in the operation of the virtualization
storage device 100A. Further, a work area is set in the control
memory 140, and various tables described later are also stored
therein.
[0110] Incidentally, one or a plurality of disk drives 161 may be
used as the cache disk. Further, the cache memory 130 and control
memory 140 may be configured to be separate memories, or a part of
the memory area of the same memory may be used as the cache area,
and the other memory area may be used as the control area.
[0111] The connection control unit 150 mutually connects the
respective CHAs 110, respective DKAs 120, cache memory 130 and
control memory 140. The connection control unit 150, for instance,
can be configured as a crossbar switch or the like.
[0112] The memory unit 160 has a plurality of disk drives 161. As
the disk drive 161, for example, various memory apparatuses such as
a hard disk drive, flexible disk drive, magnetic tape drive,
semiconductor memory drive and optical disk drive as well as the
equivalents thereof may be used. Further, for instance, different
types of disks such as a FC (Fibre Channel) disk and a SATA (Serial
AT Attachment) disk may coexist in the memory unit 160.
[0113] The service processor (SVP) 170 is respectively connected to
each CHA 110 via an internal network such as a LAN. The SVP 170 is
able to send and receive data to and from the control memory 140 or
DKA 120 via the CHA 110. The SVP 170 extracts various types of
information in the first virtualization storage device 100A and
provides this to the management terminal 20.
[0114] Since the second virtualization storage device 100B can also
be configured the same as the first virtualization storage device
100A, the explanation thereof is omitted. Nevertheless, the
respective virtualization storage devices 100A, 100B do not have to
be configured the same.
[0115] The external storage device 200 may be configured
approximately the same as the virtualization storage devices 100A,
100B, or may be configured more simple than the respective
virtualization storage devices 100A, 100B.
[0116] Here, care should be given to the network configuration of
the storage system. As described above, the upper level network CN1
connecting the host 10 and respective virtualization storage
devices 100A, 100B and the lower level network CN2 mutually
connecting the respective storage devices 100A, 100B, 200 are
respectively configured as a separate communication network.
Therefore, large quantities of data can be transferred with the
lower level network CN2 without directly influencing the upper
level network CN1.
[0117] Explanation is not provided with reference to FIG. 4. FIG. 4
is an explanatory diagram showing the memory configuration of the
storage system. Foremost, the configuration of the virtualization
storage devices 100A, 100B is explained taking the first
virtualization storage device 100A as an example.
[0118] The memory configuration of the first virtualization storage
device 100A, for example, can be broadly classified into a physical
memory hierarchy and a logical memory hierarchy. The physical
memory hierarchy is configured from a PDEV (Physical Device) 161,
which is a physical disk. PDEV corresponds to the foregoing disk
drive 161.
[0119] The logical memory hierarchy may be configured from a
plurality of (e.g., two types of) hierarchies. One logical
hierarchy may be configured from a VDEV (Virtual Device) 162, and a
virtual VDEV (hereinafter sometimes referred to as "V-VOL") 163
which is treated like the VDEV 162. The other logical hierarchy may
be configured from a LDEV (Logical Device) 164.
[0120] The VDEV 162, for example, is configured by grouping a
prescribed number of PDEVs 161 such as in a set of fours (3D+1P),
or a set of eights (7D+1P). The memory areas provided respectively
from each PDEV 161 belonging to the group are assembled to form a
single RAID storage area. This RAID memory area becomes the VDEV
162.
[0121] In contrast to the VDEV 162 being created on a physical
memory area, the V-VOL 163 is a virtual intermediate memory
apparatus that does not require a physical memory area. The V-VOL
163 is not directly associated with a physical memory area, and is
a virtual existence to become the receiver for mapping an LU
(Logical Unit) of the external storage controller device 200. This
V-VOL 163 corresponds to an intermediate volume.
[0122] At least one or more LDEVs 164 may be provided on the VDEV
162 or V-VOL 163. The LDEV 164, for instance, may be configured by
dividing the VDEV 162 in a fixed length. When the host 10 is an
open host, by the LDEV 164 being mapped with the LU 165, the host
10 will recognize the LDEV 164 as a single physical disk. An open
host can access a desired LDEV 164 by designating the LUN (Logical
Unit Number) or logical block address. Incidentally, a mainframe
host will directly recognize the LDEV 164.
[0123] The LU 165 is a device that can be recognized as a logical
unit of SCSI. Each LU 165 is connected to the host 10 via the
target port 111T. At least one or more LDEVs 164 may be
respectively associated with each LU 165. Incidentally, as a result
of associating a plurality of LDEVs 164 to a single LU 165, the LU
size can be virtually expanded.
[0124] A CMD (Command Device) 166 is a dedicated LU to be used for
transferring commands and statuses between the I/O control program
operating on the host 10 and the storage device 100.
[0125] For example, a command from the host 10 is written in the
CMD 166. The first virtualization storage device 100 executes the
processing according to the command written in the CMD 166, and
writes the execution result thereof as the status in the CMD 166.
The host device 10 reads and confirms the status written in the CMD
166, and writes the processing contents to be executed subsequently
in the CMD 166. As described above, the host device 10 is able to
give various designations to the first virtualization storage
device 100A via the CMD 166.
[0126] Incidentally, the command received from the host device 10
may also be processed directly by the first virtualization storage
device 100A without being stored in the CMD 166. Moreover, the CMD
may be created as a virtual device without defining the actual
device (LU) and configured to receive and process the command from
the host device 10. In other words, for example, the CHA 110 writes
the command received from the host device 10 in the control memory
140, and the CHA 110 or DKA 120 processes this command stored in
the control memory 140. The processing results are written in the
control memory 140, and transmitted from the CHA 110 to the host
device 10.
[0127] An external storage device 200 is connected to an initiator
port (External Port) 111E for external connection of the first
virtualization storage device 100A via the lower level network
CN2.
[0128] The external storage device 200 has a plurality of PDEVs
220, a VDEV 230 set on a memory area provided by the PDEV 220, and
one or more LDEVs 240 that can be set on the VDEV 230. And, each
LDEV 240 is respectively associated with the LU 250. The PDEV 220
corresponds to the disk drive 220 of FIG. 3. The LDEV 240
corresponds to a "separate logical volume", and corresponds to the
external volume 3A of FIG. 1.
