U.S. patent application number 12/143192 was filed with the patent office on 2008-10-23 for storage operation management program and method and a storage management computer.
Invention is credited to Masayasu ASANO, Kenichi Shimooka, Daisuke Shinohara.
Application Number | 20080263111 12/143192 |
Document ID | / |
Family ID | 33410550 |
Filed Date | 2008-10-23 |
United States Patent
Application |
20080263111 |
Kind Code |
A1 |
ASANO; Masayasu ; et
al. |
October 23, 2008 |
STORAGE OPERATION MANAGEMENT PROGRAM AND METHOD AND A STORAGE
MANAGEMENT COMPUTER
Abstract
A managing computer (manager) manages replication of data areas
inside a storage or among storages. A storage volume management
module of the manager manages policy of the volume and its
properties. When replication of a volume is set, a volume of a
replication destination appropriate for a volume of a replication
source is generated using the policy and the properties to form a
replication pair. A route management module of the manager and its
volume management module bring the policy and the properties of the
volumes, and the policies and conditions of lines into conformity
with one another. When any fault occurs in a line route used for
data transfer during replication of the volume, a separate line
route is utilized, and a fault countermeasure is taken for
replication of the volumes in accordance with the policy and
properties of the volumes.
Inventors: |
ASANO; Masayasu; (Yokohama,
JP) ; Shimooka; Kenichi; (Yokohama, JP) ;
Shinohara; Daisuke; (Yokohama, JP) |
Correspondence
Address: |
MATTINGLY, STANGER, MALUR & BRUNDIDGE, P.C.
1800 DIAGONAL ROAD, SUITE 370
ALEXANDRIA
VA
22314
US
|
Family ID: |
33410550 |
Appl. No.: |
12/143192 |
Filed: |
June 20, 2008 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11648655 |
Jan 3, 2007 |
|
|
|
12143192 |
|
|
|
|
10650851 |
Aug 29, 2003 |
7191198 |
|
|
11648655 |
|
|
|
|
Current U.S.
Class: |
1/1 ;
707/999.204; 711/E12.001; 714/E11.078; 714/E11.103 |
Current CPC
Class: |
H04L 45/00 20130101;
G06F 11/2069 20130101; Y10S 707/99952 20130101; Y10S 707/99955
20130101; G06F 11/2007 20130101; G06F 11/2071 20130101; Y10S
707/99953 20130101 |
Class at
Publication: |
707/204 ;
711/E12.001 |
International
Class: |
G06F 12/00 20060101
G06F012/00 |
Foreign Application Data
Date |
Code |
Application Number |
May 8, 2003 |
JP |
2003-130297 |
Claims
1.-17. (canceled)
18. A storage system coupled to a computer, comprising: a plurality
of storage devices comprising a plurality of volumes; and a
managing computer coupled to the plurality of storage devices
having capacity information representing volume capacities of the
plurality of volumes and read/write information representing
read/write frequencies of the plurality of volumes; wherein the
management computer receives a request designating a copy source
volume among the plurality of volumes, selecting a copy destination
volume, which capacity is the same or larger than a capacity of the
copy source volume, from the plurality of volumes based on the
capacity information, wherein a copy source storage device among
the plurality of storage devices corresponding to the copy source
volume forms a copy pair between the copy source volume and the
copy destination volume, and copies the contents of the copy source
volume to the copy destination volume, and wherein the copy source
storage device releases the copy pair after the copy, according to
a part of the read/write information corresponding to the copy
source volume.
19. A storage system according to claim 18, wherein the copy
destination volume is included in a copy destination storage
device, which is one of the plurality of storage devices, and
wherein the selection of the copy destination volume is based on a
distance between the copy source storage device and the copy
destination storage device.
20. A storage system according to claim 18, wherein the read/write
information includes a plurality of values for representing the
read/write frequencies of the plurality of volumes, each of the
plurality of values corresponding to a read/write frequency of each
of the plurality of volumes.
21. A storage system according to claim 20, wherein the part of the
read/write information corresponding to the copy source volume is
one of the plurality of values corresponding to the copy source
volume.
Description
[0001] This application is a continuation application of U.S.
application Ser. No. 11/648,655 filed Jan. 3, 2007, now allowed,
which is a continuation of U.S. application Ser. No. 10/650,851,
filed Aug. 29, 2003, now U.S. Pat. No. 7,191,198.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] This invention relates to a storage operation management
program, a storage operation management method and a storage
managing computer. More particularly, the invention relates to a
storage operation management program, a storage operation
management method and a storage managing computer each capable of
operating and managing replication of data area inside the same
storage or among storages in accordance with information of the
data areas of the storages.
[0004] 2. Description of the Related Art
[0005] Technologies such as a snapshot function for doubling data
inside a storage (e.g. JP-A-2001-306407) and a remote copy function
for doubling data among storages in consideration of backup to cope
with disasters are known as prior art technologies for improving
versatility of computer systems and for achieving storages capable
of backing up data for a non-stop operation of the computer
systems.
[0006] The snapshot function is the one that makes it possible to
conduct data backup on the on-line basis while continuing the
business operations in a computer system that is to continuously
operate for 24 hours and for 365 days without interruption.
[0007] According to the remote copy function, a storage of a
computer system operating normally transfers updating data to a
storage of a computer system installed and operating at a different
remote place so that the data stored in the storage of the remote
place can be brought substantially equivalent to the data stored in
the storage of the computer system normally operating, and the loss
of data can be minimized even when any accident occurs in the
computer system normally operating.
[0008] To efficiently process a plurality of remote copy requests
in the remote copy function, a technology is known that decides a
remote copy schedule and a route of a line in consideration of
access frequency of data areas and a condition of replication
line.
