CN100465871C - Memory device system - Google Patents

Abstract

This invention provides one memory device system with disc array comprises the following parts: identification disc device kind to manage programs, that are driven management unit; memory driver management unit to use the information driver form; the disc array comprises disc summed time management program, which comprises driver set life part to orderly switch ATA and to compute working summed time testing part.

Description

Memory device system

Technical field

The present invention relates generally to the control method and the memory device system of memory device system.

Background technology

It is disk array that the memory device system that is connected with computing machine has a kind of.Disk array also is known as RAID (Redundant Arrays of Inexpensive Disks promptly, redundant array of inexpensive disks), is to have a plurality of dish devices of array-like configuration and the memory device system of their control part of control.In disk array, read request (request of sense data) and write request (writing the request of data) are handled in the parallel work-flow of utilization dish device at high speed, and to the data additional redundancy.Disk array is as at " ACase for Redundant Arraysof Inexpensive Disks (RAID) ", DavidA.patterson, Garth Gibson and RandyH.Katz, Computer Science Division Department of Electrical Engineering andComputer Sciences, disclosed among the University of California Berkeley, kind and its structure according to additional redundant data are categorized as 5 grades.Have, so-called dish device is the memory storage with hard disk and CD etc. again.

On the other hand, in the dish device, have the different several kinds of interface that are used for other equipment connections.For example, be to have according to ANSI T11 to carry out the dish device (below be called " FC coils device ") of optical-fibre channel (Fibre Channel, below be called " the FC ") interface of normalizing operation and have the dish device (below be called " ata disk device ") that carries out ATA (At Attachment) interface of normalizing operation according to ANSI T13.

The ata disk device is less expensive, is used in home-use desk-top computer etc.On the other hand, FC dish device is used in the service server that requires reliability etc.Therefore, FC dish device is compared with the ata disk device, the handling property height of the input and output of data (below be called " I/O "), and, have can continuous working in 365 days in 24 hours high reliability.

Under the situation in the memory device system that these dish devices is used in disk array etc., distinguish use dish device sometimes according to desired performance and price etc. in the disk array.For example, disclosing in JP-A-10-301720 by will be than the also high dish device of dish device performance that is used to store data, is the dish device of parity checking as the storage redundancy data, seeks to improve the technology of dependability that is stored in the data in the disk array.

Summary of the invention

In recent years, in order to cut down the cost of data management, have the requirement of wanting to compile the data in the dish device that is stored in main frame. still, if only disk array is physically accumulated one, the situation of a plurality of disk arrays different with performance with using kind is compared, and interchangeability is impaired.Therefore, so-called by in a disk array, being mixed with high price high reliability high-performance disk device and the low dish device of cheap reliability and distinguish by purposes and to use, to suppress the price of disk array and to carry out the requirement that array more flexibly constructs more and more higher.

But former, the reliability of management dish device is difficult to uniformly, therefore, does not also consider reliability, for example device lifetime different dish device and construct and manage memory device system such as disk array of mixing pan device self.

Thereby, the present invention is a kind of storage system, it is characterized in that comprising: a plurality of dish devices that comprise the first dish device and the second dish device, and the control part that is connected with described a plurality of dish devices, the predetermined life-span of the wherein said first dish device is different with the predetermined life-span of the described second dish device, described control part is used to preserve the accumulation effective time of each described dish device, receive described a plurality of dish devices each dish device the described predetermined life-span and preserve the described predetermined life-span, utilize the described predetermined life-span of each described dish device and described accumulation to manage described storage system effective time.That is, be memory device system with control part of the different memory storage of the reliability of multiple arrangement self and these memory storages of control.At this, control part has the interface that connects these a plurality of memory storages, in order to the reliability of detection of stored device self and manage detected information.Specifically, by detecting the kind of each memory storage, determine the life-span (up time of the regulation that gives according to the kind of memory storage) of this memory storage. in addition, also consider by the life-span of the detected memory storage of management and the service time of this memory storage, management certain the time the example of remaining up time (residual life time) of this dish device.

In addition, the control part of memory device system of the present invention according to the residual life time of memory storage, changes the use-pattern of its memory storage.For example,, obtain usefulness as the backup of data, perhaps, stop and replacing the use of this memory storage with regard to using this memory storage if lack than certain official hour the excess time of the memory storage of current management.