[0129] The LU 250 (i.e., LDEV 240) of the external storage device
200 is mapped to the V-VOL 163. For example, the "LDEV 1", "LDEV 2"
of the external storage device 200 are respectively mapped to the
"V-VOL 1", "V-VOL 2" of the first virtualization storage device
100A via the "LU 1", "LU 2" of the external storage device 200.
And, "V-VOL 1", "V-VOL 2" are respectively mapped to the "LDEV 3",
"LDEV 4", and the host device 10 is thereby able to use these
volumes via the "LU 3", "LU 4".
[0130] Incidentally, the VDEV 162, V-VOL 163 may adopt the RAID
configuration. In other words, a single disk drive 161 may be
assigned to a plurality of VDEVs 162, V-VOLs 163 (slicing), and a
single VDEV 162, V-VOL 163 may be formed from a plurality of disk
drives 161 (striping).
[0131] Since the second virtualization storage device 100B may have
the same hierarchical memory configuration as the first
virtualization storage device 100A, the explanation thereof is
omitted.
[0132] FIG. 5 is an explanatory diagram showing the schematic
configuration of the management table T1A and attribute table T2A
used by the first virtualization storage device 100A. Each of these
tables T1A, T2A may be stored in the control memory 140.
[0133] The management table T1A is used for uniformly managing the
respective external volumes 240 dispersed in the storage system.
The management table T1A, for instance, may be configured by
respectively associating a network address (WWN: World Wide Name)
for connected to the respective external volumes 240, a number
(LUN: Logical Unit Number) of the respective external volumes 240,
volume size of the respective external volumes 240, an external
volume number, owner right information and transfer status
flag.
[0134] Here, an external volume number is identifying information
for uniquely specifying the respective external volumes 240 in the
storage system. Owner right information is information for
specifying the virtualization storage devices having the authority
to use such external volume. When "0" is set in the owner right
information, it shows that such external volume 240 is unused. When
"1" is set in the owner right information, it shows that one's own
device has the usage authorization to use such external volume 240.
Further, when "-1" is set in the owner right information, it shows
that the other virtualization storage devices have the usage
authorization to use such external volume 240.
[0135] Specifically, with respect to the external volume 240 to
which "1" is set in the owner right information in the management
table T1A used by the first virtualization storage device 100A, the
first virtualization storage device 100A has the usage
authorization thereof. Similarly, with respect to the external
volume 240 to which "-1" is set in the management table T1A, the
second virtualization storage device 100B has the usage
authorization thereof. As described above, when the owner right
information is set as "1" in one management table regarding a
certain external volume 240, the ownership right information of
such external volume is set to "-1" in the other management table.
By referring to the owner right information, whether such external
volume is under the control of one of the virtualization storage
devices, or is an unused volume can be known.
[0136] Incidentally, in the present embodiment, since only two
virtualization storage devices 100A, 100B are shown, by setting
either "1" or "-1" in the owner right information, the affiliation
of such external volume 240 can be specified. In addition to the
above, if there are three or more virtualization storage devices in
the storage system, as the owner right information, for instance,
the case number assigned to the respective virtualization storage
devices may also be set. In other words, identifying information
capable of uniquely specifying the respective virtualization
storage devices in the storage system may be used as the owner
right information.
[0137] The transfer status flag is information showing that the
external volume 240 is being transferred from one virtualization
storage device to the other virtualization storage device. When "1"
is set in the transfer status flag, this shows that the owner right
of such external volume 240 is being changed. Meanwhile, when "0"
is set in the transfer status flag, this shows that such external
volume 240 is in a normal state, and the owner right is not being
changed.
[0138] The attribute table T2A is a table for managing various
types of attribute information of the respective external volumes
240. The attribute table T2A, for example, may be configured by
associating the LU number of the respective external volumes 240,
path definition information, replication configuration information,
replication status information, and replication bitmap information.
Path definition information is information for showing, via which
port of which CHA 110, the logical volume 164 connected to such
external volume 240 is to be accessed by the host 10. A plurality
of paths may be set in the path definition information. One path is
the normally used primary path, and the other path is an alternate
path to be used when there is failure in the primary path.
[0139] The replication configuration information is information
showing the correspondence of the volumes configuring a copy-pair.
A volume in which "P" is set in the replication configuration
information is a primary volume (copy source volume), and a volume
in which "S" is set in the replication configuration information is
a secondary volume (copy destination volume). Incidentally, the
numbers appended to "P" and "S" are serial numbers for identifying
the respective copy-pairs.
[0140] The replication status information is information showing
the status of the respective volumes configuring the copy-pair.
When "Pair" is set in the replication status information, the
volume thereof is in synchronization with the volume of the other
party, and shows that the respective volumes forming the copy-pair
are maintaining the same memory contents. When "Resync" is set in
the replication status information, this shows that the volume
thereof and the volume of the other party are in resynchronization.
When "Simplex" is set in the replication status information, this
shows that the volume thereof is not a target of replication. When
"Suspend" is set in the replication status information, this shows
that the volume thereof has not been updated with the volume of the
other party.
[0141] The replication bitmap information is information showing
the updated position of the data in the volume thereof. For
example, a flag showing whether the data has been updated is
prepared for each segment, and this means that, in a segment with
"1" set to the flag, the data thereof has been updated. For
example, when managing the existence of the update of data
regarding a logical volume 164 having a volume size of 1 TB in a
segment size of 1 MB, the size of the replication bitmap
information will be 128 KB. When the first virtualization storage
device 100A is able to set n number of logical volumes 164, the
total size of the replication bitmap information will be
n.times.128 KB. When n is 16384, the total size of the replication
bitmap information will be 16384.times.128 KB=2048 MB.