[0009] All the prior art technologies described above are directed
to reduce a processing load of a replication processing from the
data area of a replication source to the data area of a replication
destination and to improve transfer efficiency. In other words,
these technologies are based on the premise that the data area of
the replication source and the data area of the replication
destination coexist.
[0010] On the other hand, the number of computer systems that use a
plurality of storages connected with one another through SAN
(Storage Area Network) or IP (Internet Protocol) by use of a
dedicated storage network (mainly, Fibre Channel) and share data of
a large capacity distributively stored in these storages has
increased with the increase of the data capacity. Among the
storages connected by SAN, etc, some of them are sometimes products
of different manufacturers or have different performance such as an
access speed. It has thus become more difficult to select, generate
and designate the data area having reliability and the access speed
corresponding to those of policy of users and applications such as
the policy for a financial system, a policy for various management
businesses for companies, a policy for databases, and so forth. A
prior art technology that makes it possible to generate a data area
coincident with the policy of the user is also known.
SUMMARY OF THE INVENTION
[0011] However, the prior art technologies described above do not
consider generation of a data area of a replication source and a
data area of a replication destination for doubling data by
snapshot or remote copy.
[0012] To efficiently set data doubling in accordance with user
policy from among various storages connected to SAN or IP, it
becomes necessary to generate a data area of a replication
destination similar to a data area of a replication source that
satisfies reliability and an access speed of the user policy such
as the financial system, various management businesses, databases,
etc, described above and to generate a replication pair of the data
areas. When any fault occurs in the replication operation for
doubling data in the business such as in the financial system for
which data reliability is severely required, the prior art
technologies described above executes again replication after the
fault is restored. It may be possible to change a route of
replication from the data area of the replication source to the
data area of the replication destination in data doubling.
Depending on the policy of the data area and on the operation
policy of data doubling as a whole, however, it is necessary to
decide the operation by judging whether the route of replication is
to be changed or to wait for the restoration of the fault.
Consequently, it is difficult to efficiently operate the overall
system.
[0013] In view of the problems described above, it is an object of
the invention to provide a storage operation management program, a
storage operation management method and a storage managing computer
each capable of operating and managing replication of data areas
inside the same storage or among storages in accordance with
information of the data areas of the storages.
[0014] According to a feature of the invention, a storage operation
management program for operating and managing replication of data
areas inside a storage or among a plurality of storages comprises a
processing step of accepting a request for generation of a data
area of a replication destination for a data area of a replication
source; a process step of retrieving a data area capable of
becoming a replication destination coincident with properties of a
data area corresponding to policy of the data area of the
replication source from existing data areas; and a process step of
instructing the storage to generate a replication pair of the data
areas.
[0015] According to another feature of the invention, a storage
operation management method for operating and managing replication
of data areas inside a storage or among a plurality of storages
comprises a processing step of accepting a request of generation of
a data area of a replication destination for a data area of a
replication source; a process step of retrieving a data area
capable of becoming the replication destination coincident with
properties of a data area corresponding to policy of the data area
of the replication source from existing data areas; and a process
step of generating a replication pair on the basis of the retrieval
result.
[0016] More concretely, in an aspect of the invention, when a data
area for managing and replicating policy of a data area and its
properties and a data area for replication is selected, the policy
and properties of a data area of a replication source are acquired,
a data area of a replication destination coincident with policy and
properties of the data area of the replication source is generated
and a replication pair of the data areas is generated.
Consequently, an operation of data doubling in accordance with the
policy and properties of the data area becomes explicitly possible,
and automation of setting of data doubling in accordance with the
policy of the data area and data doubling can be effectively
constituted from among a plurality of storages connected through
SAN or IP.
[0017] In another aspect of the invention, the policy of the data
area of the replication source is brought into conformity in the
aspect of the operation, too. When any fault or delay of a
replication process occurs in a route connecting a data area of a
replication source and a data area of a replication destination in
the case of remote copy, for example, judgment is made in
accordance with the policy of the data area of the replication
source as to whether another connectable route is set, or data to
be doubled is stored in a cache inside a storage to wait for
restoration of the fault of a line, or replication data having low
priority is omitted. Consequently, the operation of data doubling
can be effectively made.
[0018] According to the invention, the resolution method described
above manages the policy of the data areas of the storages and
their property, can conduct automatic setting of data doubling in
accordance with the policy of the data area by utilizing the policy
and can operate data doubling flexibly and explicitly for
users.
[0019] Other objects, features and advantages of the invention will
become apparent from the following description of the embodiments
of the invention taken in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0020] FIG. 1 is a block diagram showing a construction of a
storage operation management system according to a first embodiment
of the invention;
[0021] FIG. 2 is an explanatory view useful for explaining a table
group managed by a volume management module 105;
[0022] FIG. 3 is a flowchart useful for explaining a processing
operation for setting replication of volumes in accordance with
policy of a volume of replication source and generating a pair of
the volume of the replication source and a volume of the
replication destination;
[0023] FIG. 4 is a block diagram showing a construction of a
storage operation management system according to a second
embodiment of the invention;
[0024] FIG. 5 is an explanatory view useful for explaining a table
group managed by a route management module 401;
[0025] FIG. 6 is a flowchart useful for explaining a processing
operation for generating a replication in accordance with policy of
a line for connecting volumes when a volume replication is
generated between storages in the second embodiment of the
invention;
[0026] FIG. 7 is a flowchart useful for explaining a processing
operation of fault shooting when any fault occurs in a route of
replication during the volume replication operation between the
storages;
[0027] FIG. 8 is a flowchart useful for explaining a processing
operation for generating a replication pair in another storage;
and
[0028] FIG. 9 is a block diagram showing a construction of a
storage operation management system according to a third embodiment
of the invention.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0029] A construction of a storage operation management system and
an operation management method according to preferred embodiments
of the invention will be hereinafter explained in detail with
reference to the accompanying drawings.
[0030] FIG. 1 is a block diagram showing a construction of a
storage operation management system according to a first embodiment
of the invention.
[0031] The storage operation management system according to the
first embodiment of the invention includes a managing computer
(manager) 100 and a plurality of storages 11a to 11n that are
connected to one another through a data network 130 and through a
network 140. The storage 11a includes volumes (data areas,
hereinafter called merely "volumes") 113a, 113b, . . . , 113n as
management areas for storing data that are managed in practice by a
computer, a space area 114 that will be divided and set as volumes
in future, data communication equipment 110 for transmitting and
receiving data I/O of read/write data of the volumes, network
communication equipment 111 for communicating with the managing
computer 100, etc, and a controller 112 for practically controlling
the storages. The data communication equipment 110 and the network
communication equipment 111 may be constituted and arranged as one
communication equipment by connecting them together through IP
(Internet Protocol) connection in a connection form (such as
Ethernet (trade mark)).
[0032] The storage 11b, too, includes volumes 123a, 123b, . . . ,
123n, a space area 124, data communication equipment 120, network
communication equipment 121 and a controller 122 in the same way as
the storage 11a. The other storages 11b, . . . , 11n have also the
same construction as the storage 11a.
[0033] The data network 130 is a cable as a route of data
communication of the storages and the computers. The network 140 is
a route that can communicate with the managing computer 100, etc.
These networks may be either mere buses or LAN. The data network
130 and the network 140 are the same with each other depending on
the communication form and may be Ethernet (trademark), for
example.
[0034] The managing computer 100 includes a memory 101, a CPU 102,
network communication equipment 103 and a storage device 104. The
storage device 104 includes therein a volume management module 105
and a later-appearing volume management module table group 200
managed by the volume management module 105.
[0035] The volume management module 105 achieves the processing in
the embodiment of the invention. The volume management module 105
includes a volume generation part 106 for instructing generation of
the volumes to the storages 11a, 11b, . . . , 11n, a volume pair
generation part 107 for generating a replication pair from among
the volumes generated by the instruction of the volume generation
part 106, and a volume pair selection part 108 for selecting the
volume pair that can be replicated from policy of the volumes (for
replication source, for replication destination, for both, etc) and
their properties (performance, reliability, etc) and from policy of
the storages (for financial systems, for various management
businesses of companies, for database, etc) and their properties
(performance, reliability, etc). The volume management module 105
is accomplished when software, not shown, stored in the storage
device of the managing computer 100 is written into the memory 101
and is executed by the CPU 102.
[0036] When the storage operation management system shown in FIG. 1
replicates the volumes between the storages, such as when a volume
113a of the storage 11a is a volume of the replication source and
the volume 123a of the storage 11b is a volume of a replication
destination, the controller 112 of the storage 11a and the
controller 122 of the storage 11b cooperate with each other,
transmit the content of the volume 113a to the volume 123a through
the data network 130 and replicate the content. In this case, the
respective computers using the volume 113a and the volume 123a and
the controllers 112 and 122 may cooperate with one another and
conduct replication.
[0037] FIG. 2 explains the table group that the volume management
module 105 manages. The volume management module table group 200
managed by the volume management module 105 includes a volume
information table 210, a storage information table 220 and a pair
information table 230.
[0038] The volume information table 210 stores a volume ID 211
allocated to identify all the volumes of the storages managed by
the managing computer 100, a storage ID 212 representing an
identifier of the storage to which the volume belongs, a storage
volume ID 213 representing the identifier of the volume managed
inside each storage in the storage, a policy 214 representing the
use policy of the volume such as for the replication source, for
the replication destination, for both replication source and
destination, for the replication source for remote copy, each being
designated by the user and the application, a volume capacity 215,
Read/Write 216 representing read/write frequency from and to the
volume set by the policy of the user and the application (for
financial system, for various management businesses of companies,
for database, etc), performance 217 representing a read/write speed
of the volume, reliability 218 representing reliability of the
volume in terms of a numerical value, and pairing
possible/impossible information 219 representing whether or not the
volume can be set as the volume that can be paired as the
replication source with volume of the replication destination.
[0039] The storage ID 212 is an identifier of each storage 11a,
11b, . . . , 11n. The storage volume ID 213 is an identifier of
each volume 113a, 113, . . . , 113n. Values of from 0 to 10 are
entered as Read/Write information 216. In the case of the value 1,
the policy is limited to only Read. As the value increases from 2,
frequency of Write becomes higher. In the case of 10, the policy is
limited to only Write. Values acquired by normalizing performance
such as a read/write speed, etc between 1 and 10, that is, values
of from 1 to 10, are entered to the performance information 217
representing performance of the volume. In the case of the value 1,
access performance to the volume is the lowest and in the case of
10, access performance reaches maximum. Reliability information 218
representing reliability of the volume in terms of a numerical
value is acquired by normalizing reliability between 1 and 10, that
is, values from 1 to 10, are entered. The value 1 represents the
highest degree of fault occurrence of the volume and the value 10
represents the lowest degree of fault occurrence. When setting is
made in only the RAID level, for example, the managing computer 100
decides a rule such that the volume of RAID0 is 1, RAID1 is 10,
RAID5 is 5, and so forth, and executes the management operation. In
other words, Read/Write information 216, performance information
217 and reliability information 218 represent the properties of the
corresponding volume.
[0040] The storage information table 220 stores information
representing each of the storage ID 221 as the identifier of the
storage managed by the managing computer 100, a space capacity 222
representing the capacity of a space area the storage does not yet
set and use as the volume, policy 223 to be given to the storage,
maximum reliability 224 that can be achieved when the storage
generates the volume by use of the space area and maximum
performance 225 that can be achieved when the storage generates the
volume by use of the space area.