[0142] As described above, when only focusing attention on the
replication bitmap information, the table size of the attribute
table T2A will be enormous. According, when the entirety of this
attribute table T2A is to be transferred to the second
virtualization storage device 100B, the control memory 140 of the
second virtualization storage device 100B will be compressed. Thus,
in the present embodiment, among the information stored in the
attribute table T2A, only the information relating to the volume to
be transferred to the second virtualization storage device 100B is
transferred to the second virtualization storage device 100B. In
other words, attribute information is transferred to the necessary
extent. Thereby, the data volume to be transferred can be reduced,
the time required for creating the attribute table can be
shortened, and the compression of the memory resource (control
memory 140) of the second virtualization storage device 100B, which
is the transfer destination, can be prevented.
[0143] Incidentally, in addition to the foregoing items, for
instance, information such as the device type (disc device or tape
device, etc.), vendor name, identification number of the respective
storage devices and so on may also be managed. Such information may
be managed with either the management table T1A or attribute table
T2A.
[0144] FIG. 6 is an explanatory diagram showing the schematic
configuration of the management table T1B and attribute table T2B
used by the second virtualization storage device 100B. The
management table T1B, as with the management table T1A described
above, for instance, is configured by associating a network address
such as WWN, an LU number, volume size, an external volume number,
owner right information and a transfer status flag. The management
table T1A and management table T1B are configured the same
excluding the owner right information.
[0145] The attribute table T2B, as with the attribute table T2A
described above, is also configured by associating an LU number,
path definition information, replication configuration information,
replication status information and replication bitmap information.
Nevertheless, as described above, in order to effectively use the
memory resource of the second virtualization storage device 100B,
it should be noted that only the attribute information of the
volume under the control of the second virtualization storage
device 100B is registered in the management table T2B.
[0146] FIG. 7 is an explanatory diagram showing the schematic
configuration of the path setting information T3 to be used by the
volume management unit 12 of the host 10. This path setting
information T3 may be stored in the memory of the host 10 or a
local disk.
[0147] The path setting information T3 includes information
relating to the primary path to be used in normal times, and
information relating to the alternate path to be used in abnormal
times. Each path, for instance, is configured by including
information for specifying the HBA 11 to be used, port number of
the access destination, and LU number for identifying the volume of
the access target.
[0148] Although a plurality of alternate paths is described in the
path setting information T3, the alternate path described first is
a normal alternate path, and the subsequently described alternate
path is a path unique to the present embodiment. In other words,
the second alternate path is a path set upon transferring the
volume from the first virtualization storage device 100A to the
second virtualization storage device 100B.
[0149] The lower part of FIG. 7 shows a frame format of the
situation of switching from the primary path to the alternate path.
Here, explained is a case where the volume 420 of "#0" is
transferred from the first virtualization storage device 100A to
the second virtualization storage device 100B.
[0150] Before the transfer, by accessing the Port #0 from the HBA
#0 as shown with the thick line in FIG. 7, the host 10 is able to
read and write data from and into the logical volume of the first
virtualization storage device 100A. In the first virtualization
storage device 100A, the external volume 240 is accessed from the
Port #1 based on the access from the host 10.
[0151] When transferring the volume, information for the host 10 to
access the transferred volume is added to the path setting
information T3 as the second alternate path. And, the first
virtualization storage device 100A rejects the access request
regarding the transferred volume.
[0152] Therefore, even if the host 10 tries to access the
transferred volume via the primary path shown with the thick line
in FIG. 7, such access will be rejected by the first virtualization
storage device 100A. Thus, the host 10 tries re-accessing such
transferred volume by switching to the first alternate path (HBA
#1.fwdarw.Port #2.fwdarw.LU #0) shown with the dotted line in FIG.
7. Nevertheless, this access is also rejected by the first
virtualization storage device 100A.
[0153] Then, the host 10 tries to access the volume by switching to
the second alternate path (HBA #1.fwdarw.Port #4.fwdarw.LU #0)
shown with the dashed line in FIG. 7. The second alternate path is
a path to the second virtualization storage device 100B, which is
the volume transfer destination. When the access request from the
host 10 is in a processible state, the second virtualization
storage device 100B processes this access request, and returns the
processing result to the host 10. The processible state of the
access request means that even when the access request from the
host 10 is processed, inconsistency in the data stored in the
volume will not occur. This will be described in detail later.
[0154] As described above, when the host 10 is unsuccessful in
accessing via the primary path, it switches to the first alternate
path, and, when it is unsuccessful in accessing via the first
alternate path, it switches to the second alternate path.
Accordingly, until the access request of the host 10 is accepted,
some time (path switching time) will be required. Nevertheless,
this path switching time is not wasteful time. This is because, as
described later, destage processing to the transferred volume can
be performed during such path switching time. In the present
embodiment, merely by adding a new path to the path setting
information T3 stored in the host 10, the access destination of the
host 10 can be switched.
[0155] FIG. 8 is a flowchart showing the outline of the processing
for searching the external volume existing in the storage system
and registering this in the management table T1A. Here, an example
of a case where the first virtualization storage device 100A
executes the processing is explained.
[0156] Foremost, the first virtualization storage device 100A
issues a command ("Test Unit Ready") toward the respective external
storage devices 200 for confirming the existence thereof (S11).
Each external storage device 200 operating normally will return a
Ready reply having a Good status as the response to such command
(S12).
[0157] Next, the first virtualization storage device 100A issues an
"Inquiry" command to each external storage device 200 in which the
existence thereof has been confirmed (S13). Each external storage
device 200 that received this command, for instance, transmits
information regarding the device type and so on to the first
virtualization storage device 100A (S14).
[0158] The first virtualization storage device 100A issues a "Read
Capacity" command to each external storage device 200 (S15). Each
external storage device 200 transmits the size of the external
volume 240 to the first virtualization storage device 100A
(S16).
[0159] The first virtualization storage device 100A transmits a
"Report LUN" command to each external storage device 200 (S17).
Each external storage device 200 transmits the LUN quantity and LUN
number to the first virtualization storage device 100A (S18).
[0160] The first virtualization storage device 100A registers the
information acquired from each external storage device 200 in the
management table T1A and attribute table T2A, respectively. As
described above, the first virtualization storage device 100A is
able to respectively create the management table T1A and attribute
table T2A by issuing a plurality of inquiry commands.