[0041] In the case of the storage 11a, for example, the value of
the space capacity 222 represents the capacity of the space area
114. When the storage ID 221 has the same value as the value of the
storage ID 212 of the volume information table 210, it represents
the same storage.
[0042] The pair information table 230 represents information of
replication of the volumes the managing computer 100 manages. This
table includes information of each of a pair ID 231 for identifying
pair information, a main volume ID 232 representing a volume ID of
the volume of the replication source, a sub-volume ID 233
representing a volume ID of the volume of the replication
destination and a replication type 234 representing a replication
type.
[0043] The value of the main volume ID 232 and the value of the
sub-volume 233 represent the same volume when the values of the
volume ID in the volume information table 210 are the same. A value
"synchronous" or a value "asynchronous" is entered to the
replication type 234. In the case of "synchronous", replication to
the volume of the replication destination is made when write occurs
in the volume of the replication source. In the case of
"asynchronous", replication is made in a unit of a certain schedule
(every predetermined time, data quantity per replication). The user
or the application may set this replication type. Alternatively,
the replication type may be set as a part of the processing of the
volume management module that will be explained next.
[0044] FIG. 3 is a flowchart useful for explaining a processing
operation for setting replication of the volume in accordance with
the policy of the volume of the replication source and then
generating the pair of the volume of the replication source and the
volume of the replication destination. Next, this flowchart will be
explained. The volume management module 105 executes the processing
that will be explained with reference to this flowchart.
[0045] First, the volume management module 105 acquires the pair
generation request of the volume to be replicated with the ID of
the volume of the replication source from the user and the
application (Step 300).
[0046] At this time, the volume management module 105 may start
executing the processing of Step 300 after the volume of the
replication source is generated but before it receives the ID of
the volume. In other words, when receiving the replication
generation request, the volume management module 105 may create the
volume of the replication source and the volume of the replication
destination during a series of processing and may execute a
plurality of settings. Alternatively, the volume management module
105 may acquire a request that the replication is created by use of
the volumes inside the storage or the volumes between the
storages.
[0047] Next, information of the policy 223 of the storage is
retrieved on the basis of the storage ID, and whether or not the
storage having a volume capable of forming the pair with the volume
of the replication source exists is judged. When the storage has
the policy not capable of creating the volume as the replication
source, the flow shifts to Step 308 and if not, to Step 302 (Step
301).
[0048] When the request for setting replication of the volumes
exists between the storages in Step 301 and when all the storages
other than the storage to which the volume of the replication
source belongs, that is, all the storages other than the storage to
which the volume of the replication source registered to the
storage information table 200, have registered their policy to the
limited policy of "for main storage" that generates only the volume
of the replication source to the policy 223 as typified by the
storage of the storage ID3 in the storage information table in FIG.
2, the flow proceeds to Step 308 to notify that the storage capable
of forming the volume of the replication destination does not
exist. The flow may proceed to next Step 302 when the storage
information table 220 does not manage the information of policy
223.
[0049] When the storage capable of forming the pair is found
existing in the volume as a result of judgment in Step 301, whether
or not that storage has the policy of forming the pair is judged.
In other words, when the volumes having the policy not capable of
becoming the volume of the replication destination other than the
volumes belonging to the storage having the policy not capable of
forming the volume of the replication destination are retrieved and
when all the volumes so retrieved cannot become the volume of the
replication destination, the flow proceeds to Step 305 and if not,
the flow proceeds to Step 303 (Step 302).
[0050] In Step 302, when the request for setting replication of the
volume exists between the storages, for example, and when the
volume of the replication destination cannot be designated in case
all the policies of the volumes other than the storage to which the
volume of the replication source belongs are "for replication
source" as typified by the volume ID 1 in the volume information
table 210, the volume of the replication source cannot be
designated and the flow therefore proceeds to Step 305. If not, the
flow proceeds to Step 303. When replication of the volume is set
inside the storage, the volume of the replication destination
cannot be designated if all the policies of the volumes of the
storage to which the volume of the replication source belongs are
"for main volume". Therefore, the flow proceeds to Step 305 and to
Step 303 when the policy is not "for main volume".
[0051] When the storage is found having the policy of forming the
pair in the judgment of Step 302, whether or not the volume capable
of forming the pair exists is judged. In other words, the volumes
that cannot be used at present as the replication destination other
than the volumes that cannot form the volume of the replication
destination are retrieved on the basis of information of pairing
impossible/impossible 219. When the object volumes all have the
information "impossible" in the pairing possible/impossible
information 219 as represented by the volume ID2 of the volume
information table, the existing volumes managed by the managing
computer cannot accomplish the replication destination. Therefore,
the flow proceeds to Step 305, and to Step 304 when they can (Step
303).
[0052] When the volume capable of forming the pair is found
existing in the judgment of Step 303, whether or not the properties
accessorial to the volume, that is, the values of capacity 215,
Read/Write 216, performance 217 and reliability 218 of the volume
information table 210 are coincident are checked. The flow proceeds
to Step 307 when the coincident volume exists and to Step 305 when
not. The term "coincidence" hereby used means in principle that the
values are the same. Even when the volume having properties
coincident with these properties does not exist, a volume having
the property values approximate to these values may be used
depending on the operation principle. As to the values of capacity
215, performance 217 and reliability 218 described above, the
values of the replication destination may be the same or greater.
For example, when the volume having a value 3 as the value of the
volume ID 211 of the volume information table 210 is designated as
the volume of the replication source during the processing of Steps
300 to 304, it becomes possible to set the volume having a value 6
of the volume ID 211 of the volume information table 210 to the
volume of the replication destination. In other words, it is
possible to set the volume having properties that are the same as
or greater than the properties accessorial to the replication
source to the volume of the replication destination (Step 304).
[0053] When a volume coincident with the properties accessorial to
the volume described above is not found in the judgment of Step
304, it means that no corresponding volume as the volume of the
replication destination exists among the existing volumes managed
in the volume information table. Therefore, whether or not this
corresponding volume can be generated from the space area of the
storage is judged. In other words, whether or not the volume
coincident with the properties of the volume of the replication
source can be generated is checked from the storage information
table 220 with the exception of the storages not having the space
area, and whether or not the volume coincident with the volume of
the replication source can be generated is checked from capacity
222, reliability 224 and maximum performance 225 inside the storage
information table 220. When it can, the flow proceeds to Step 306
and when it cannot, the flow proceeds to Step 308. When the volume
having the volume ID 211 value of 7 in the volume information table
210 is designated as the volume of the replication source, for
example, other volumes cannot be set as the volume of the
replication destination in case that the properties are judged as
non-coincident and designation of the volume is judged as
impossible in Step 304. Therefore, this processing judges whether
or not a volume of the replication source can be generated afresh
from the space area is judged. At this time, it is possible to know
from the information of the storage information table 220 that the
volume of the replication destination can be achieved from the
storage of the storage ID4 (Step 305).
[0054] When the judgment result of Step 305 represents that the
corresponding volume can be generated from the space area, the
instruction is given to the controller of the storage to generate
the volume of the replication destination coincident with the
volume of the replication source, and the volume management table
210 and the storage information table 220 are updated. The
controller of the storage generates the volume in accordance with
the instruction described above. When the volume of the volume ID7
is the replication source as in the example taken in Step 305, the
volume having the properties coincident with those of the volume of
the volume ID7 is generated from the storage of the storage ID4,
and that volume is registered to the volume information table (Step
306).
[0055] When the volume coincident with the volume of the
replication source is found existing among the existing volumes
after the processing of Step 306 or in the judgment of Step 304,
the replication pair is generated for the storage of the
replication source and the storage of the replication destination
by use of the corresponding volume as the replication source, and
the volume of the replication source and the volume of the
replication destination set afresh are registered to the pair
information table 230. The processing is then finished (Step
307).
[0056] When the storage capable of forming the pair is not found in
the judgment of Step 301 or when the judgment result of Step 305
represents that the corresponding volume cannot be generate from
the space area, the effect that the pair cannot be set with volumes
of the replication destination indicated by the information 219
representing whether or not the volume can be set as the volume of
the replication destination of the volume information table is
indicated, and the processing is finished (Step 308).
[0057] When the processing of Step 307 described above is the
processing for the volume in which the value of Read/Write 216 of
the properties of the volume is only Read, a subsequent replication
operation does not exist once replication is made. Therefore, a
report may be given to notify that setting of replication is
released after replication is finished once. In the case of
replication of the volume in which Read occupies the major
proportion, replication does not occur so frequently. Therefore,
this property may be notified to the user as the factor for making
the asynchronous replication schedule. The change of the
asynchronous schedule may be urged to the user depending on the
policy of the volume and on the load of replication.
[0058] Because the processing of Step 308 described above fails to
provide the volume of the replication source, the report of the
failure is made. In this case, a message may be outputted on the
basis of the condition of the point of the occurrence of the
failure. When Step 301 fails, for example, a message to the effect
that "storage capable of generating volume of replication
destination does not exist" may be additionally outputted.
[0059] In the processing described above, the volume pair
generation part 107 executes the processing of Steps 300 and 307,
the volume generation part 106 executes the processing of Step 306
and the volume pair selection part 108 executes the processing of
Steps 301 to 305 and 308.
[0060] This embodiment executes the processing described above and
can generate the replication by setting the volume of the
replication destination in accordance with the policy of the volume
of the replication source and its properties.
[0061] When a request for generating a plurality of pairs exists in
a series of processing described above, the processing explained
with reference to FIG. 3 is executed either repeatedly or in
parallel and setting of the pairs is executed so that all the
designated pairs can be generated. When any error occurs during the
generation process of the plurality of pairs and the processing of
Step 308 is therefore executed, the pairs that have been set so far
are omitted.
[0062] The judgment steps from Steps 302 to 304 in the series of
the process steps described above are the sequence for finding out
the storage of the replication destination from the existing
volumes that have already been generated. The process steps of
Steps 305 and 306 are the sequence for generating afresh the volume
of the replication destination. However, it is possible to change
the process steps in such a fashion that the processing of Steps
305 and 306 is executed after Step 301, and then the processing
proceeds to Step 302 when Step 305 has the condition in which
volume cannot be generated and to Step 308 when Step 305 has the
condition in which the volume cannot be selected. When the volume
cannot be selected in the sequence of Steps 302 to 304 after Step
301, it is also possible to change the processing in such a fashion
that the processing proceeds to Step 308 and is then finished.
[0063] The processing described above may further be changed in
such a fashion as to proceed to Step 305 after Step 301 and then to
Step 308 when the volume cannot be generated, and to terminate the
processing. Furthermore, the processing of Steps 302 to 304 and the
processing of Steps 305 and 306 may be executed in parallel after
Step 301, and the volume that can be set as the replication
destination may be displayed on the screen of the managing computer
for the submission to the user and to the application.
[0064] When the volume of the replication source and the volume of
the replication destination are generated as the replication pair
in Step 307 in the processing described above, the connection
distance between the storage to which the volume of the replication
source belongs and the storage to which the volume of the
replication destination belongs may be added as a condition in
accordance with the policy of the volume of the replication source
and its properties. For example, when the policy of the volume of
the replication source is "volume for which most important data
must be secured at the time of accident", the volume having the
greatest inter-storage distance is preferentially selected.