[0161] Incidentally, the configuration of the storage system may
change by one of the external storage devices 200 being removed, or
a new external storage device 200 being added. When the
configuration of the storage system is changed, for example, the
first virtualization storage device 100A is able to detect such
change in configuration based on command and notifications such as
RSCN (Registered State Change Notification), LIP (Loop
Initialization Primitive), SCR (State Change Registration) or SCN
(State Change Notification). Incidentally, the foregoing processing
may also be executed by the second virtualization storage device
100B.
[0162] Next, the method of the virtualization storage devices 100A,
100B using the external volume 240 to process the access request
from the host 10 is explained. Here, although explained is a case
where the first virtualization storage device 100A processes the
access request, the second virtualization storage device 100B may
also perform the same processing. Foremost, the processing method
of a write command is explained. As the method for processing the
write command, two methods; namely, the synchronous transfer mode
and asynchronous transfer mode may be considered.
[0163] In the case of the synchronous transfer mode, when the first
virtualization storage device 100A receives a write command from
the host 10, the first virtualization storage device 100A stores
the write data received from the host 10 in the cache memory 130,
and thereafter transfers the write data to the external storage
device 200 via the communication network CN2. When the external
storage device 200 receives the write data and stores this in the
cache memory, it transmits a reply signal to the first
virtualization storage device 100A. When the first virtualization
storage device 100A receives the reply signal from the external
storage device 200, it transmits a write completion report to the
host 10.
[0164] As described above, in the synchronous transfer mode, after
the write data is transferred to the external storage device 200,
the completion of the write command processing is notified to the
host 10. Accordingly, in the synchronous transfer mode, a delay
will arise in the time of waiting for the reply from the external
storage device 200. Thus, the synchronous transfer mode is suitable
in cases where the distance between the first virtualization
storage device 100A and external storage device 200 is relatively
short. Contrarily, if the first virtualization storage device 100A
and external storage device 200 are far apart, generally speaking,
the synchronous transfer mode is not suitable due to problems of
delays in reply and delays in propagation.
[0165] Contrarily, in the case of an asynchronous transfer mode,
when the first virtualization storage device 100A receives a write
command from the host 10, it stores the write data in the cache
memory 130, and thereafter immediately issues a write completion
report to the host 10. After issuing the write completion report to
the host 10, the first virtualization storage device 100A transfers
the write data to the external storage device 200. The write
completion report to the host 10 and the data transfer to the
external storage device 200 are conducted asynchronously.
Accordingly, in the case of the asynchronous transfer mode, the
write completion report can be transmitted to the host 10 quickly
irrelevant to the distance between the first virtualization storage
device 100A and external storage device 200. Thus, the asynchronous
transfer mode is suitable when the distance between the first
virtualization storage device 100A and external storage device 200
is relatively long.
[0166] FIG. 9 is an explanatory diagram showing the case of the
asynchronous transfer mode. In FIG. 9 and FIG. 10, the
virtualization storage devices 100A, 100B are not differentiated,
and will be referred to as the "virtualization storage device 100".
Further, the management tables T1A, T1B are not differentiated, and
will be referred to as the "management table T1".
[0167] The host 10 issues a write command to a prescribed LU 165 of
the virtualization storage devices 100 (S31). The LU 165 is
associated with the LU 250 of the external storage device 200 via
the V-VOL 163. The LU 165 of the virtualization storage devices 100
is an access target from the host 10, but the external LU 250 is
actually storing the data. Therefore, for instance, the LU 165 may
be referred to as an "access destination logical memory apparatus"
and the LU 250 may be referred to as a "data storage destination
logical memory apparatus", respectively.
[0168] When the virtualization storage devices 100 receives a write
command from the host 10, it specifies the LU targeted by such
write command, refers to the management table T1 and determines
whether this LU is associated with an external volume. When this is
a write command to an LU associated with an external volume, the
virtualization storage device 100 transmits a write command to the
external storage device 200 having such external volume (S32).
[0169] After the write command is issued, the host 10 transmits the
write data with the LU 165 as the write target to the
virtualization storage devices 100 (S33). The virtualization
storage device 100 temporarily stores the write data received from
the host 10 in the cache memory 130 (S34). After the virtualization
storage device 100 stores the write data in the cache memory 130,
it reports the completion of writing to the host 10 (S35).
[0170] After converting the address and so on, the virtualization
storage device 100 transmits the write data stored in the cache
memory 130 to the external storage device 200 (S36). The external
storage device 200 stores the write data received from the
virtualization storage device 100 in the cache memory. And, the
external storage device 200 reports the completion of writing to
the virtualization storage device 100 (S37). The external storage
device 200, for example, looks out for a period with few I/O, and
writes the write data stored in the cache memory in the memory
apparatus 220 (destage processing). In the asynchronous transfer
mode, after write data is received from the host 10, the write
completion can be sent to the host 10 in a short reply time
.delta.1.
[0171] FIG. 10 shows a case of the synchronous transfer mode. Upon
receiving the write command issued from the host 10 (S41), the
virtualization storage device 100 specifies the external volume (LU
250) associated with the access destination volume (LU 165) of the
write command, and issues a write command to such external volume
(S42).
[0172] When the virtualization storage device 100 receives the
write data from the host 10 (S43), it stores this write data in the
cache memory 130 (S44). The virtualization storage device 100
transfers the write data stored in the cache memory 130 to the
external storage device 200 such that it is written in the external
volume (S45). After storing the write data in the cache memory, the
external storage device 200 reports the completion of writing to
the virtualization storage device 100 (S46). When the
virtualization storage device 100 confirms the completion of
writing in the external storage device 200, it reports the
completion of writing to the host 10 (S47). In the synchronous
transfer mode, since the report of the write completion to the host
10 is made upon waiting for the processing in the external storage
device 200, the reply time .delta.2 will become long. The reply
time .delta.2 of the synchronous transfer mode is longer than the
reply time .delta.1 of the asynchronous transfer mode
(.delta.2.gtoreq..delta.1).