Further, the site of the storage and its position of either one, or
both, of the replication source and the replication destination may
be used as a condition for generating the replication pair of the
volumes of both replication source and destination. For example,
when the policy of the volume of the replication source is "volume
whose leak is inhibited by company rule or law", the generation of
the replication pair is permitted only in a specific country, a
specific city, a specific company and a specific department. The
policy of such a volume may be stored in and managed by the volume
information table 210.
[0065] FIG. 4 is a block diagram showing a construction of storage
operation management system according to a second embodiment of the
invention. When a plurality of connection methods for connecting
storages exists in the volume replication operation between the
storages, the second embodiment of the invention hereby explained
executes replication in accordance with a condition of a route of a
volume and its policy.
[0066] In the construction of the first storage operation
management system of the invention shown in FIG. 1, the second
embodiment shown in FIG. 4 employs a construction in which a route
management module 401 and a later-appearing route management module
table group managed by the path management module are added to the
managing computer 100 of the first embodiment, a cache 415, 425,
435 is further added to each of the plurality of storages 11a, 11b
and 11c, and switches A440 and B450 are disposed to mutually
connect the storages 11a, 11b and 11c, and data networks 460 to 464
are further added. Incidentally, though FIG. 4 shows only three
storages 11a, 11b and 11c, they are shown as the representatives of
the storages 11a, 11b, . . . , 11n shown in FIG. 1. The switches
A440 and B450 and the data networks 461 and 463 for connecting
these switches may use a public communication network.
[0067] The route management module 401 accomplishes the processing
in the second embodiment of the invention. The processing is
accomplished as software stored inside the storage device 104 of
the managing computer 100 is read into the memory 101 and is
executed by the CPU 102. The data route between the storages is
accomplished through the switch A440, the switch B450 and the data
networks 460 to 464. The data networks 460 to 464 are the cables in
the same way as the network 130 shown in FIG. 1. The storage 11a,
for example, can be connected to the storage 11b through the data
network 460, the switch A440, the data network 461, the switch B450
and the data network 462. The storage 11a can also be connected to
the storage 11b through another route extending from the data
network 460, the switch A440, the data network 463, the switch B450
to the data network 462 as a different route.
[0068] FIG. 5 is an explanatory view useful for explaining the
table group that the route management module 401 manages. The route
management module table group 500 managed by the route management
module 401 includes a route information table 510 and a cache
information table 520.
[0069] The route information table 510 is a table for managing
information of the route of the data network used for replicating
the volume. This table 510 includes a route ID 511 as an identifier
representing the replication route between the storages, an actual
route 512 by cable information such as a cable name as the network
for generating the route, a condition 513 representing the
condition of the route 513, an associated storage 514 representing
the storage connected to the route, an associated pair 515 for
identifying a replication pair of the volumes using the route and a
policy 516 representing the properties of the route. The managing
computer 100 acquires the condition of the route from the switches
A and B, the storages, etc, and manages this route information
table 510. Incidentally, information of the security level of the
route may be stored in and managed by the route information table
510, and the policy 516 may store a concrete maximum transfer
speed, etc, of the route instead of the conditions "high speed" and
"low speed".
[0070] In the explanation described above, the route 512 represents
the route between the storages to which the replication pair is set
or which have the possibility of setting. In the case of the
example shown in the drawing, a route having the route ID 1 is
represented as the route that connects the storage 11a to the
storage 11b through the data network 460, the switch A440, the data
network 461, the switch B450 and the data network 462.
[0071] The condition 513 registers the condition as to whether or
not each route operates normally. In this embodiment, the condition
includes "normal" and "abnormal", and the condition in which the
line does not operate normally is registered as "abnormal". When
the load to the line becomes high and the line cannot satisfy the
policy, the condition may also be changed to "abnormal". In this
case, it is preferred that the route information table manages the
value of the network load at which the policy cannot be
satisfied.
[0072] When the values of the pair ID 231 of the pair information
table 230 are the same, the value of the associated pair 515
presents the same volume replication pair.
[0073] The cache information table 520 is a table for managing a
cache use ratio prepared for speeding up replication, that is,
information of use capacity/maximum capacity, for each storage and
stores the storage ID 521 and information of the cache use ratio
522 of each storage. The caches in this instance correspond to the
caches 415, 425 and 435.
[0074] FIG. 6 is a flowchart useful for explaining the processing
for generating the replication depending on the policy of the line
for connecting the volumes when volume replication is made between
the storages in the second embodiment of the invention. Next, this
processing will be explained. The processing shown in this
flowchart is the one that is contained in the volume management
module 105 and in the route management module 401.
[0075] First of all, the volume management module 105 acquires the
generation request of the volume pair to be replicated from the
user and from the application together with the ID of the volume of
the replication source (Step 600).
[0076] At this time, the volume management module 105 may start
processing Step 600 after the volume of the replication source is
generated but before it receives the ID of the volume. In other
words, when receiving the replication generation request, the
volume management module 105 may generate the volume of the
replication source and the volume of the replication destination in
a series of processing and may set replication. Alternatively, the
volume management module 105 may acquire the request that the
replication is created by use of the volumes inside the storages or
the volumes between the storages.
[0077] Next, the volume management module 105 retrieves information
of the policy 223 of the volume of the replication source acquired
in Step 600, and causes the route management module 401 to retrieve
the policy of the line between the storage to which the volume of
the replication source belongs and the volume that can be
connected. Next, whether or not the line that can accomplish the
policy of the volume of the replication source exists is judged.
When the line having the policy capable of generating the volume of
the replication destination exists, the flow proceeds to the
processing of Step 602 and to Step 603 when such a line does not
exist (Step 601).