[0173] As described above, the respective virtualization storage
devices 100A, 100B are able to incorporate and use the external
volume 240 of the external storage device 200 as though it is a
virtual internal volume.
[0174] Next, the method of transferring the external volume 240
being used by the first virtualization storage device 100A to the
second virtualization storage device 100B is explained.
Incidentally, the external volume 240 may also be transferred from
the second virtualization storage device 100B to the first
virtualization storage device 100A.
[0175] FIG. 11 is a flowchart showing the processing for
designating the transfer of the volume to the respective
virtualization storage devices 100A, 100B.
[0176] For example, when the user provides a designation to the
management terminal 20, the monitoring unit 21 acquires performance
information from the first virtualization storage device 100A
(S51). The monitoring unit 21 displays the acquired performance
information on a terminal screen of the management terminal 20
(S52). This performance information corresponds to the information
showing the "load status", and, for instance, includes the
input/output per second (IOPS), CPU usage rate, cache memory usage
rate and so on.
[0177] The user discovers whether there is a high-load CPU based on
the performance information displayed on the screen of the
management terminal 20 (S53). This CPU represents the CPU built in
the CHA 110. Next, the user confirms that every CPU of other CHAs
110 is of a load that is greater than a prescribed value (S54).
[0178] And, in order to alleviate the load of the high-load CHA
110, the user determines the transfer of the external volume 240
under the control of such CHA 110 (S55). Subsequently, the user
sets a path of the transfer destination (S56). In other words, the
user defines the path information regarding which port the host 10
will use for the access in the second virtualization storage device
100B, which is the transfer destination (S56). The defined path
information is added to the host 10. Finally, the user designates
the transfer of such external volume 240 to the respective
virtualization storage devices 100A, 100B (S57).
[0179] In other words, the user specifies the external volume that
is being the bottleneck in the first virtualization storage device
100A, which is the transfer source (switching source) (S53 to S55)
based on the monitoring result of the monitoring unit 21 (S51,
S52), and designates the start of transfer by defining the path of
the transfer destination (S56, S57). Each of the foregoing
processing steps may also be conducted automatically.
[0180] FIG. 12 is an explanatory diagram showing an example of a
screen showing the monitoring result of the monitoring unit 21. The
monitoring unit 21 is able to respectively acquire performance
information from the respective virtualization storage devices
100A, 100B, and display such performance information upon
performing statistical processing or creating a graphical chart
thereof.
[0181] In the selection unit G11, it is possible to select which
load status regarding which resource among the various resources in
the storage system is to be displayed. Here, as the resource, for
instance, "network", "storage", "switch" and so on may be
considered.
[0182] When the user selects "storage", the user may further select
one of the virtualization storage devices 100A, 100B. Further, when
the user selects one of the virtualization storage devices 100A,
100B, the user may make a more detailed selection. As such detailed
selection, "port" and "LU" may be considered. As described above,
the user is able to select in detail the desired target for
confirming the load status.
[0183] For example, in the first display unit G12, the overall
status of the selected virtualization storage device can be
displayed as a list among the virtualization storage devices 100A,
100B. With the second display unit G13, for example, a more
detailed monitoring target status, such as the "port" and "LU", can
be displayed. Further, with the third display unit G13, the load
status can be displayed as a graph.
[0184] The user is able to relatively easily determine which part
of which virtualization storage device is a bottleneck based on the
performance monitoring screen as shown in FIG. 12. Thus, the user
is able to determine the volume to be transferred based on such
determination.
[0185] FIG. 13 is a flowchart showing the situation of newly adding
a second virtualization storage device 100B to the storage system
in a state where the first virtualization storage device 100A is in
operation, and transferring one or a plurality of volumes from the
first virtualization storage device 100A to the second
virtualization storage device 100B. Incidentally, in FIG. 13 and so
on, the first virtualization storage device 100A is abbreviated as
the "first storage" and the second virtualization storage device
100B is abbreviated as the "second storage", respectively.
[0186] The user will be able to comprehend the load status of the
first virtualization storage device 100A with the methods described
with reference to FIG. 11 and FIG. 12. As a result, the user will
be able to determine the additional injection of the second
virtualization storage device 100B.
[0187] Foremost, the user or engineer of the vendor performs
physical connection procedures of the newly introduced second
virtualization storage device 100B (S61). Specifically, the host
connection interface 111T of the second virtualization storage
device 100B is connected to the upper level network CN1, the
external storage connection interface 111E of the second
virtualization storage device 100B is connected to the lower level
network CN2, and the SVP 170 of the second virtualization storage
device 100B is connected to the network CN3.
[0188] Next, the second virtualization storage device 100B acquires
the memory contents of the management table T1A from the first
virtualization storage device 100A (S62). Based on such acquired
contents, the second virtualization storage device 100B creates a
management table T1B. The second virtualization storage device 100B
respectively detects the external volumes 240 in the storage system
based on the management table T1B (S63).
[0189] When the user designates the transfer of the volume from the
management terminal 20 (S64), the second virtualization storage
device 100B connects the designated external volume 240 to the
V-VOL 163 via the interface 111E (S65).
[0190] Details of external connection are shown in FIG. 17. Thus,
explanation will foremost be made with reference to FIG. 17. The
second virtualization storage device 100B acquires attribute
information relating to the transfer target volume from the storage
device of the transfer source; that is, the first virtualization
storage device 100A (S151). The second virtualization storage
device 100B registers the attribute information other than the path
information among the acquired attribute information in the
attribute table T2B (S152). The second virtualization storage
device 100B newly sets path definition information regarding the
transfer target volume (S153).
[0191] Here, the user selects the logical volume 164 to be accessed
from the host 10 as the transfer target. When the selected logical
volume 164 is connected to the external volume 240, in hindsight,
the external volume 240 connected to such logical volume 164 will
be reconnected to a separate logical volume 164 of the transfer
destination storage device (100B). As described above, the
virtualization storage devices 100A, 100B connect the external
volume 240 to the logical volume 164 via the V-VOL 163, and are
able to use this as though it is one's own internal memory
apparatus.
[0192] Returning to FIG. 13, the volume management unit 12 of the
host 10 adds the path information for accessing the transferred
volume to the path setting information T3 (S66). In other words,
path information for accessing the logical volume 164 connected to
the external volume 240 via a prescribed port of the second
virtualization storage device 100B is set.
[0193] The first virtualization storage device 100A sets an owner
right regarding the external volume 240 designated as the transfer
target (S67). In other words, "-1" is set in the owner right
information regarding the transfer target volume. The first
virtualization storage device 100A notifies the set owner right
information to the second virtualization storage device 100B
(S68).
[0194] When the second virtualization storage device 100B acquires
the owner right information from the first virtualization storage
device 100A (S69), it registers the acquired owner right
information in the management table T1B (S70). Here, the owner
right information is registered in the management table T1B upon
the value thereof being changed to "1". This is because the usage
authorization of the transfer target volume has been transferred to
the second virtualization storage device 100B. The second
virtualization storage device 100B reports the completion of
registration of the owner right information to the first
virtualization storage device 100A (S71). The first virtualization
storage device 100A receives the setting completion report of the
owner right information from the second virtualization storage
device 100B (S72).
[0195] When the access request relating to the transfer target
volume is issued by the host 10 (S73), the first virtualization
storage device 100A starts the destage processing without
processing the access request (S74). Access processing in the
transfer source before the completion of transfer will be described
later with reference to FIG. 14. The second virtualization storage
device 100B receives a notice indicating the completion of destage
processing from the first virtualization storage device 100A
(S75).
[0196] Meanwhile, when the command processing issued to the first
virtualization storage device 100A is rejected, the host 10 refers
to the path setting information T3, switches to a different path
(S76), and reissues the command (S77). Here, for the sake of
convenience of explanation, the switch shall be from the primary
path passing through the first virtualization storage device 100A
to the second alternate path passing through the second
virtualization storage device 100B.
[0197] When the second virtualization storage device 100B receives
a command from the host 10, it performs access processing (S78). If
at the point in time of receiving the command the destage
processing of the transfer target volume is complete, normal access
processing will be performed. If the destage processing is not
complete, however, different access processing will be performed.
Access processing in the transfer destination before the completion
of the transfer will be described later with reference to FIG. 15.
Incidentally, the flow shown in FIG. 13 is merely an example, and,
in reality, there are cases where the order of steps will be
different.
[0198] FIG. 14 is a flowchart showing the details of S74 in FIG.
13. When the first virtualization storage device 100A, which is the
transfer source storage device, receives a command from the host 10
(S81: YES), it analyzes the access target of such command. The
first virtualization storage device 100A determines whether the
command in which the logical volume 164 connected to the external
volume 240 of its own usage authorization is the access target
(S82). In other words, the first virtualization storage device 100A
determines whether the command is an access request relating to the
external volume 240 in which it has the owner right.
[0199] When the first virtualization storage device 100A determines
that it is an access to the logical volume 164 connected to the
external volume 240 in which it does not have the usage
authorization; that is, the external volume 240 in which "-1" is
set in the owner right information (S82: NO), the command
processing from the host 10 is rejected (S83). Refection of the
command processing, for instance, may be made by not replying for a
prescribed period of time (negative rejection), or by notifying the
host 10 that processing is impossible (positive rejection).
[0200] The first virtualization storage device 100A starts the
destage processing of dirty data regarding the external volume 240
in which the access was requested from the host 10 (S84). And, when
the destage processing is complete (S85: YES), the first
virtualization storage device 100A notifies the second
virtualization storage device 100B to such effect (S86).
[0201] Amore detailed explanation is now provided. The access
target of the host 10 is the logical volume 164 of the first
virtualization storage device 100A. The logical volume 164 is
selected as the transfer target. and, this logical volume 164 is
connected to the logical volume 240 of the external storage device
200.
[0202] Here, the first virtualization storage device 100A is
processing the write command in the asynchronous transfer mode.
Accordingly, the first virtualization storage device 100A reports
the completion of writing to the host 10 at the time the write data
received from the host 10 is stored in the cache memory 130. The
write data stored in the cache memory 130 is transferred to the
external storage device 200 in a prescribed timing, and reflected
in the external volume 240.
[0203] At the stage before the write data is written in the
external volume 240, the data stored in the cache memory 130 of the
first virtualization storage device 100A and the data stored in the
external volume 240 are different. Updated data regarding a certain
segment or a segment group is stored in the cache memory 130, and
old data before the update is regarding the same segment or segment
group is stored in the external volume 240. As described above,
data that is not reflected in the external volume 240 and which
does not coincide with the memory contents of the cache memory 130
and the memory contents of the external volume 240 is referred to
as dirty data. Incidentally, data in which write data is written in
the external volume 240 and which coincides with the memory
contents of the cache memory 130 and the memory contents of the
external volume 240 is referred to as clean data. The processing of
writing and reflecting the dirty data stored in the cache memory
130 of the first virtualization storage device 100A into the
external volume 240 is referred to as destage processing.
[0204] In the present embodiment, in order to maintain the
consistency of data before and after the transfer of volume, when
the owner right is changed, the first virtualization storage device
100A, which is the transfer source, with perform destage processing
without processing the access request from the host 10.
[0205] Meanwhile, when the access target from the host 10 is a
logical volume 164 other than the transfer target (S82: YES), the
first virtualization storage device 100A identifies the type of
command (S87), and performs normal access processing.
[0206] When it is a write command, the first virtualization storage
device 100A stores the write data received from the host 10 in the
cache memory 130 (S88), and notifies the host 10 of the completion
of writing (S89). Next, while looking out for a prescribed timing,
the first virtualization storage device 100A refers to the
management table T1A, confirms the path to the external volume 240
(S90), and transfers the write data to the external volume 240
(S91).
[0207] When it is a read command, the first virtualization storage
device 100A reads the data requested from the host 10 from the
external volume 240 (S92), and transfers this data to the host 10
(S93). Incidentally, when reading data from the external volume
240, the management table T1A is referred to. Further, when the
data requested from the host 10 already exists on the cache memory
130 (when the data has been sliced), the first virtualization
storage device 100A transfers the data stored in the cache memory
130 to the host 10 without accessing the external volume 240.