[0078] When the generation request of the volumes exists between
the storages and the volume of the replication source has the
policy "Connection is to be made through high-speed line in
replication of volumes" in the processing of Step 601, for example,
the storage ID to which the volume of the replication source
belongs is entered into the associated storage 514 of the route
information table 510 of the line retrieved, and the route having
the line that can be used as the "high-speed" policy is retrieved.
When the corresponding line exists, the processing proceeds to Step
602 and when the corresponding line does not exist, to the
processing of Step 603.
[0079] When the corresponding line exists in the judgment of Step
601, the processing shifts to the processing of Step 301 and to the
following Steps explained with reference to the flow of FIG. 3. In
this case, the processing from Step 301 is executed for the
associated storage acquired in Step 601 (Step 602).
[0080] When the corresponding line is not found in the judgment of
Step 601, the volume of the replication destination cannot be
provided, and the report is made to this effect. In this case, a
message may be outputted on the basis of the condition of the
occurrence point of the error. For example, a message "Line capable
of generating volume of replication destination does not exist" may
be additionally outputted.
[0081] When the processing described above is executed, the
connectable line can be selected in accordance with the policy of
the line and with the policy of the volume designated as the
replication source.
[0082] FIG. 7 is a flowchart useful for explaining the processing
operation of fault shooting when any fault occurs in the
replication route during the volume replication between the
storages. Next, this operation will be explained. The route
management module 401 executes this operation. However, the table
of the table group 200 managed by the volume management module 105
is sometimes called out, and the volume management module 105
operates in this case.
[0083] It will be assumed that the fault occurs in a plurality of
routes during the volume replication operation between the storages
and this route fault is detected. For example, it will be assumed
that setting of replication exists with the volume 113a of the
storage 11a in FIG. 4 being the replication source and the volume
123a of the storage 11b being the replication destination. When the
replication type 234 of the pair information table 230 is
"synchronous" and the Write request is raised for the volume 114a,
or when the replication type 234 is "asynchronous" and the user or
the application designates the start of replication, the controller
of the storage start replication. When the controller of the
storage or the switch A440 or B450 detects the fault of the route
after the start of this replication, detection of the fault is
notified to the managing computer 100 through the storage network
communication equipment and the network, and the managing computer
100 receives the communication content (Step 700).
[0084] The route fault includes the case where the fault is
notified from the storage or the switch and the case where the
managing computer 100 periodically asks the fault of the storage or
the switch. When the condition of the route having the value 1 of
the route ID 511 is "abnormal" as shown in the route information
table 510, for example, the managing computer 100 receives the
route fault in Step 700.
[0085] When the route fault is detected, whether or not another
route should be searched for the volume 113a of the replication
source failing replication is judged depending on the policy 213 of
the volume ID 211 corresponding to the volume 113a (Step 701).
[0086] When the search of another route is judged as necessary, the
flow proceeds to Step 702 and when not, to Step 706. When the
volume is the one having high importance in the policy 213 of the
volume, for example, replication must be made quickly and the flow
may proceed at this time to Step 702. When the policy of the volume
is not particularly important, the flow may proceed to Step
706.
[0087] When the judgment of Step 701 represents that another route
must be searched, whether or not another route capable of reaching
the same storage exists among the routes managed by the managing
computer 100 is checked (Step 702).
[0088] At this time, the managing computer 100 may give an
instruction to search data capable of being actually connected
through the switch or the storage. Alternatively, it is possible to
register in advance the storage route and to utilize this
information. In this case, it is only necessary to register the set
of the storages for each route as another attribute to the route
management table 510. When the route exists and is not registered
to the route management table 510, the route ID 511, the route 512,
the condition 513, the associated storage 514 and the service 516
are registered. When the route exists, the flow proceeds to Step
703 and when not, to Step 706.
[0089] Incidentally, the policy 516 may be acquired from each
switch, or the user or the application may set the policy 516. It
will be assumed, for example, that the fault occurs in the route of
the data network 461 in the route connecting the storage 11a and
the storage 11b through the data network 460, the switch A440, the
data network 461, the switch B450 and the data network 462 in the
construction shown in FIG. 4, it is possible to set another route
for connecting the storages 11a and 11b through the data network
460, the switch A440, the data network 463, the switch B450 and the
data network 462.
[0090] When the route exists in Step 702, the policy of the volume
of the replication destination and the policy of the route are
compared on the basis of the policy of the network and whether or
not the policies are coincident is judged. When they are
coincident, the flow proceeds to Step 704 and when not, to Step 706
(Step 703).
[0091] When the replication request of the volume of the
replication destination is "high speed" and when the policy of the
route is "low speed" in this case, for example, the requirement
cannot be satisfied and the flow proceeds to Step 706. However,
when there is the policy in which replication is more preferably
continued even at the low line speed in accordance with the
condition of fault, replication may be carried out by use of the
low speed line. This request may be registered to the policy 213 of
the volume information table, or the user or the application may
give this request instruction as a part of the processing of Step
703.
[0092] When the policies are found coincident in the judgment of
Step 703, setting of the normal route is requested for the switches
and the storages, and the volumes mutually confirm that the
replication processing is possible. When a plurality of routes
exists, a judgment may be made to automatically select one route in
accordance with the policy, or the user or the application may be
allowed to judge by providing a plurality of results (Step
704).
[0093] Next, the managing computer gives the start instruction of
replication to the storages by use of the route set in Step 704
(Step 705).
[0094] When the judgment result proves NO in the judgment of Steps
701, 702 and 703, whether or not the replication data should be
permanently stored in the cache is judged in accordance with the
policy of the volume. In other words, when the use ratio of the
cache is high as in the storage of the storage ID3 of the cache
information table, overall efficiency of the policy will drop when
the replication data of the failed volume having low policy is
permanently stored. Therefore, judgment is made that such a
replicated data be omitted. When the judgment result represents
that the replication data should be permanently stored, the flow
proceeds to Step 708. When not, the flow proceeds to Step 707 (Step
706).