[0208] FIG. 15 is a flowchart showing the details of S78 in FIG.
13. When the second virtualization storage device 100B, which is
the transfer destination, receives a command from the host 10
(S101: YES), it analyzes the access target of such command. The
second virtualization storage device 100B determines whether the
access target of the host 10 is a logical volume 164 connected to
the external volume 240 under the control of the second
virtualization storage device 100B (S102). In other words, the
second virtualization storage device 100B determines whether the
command is an access request relating to the external volume 240 in
which it has the owner right thereof.
[0209] When the second virtualization storage device 100B
determines that this is an access request relating to the volume in
which it has the owner right thereof (S102: YES), it determines
whether the destage processing performed by the first
virtualization storage device 100A regarding the external volume
240 connected to the logical volume 164 thereof is complete (S103).
In other words, the second virtualization storage device 100B
determines whether a destage completion notification has been
acquired from the first virtualization storage device 100A
regarding such volume.
[0210] When the second virtualization storage device 100B does not
have an owner right with respect to the access target of the host
10 (S102: NO), or when the second virtualization storage device
100B has the owner right but the destage processing at the transfer
source is not complete (S103: NO), the second virtualization
storage device 100B rejects the command processing (S104). This is
in order to maintain the consistency of data regarding the transfer
target volume.
[0211] Contrarily, when the second virtualization storage device
100B has the owner right regarding the access target volume from
the host 10 (S102: YES), and the destage processing at the transfer
destination regarding the volume is complete (S103: YES), the
second virtualization storage device 100B is able to perform the
normal access processing. The normal access processing performed by
the second virtualization storage device 100B is the same as the
normal access processing performed by the first virtualization
storage device 100A.
[0212] In other words, the second virtualization storage device
100B distinguishes the type of command received from the host 10
(S105). When it is a write command, the second virtualization
storage device 100B stores the write data received from the host 10
in the cache memory 130 (S106), and thereafter notifies the
completion of writing to the host 10 (S107). And, the second
virtualization storage device 100B refers to the management table
T1B, confirms the path to the external volume 240 (S108), and
transfers the write data stored in the cache memory 130 to the
external volume and writes it therein (S109).
[0213] When it is a read command, the second virtualization storage
device 100B reads the data requested from the host 10 from the
external volume 240 (or cache memory 130) (S110), and transfers
this data to the host 10 (S111).
[0214] The foregoing explanation is an example of newly introducing
the second virtualization storage device 100B to the storage
system. Next, a case of introducing the second virtualization
storage device 100B and thereafter dispersing the load is
explained.
[0215] FIG. 16 is a flowchart showing a different example of
transferring a volume between the respective virtualization storage
devices 100A, 100B.
[0216] The user is able to comprehend the operational status of the
storage system based on the monitoring result of the monitoring
unit 21. For example, when the user judges that the load of the
first virtualization storage device 100A is heavy, the user may
issue a designation so as to transfer the external volume 240 under
the control of the first virtualization storage device 100A to the
second virtualization storage device 100B via the management
terminal 20 (S121). Further, a path for accessing via the second
virtualization storage device 100B is added to the path setting
information T3 of the host 10 based on the transfer designation
from the management terminal 20.
[0217] When the first virtualization storage device 100A receives a
transfer designation from the management terminal 20, it changes
the owner right of the external volume designated as the transfer
target from "1" to "-1", and notifies this change to the second
virtualization storage device 100B (S122).
[0218] When the second virtualization storage device 100B receives
a notice from the first virtualization storage device 100A (S123),
it sets "1" in the transfer status flag relating to the transfer
target volume and updates the management table T1B (S124), and
notifies the completion of setting of the transfer status flag to
the first virtualization storage device 100A (S125).
[0219] When the first virtualization storage device 100A receives a
notice from the second virtualization storage device 100B,
similarly, it sets "1" in the transfer status flag relating to the
transfer target volume and updates the management table T1A (S126).
And, the first virtualization storage device 100A starts the
destage processing of dirty data relating to the transfer target
volume (S127).
[0220] Before the completion of the destage processing, if a
command requesting access to the transfer target logical volume 164
is issued from the host 10 (S128), the first virtualization storage
device 100A will reject such processing (S129).
[0221] When the access processing is rejected by the first
virtualization storage device 100A, the host 10 refers to the path
setting information T3 and switches the path (S130). Here, the
explanation is regarding a case of switching from the primary path
passing through the first virtualization storage device 100A to the
alternate path passing through the second virtualization storage
device 100B. After switching the path, the host 10 reissues the
command (S131). This command may be a write command or a read
command, and let it be assumed that a write command has been issued
for the sake of convenience of explanation.
[0222] When the second virtualization storage device 100B receives
a write command from the host 10 (S132), it receives write data
transmitted from the host 10 after the write command, and stores
this in the cache memory 130 (S132). After storing the write data
in the cache memory 130, the second virtualization storage device
100B reports the completion of writing to the host 10 (S133). The
host 10 receives a processing completion notice from the second
virtualization storage device 100B (S134).
[0223] Meanwhile, when the destage processing performed by the
first virtualization storage device 100A is complete (S135), the
first virtualization storage device 100A notifies the completion of
the destage processing to the second virtualization storage device
100B (S136). When the second virtualization storage device 100B
receives this destage completion notice (S137), it resets the
transfer status flag relating to the transfer target volume (S138).
Thereby, the transfer of the volume is completed while maintaining
the consistency of the volume. After the transfer of the volume is
complete, if the host 10 issues a different command (S139), the
second virtualization storage device 100B performs the normal
access processing (S140).
[0224] Incidentally, if the command issued at 5131 is a read
command, the second virtualization storage device 100B may reject
the processing of the read command until the destage processing by
the first virtualization storage device 100A is complete.
[0225] FIG. 18 is an explanatory diagram showing a frame format of
the situation of transferring the volume according to the present
embodiment. Foremost, as shown in FIG. 18(a), let it be assumed
that only the first virtualization storage device 100A is initially
operating in the storage system. Under these circumstances, the
first virtualization storage device 100A is using all external
volumes 240.
[0226] As shown in FIG. 18(b), the user determines the introduction
of the second virtualization storage device 100B based on the load
status of the first virtualization storage device 100A, and adds
the second virtualization storage device 100B to the storage
system.
[0227] As shown in FIG. 18(c) when the user designates the transfer
of the volume 240 of "#B" and "#C" via the management terminal 20,
these volumes 240 are connected to the logical volume 164 of the
second virtualization storage device 100B. More precisely, when the
user designates the transfer of a volume regarding the logical
volume 164 of the first virtualization storage device 100A, the
external volumes 240 (#B, #C) connected to the transfer target
logical volume 164 are re-connected to the logical volume 164 of
the second virtualization storage device 100B. Thereby, at least a
part of the load of the first virtualization storage device 100A
will be transferred to the second virtualization storage device
100B, and the bottleneck in the first virtualization storage device
100A can be resolved. As a result, the response performance and
efficiency of the overall storage system can be improved.
[0228] As described above, according to the present embodiment, a
plurality of virtualization storage devices 100A, 100B may be used
to manage each of the external volumes 240. Accordingly, the load
in the storage system can be dispersed and the processing
performance of the overall storage system can be improved.
[0229] In the present embodiment, the external volume 240 can be
transferred between the respective virtualization storage devices
100A, 100B without stopping the access from the host 10. Therefore,
the volume can be transferred via online without having to shut
down the host 10, and the usability will improve.
[0230] In the present embodiment, the user merely needs to make a
designation via the management terminal 20 to transfer the external
volume 240 between the respective virtualization storage devices
100A, 100B. Accordingly, in a storage system having a plurality of
virtualization storage devices 100A, 100B capable of virtualizing
and using the external volume 240, the performance of the storage
system can be improved with a relatively simple operation.
[0231] In the present embodiment, the virtualization storage device
100A, which is the transfer source, is configured such that it can
reject the access request from the host 10 until the destage
processing relating to the transfer target external volume 240 is
complete. Therefore, the volume can be transferred while
maintaining the consistency of data.
2. Second Embodiment
[0232] The second embodiment of the present invention is now
explained with reference to FIG. 19. The present embodiment
corresponds to a modified example of the foregoing first
embodiment. In the present embodiment, the storage system
autonomously disperses the load between the respective
virtualization storage devices 100A, 100B.
[0233] FIG. 19 is a flowchart of the transfer designation
processing according to the present embodiment. This transfer
designation processing, for example, can be executed with the
management terminal 20. The management terminal 20 acquires the
performance information from the respective virtualization storage
devices 100A, 100B (S161). The management terminal 20, based on
each type of performance information, respectively calculates the
loads LS1, LS2 of the respective virtualization storage devices
100A, 100B (S162). These loads, for example, may be calculated
based on the input/output per second, CPU usage rate, cache memory
usage rate and the like.
[0234] The management terminal 20 compares the load LS1 of the
first virtualization storage device 100A and the load LS2 of the
second virtualization storage device 100B (S163). When the first
load LS1 is greater than the second load LS2 (LS1>LS2), the
management terminal 20 determines the logical volume (external
volume) to the transferred from the first virtualization storage
device 100A to the second virtualization storage device 100B
(S164). The management terminal 20, for instance, may select the
volume of the highest load in the device.
[0235] The management terminal 20 judges whether the transfer
timing has arrived (S165), and, when the transfer timing has
arrived (S165: YES), it defines the path information of the
transfer destination (S166), and respectively issues a transfer
designation to the respective virtualization storage devices 100A,
100B (S166). For example, a time frame with low access frequency
from the host 10 may be pre-selected as the transfer timing.
[0236] Meanwhile, when the second load LS2 is equal to or greater
than the first load LS1 (LS1.ltoreq.LS2), the management terminal
20 determines the volume to be transferred from the second
virtualization storage device 100B to the first virtualization
storage device 100A (S168).
[0237] The management terminal 20, as described above, looks out
for a prescribed transfer timing (S169: YES), defines the path of
the transfer destination (S170), and respectively issues a transfer
designation to the respective virtualization storage devices 100A,
100B (S171).
[0238] The present embodiment configured as described above also
yields the same effects as the foregoing embodiments. In addition,
with the present embodiment, the load dispersion between the
plurality of virtualization storage devices 100A, 100B capable of
respectively virtualizing the external volume 240 can be performed
autonomously.
[0239] Incidentally, the present invention is not limited to the
embodiments described above. Those skilled in the art may make
various additions and modifications within the scope of the present
invention.
[0240] For example, in each of the foregoing embodiments, although
a case was mainly explained where a plurality of virtualization
storage devices coexists, the present invention is not limited
thereto, the configuration may also be such that all external
volumes are transferred to the second virtualization storage
device, and the first virtualization storage device may be entirely
replaced with the second virtualization storage device.
[0241] Further, in each of the foregoing embodiments, although a
case was mainly explained where the management terminal is
configured from a separate computer, the present invention is not
limited thereto, and the configuration may be such that the
function of the management terminal is built in one of the
virtualization storage devices.
[0242] Moreover, in each of the foregoing embodiments, although a
case was mainly explained where two virtualization storage devices
are used, the present invention is not limited thereto, and the
present invention may also be applied to cases of using three or
more virtualization storage devices.
[0243] Further, in each of the foregoing embodiments, although a
case was mainly explained where the virtualization storage devices
are operated in an asynchronous transfer mode, these may also be
operated in a synchronous transfer mode. When operating the
virtualization storage devices in a synchronous transfer mode,
generally speaking, since the memory contents of the external
volume will always be updated to be the latest contents, such
memory contents may be transferred between the respective
virtualization storage devices quickly without having to wait for
the completion of the destage processing at the transfer
source.
[0244] Incidentally, when transferring a volume, the logical volume
164 of the transfer source and the logical volume 164 of the
transfer destination will be set to be of the same size.
* * * * *