[0095] When the judgment result of Step 706 represents that the
replication data should not be stored permanently in the cache, the
managing computer 100 gives the instruction to the controller of
the storage to omit the cache data of the volume to be omitted from
the cache and to reduce the use ratio of the cache (Step 707).
[0096] When the judgment result represents that the replication
data should be permanently stored in the cache after the processing
of Step 707 or in Step 706, the route fault is reported and the
processing is finished (Step 708).
[0097] When a fault of replication occurs in the processing
explained above, replication is sometimes made through a route
different from the registered route. In this case, the processing
capacity of the line of the route is sometimes affected depending
on the other replication processing. Therefore, even when the
policy of the route is "high speed", the invention checks whether
or not the policy can be kept if a processing of replication
increases. When the policy can be kept, the invention can set so as
to execute a replication processing in which the fault occurs as
another route.
[0098] As explained above, the second embodiment of the invention
can judge whether or not replication should be made through another
route when any fault occurs in a plurality of routes, or whether or
not the content of the cache should be omitted depending on the
policy of the volume, and can thus make efficient replication of
the volumes.
[0099] FIG. 8 is a flowchart useful for explaining the processing
operation for generating the replication pair in another storage
when another route cannot be secured or when replication cannot be
conducted even when the replication data is permanently stored in
the cache. Next, this processing will be explained. The processing
shown in this flowchart is the one that is contained in the volume
management module 105 and the route management module 401. The
processing explained hereby generates separately the replication
pair of the volumes in replication of the volumes for which fault
recovery cannot be made.
[0100] In the pair after the fault occurs, whether or not fault
recovery is impossible or a drastic delay occurs is judged. This
judgment is made by setting a threshold value to the time from the
occurrence of the volume fault to recovery of the fault of the
route (Step 800).
[0101] Next, whether or not the replication may be generated by use
of other storages is judged from the policy 214 of the volumes of
the volume information table 210. When the judgment result
represents that other storages are not utilized, such as when the
policy inhibits replication to other storages, the processing is as
such finished without doing anything at all (Step 801).
[0102] When the judgment result of Step 801 represents that setting
of replication may well be set to other storages, the storages that
can be registered as the volume replication route are searched, and
whether or not such storages exist is judged (Step 802).
[0103] When the policy relating to the distance is set in this
judgment, judgment may be made to the effect that the policy is
contradictory even when the route can be registered. The flow
proceeds to Step 803 when the storages having the route exist. The
processing is finished when the route does not exist. When the
route is under the condition represented by the route information
table 510 and the condition of the route 1 is "abnormal", for
example, the route 3 represents the route the condition of which is
normal and which satisfies the policy.
[0104] When the judgment result in Step 802 represents that the
storage capable of setting the route exists, the pair is generated.
This pair generation processing is the one that executes the flow
explained with reference to FIG. 3. The pair generation processing
is executed in accordance with the instruction that the processing
should be executed within the range of the storages obtained by the
processing of Step 802 (Step 803).
[0105] In the processing described above, the range management
module 401 executes the processing of each Step 800, 801, 803 and
the volume management module 105 executes the processing of Step
803.
[0106] By executing the processing described above, this embodiment
stops once the replication processing when the fault occurs but can
change setting of the pairs registered at present depending on the
policy of the volumes, and can continue replication itself
irrelevantly to the restoration speed of the route fault.
[0107] The processing of each embodiment of the invention described
above can be constituted as a processing program. The processing
program can be stored and provided in the form of storage media
such as HD, DAT, FD, MO, DVD-ROM, CD-ROM, and so forth.
[0108] FIG. 9 is a block diagram showing a construction of storage
operation system according to a third embodiment of the invention.
The third embodiment employs the construction in which the
functions of the managing computer in the constructions of the
first and second embodiments are disposed inside the storage.
[0109] The third embodiment of the invention shown in FIG. 9
represents an example where a processing part for executing
management is disposed inside the storage 11a shown in FIG. 4, for
example. The volume management module 105 and the route management
module 401 existing in the storage device 104 of the managing
computer 100 shown in FIG. 4 are disposed inside a storage device
900 of the storage 11a. The volume management module 105 and the
route management module 401 are accomplished as software stored in
the storage device 900 of the storage 11a is read and executed by
the controller 112. A data synchronization module 901 is disposed
so that mutual information is coincident while keeping data
consistency with other storages. Data consistency is kept by use of
the network communication equipment 111, etc.
[0110] The third embodiment of the invention having the
construction described above can eliminate the managing computer
100 utilized in the first and second embodiments. The third
embodiment uses the data synchronization module 901 from a
plurality of managing computers 100 when the managing computers 100
are provided to a large-scale storage operation management system.
Therefore, when one managing computer cannot keep information, the
third embodiment makes it possible to share the data of the
plurality of managing computers.
[0111] According to each embodiment of the invention described
above, the volume pair management module 105 can generate a volume
replication pair between or inside the storages in accordance with
the policy of the volumes or properties, and can generate the
volume replication pair on the basis of the policy of the
volume.
[0112] In the volume replication operation, the route management
module 401 and the volume management module 105 can classify fault
shooting in accordance with the policy of the volume and can make
fault shooting of volume replication depending on the policy of the
volume.
[0113] As explained above, the invention can efficiently operate
and manage replication of the data areas inside the same storage or
between the storages in accordance with information of the data
area of the storage.
[0114] It should be further understood by those skilled in the art
that although the foregoing description has been made on
embodiments of the invention, the invention is not limited thereto
and various changes and modifications may be made without departing
from the spirit of the invention and the scope of the appended
claims.
* * * * *