Have, the control part of memory device system of the present invention according to the up time of memory storage, is predetermined the purposes of this memory storage again.For example, memory storage that the difference up time is long and short memory storage, will the up time long set of storage devices is as normally used memory storage, and will the up time short set of storage devices is as the memory storage of store backup data.

In addition, control part of the present invention based on the kind of detected memory storage, is controlled these memory storages.For example, according to the kind of detected memory storage, this memory storage is issued distinctive order.

Have, the control part of memory device system of the present invention also can have the kind that is used to discern memory storage and manages the program of different memory storages respectively and be used for the driven management table of management information again.

In addition, memory device system of the present invention in order to manage the cumulative time, for example, is set the life-span of driver, in a planned way starts/stop the ata disk device, the cumulative time of instrumentation work.

Have again, also can constitute and have storage managing server, in order to manage aforesaid memory device system mixed environment.

In addition, memory device system of the present invention carry out between memory device system or memory device system in data duplicate, based on life-span management, distribute the logical storage become the object that these data duplicate.

Have again, in " memory storage ", comprise the memory storage both sides of physics such as the memory storage of logic and disk array.

Description of drawings

Fig. 1 is the system construction drawing in the first embodiment of the present invention.

Fig. 2 is the figure of the driven management table among first embodiment.

Fig. 3 is the structural drawing of the input part among first embodiment.

Fig. 4 is the structural drawing of the efferent among first embodiment.

Fig. 5 is the figure of an example of the LU admin table among expression first embodiment.

Fig. 6 is the process flow diagram of the driver life-span management work among first embodiment.

Fig. 7 is the system construction drawing in the second embodiment of the present invention.

Fig. 8 is the figure of the port bypass circuit among second embodiment.

Fig. 9 is the figure of the FC-ATA conversion I/F among second embodiment.

Figure 10 is the system construction drawing in the third embodiment of the present invention.

Figure 11 is the figure of the LU of expression among the 3rd embodiment to an example of admin table.

Figure 12 is the figure of the storage managing server among the 3rd embodiment.

Figure 13 is the figure of an example of the device tabulation among expression the 3rd embodiment.

Figure 14 is the figure of an example of the LU of expression among the 3rd embodiment (logical unit promptly, logical block) assessment item table.

Figure 15 is the process flow diagram of the LU allocation process among the 3rd embodiment.

Figure 16 is the process flow diagram of the main frame LU allocation process among the 3rd embodiment.

Figure 17 is the process flow diagram that duplicates destination LU allocation process among the 3rd embodiment.

Figure 18 is the process flow diagram of the distribution candidate LU retrieval process among the 3rd embodiment.

Figure 19 is the figure of an example of the LU residual life threshold value table among expression the 3rd embodiment.

Figure 20 is the process flow diagram that the LU life-span management among the 3rd embodiment is handled.

Embodiment

Fig. 1 is the figure that expression has been suitable for first embodiment of memory device system of the present invention.Be connected with office terminal 500 as the disk array 700 of an example of memory device system and computing machine (below be called " main frame ") 100.The interface that is used for being connected with main frame 100 grades considers to have FC and SCSI.Disk array 700 has disk array control part 200, FC dish group 310 and ata disk group 410.

FC dish group 310 has a plurality of FC dish devices 301.Ata disk group 410 has a plurality of ata disk devices 401.Have again, in the present embodiment illustration FC dish device and ata disk device, but in accommodation of the present invention, the dish device that adopts in the disk array is not limited to these dish devices.

Office terminal 500 has input part 510 and efferent 520, described input part 510 is key board units of being used by the user when the input setting relevant with disk array 700 etc., and described efferent 520 is to the display device of the relevant information of user's demonstration and disk array 700 etc.

Disk array control part 200 has the CPU201 that carries out the program that is used for console panel array 700, storer 202, temporary transient storage is from the cache memory 203 of the inputoutput data of main frame 100, the main frame FCI/F (interface) 204 of the data transmit-receive between main control system 100 and the disk array control part 200, the driver FCI/F205 of the data transmit-receive between control FC dish group 310 and the disk array control part 200, the driver ATAI/F206 of the data transmit-receive between control ata disk group 410 and the disk array control part 200, managing I/the F207 of the information transmit-receive between control and management terminal 500 and the disk array control part 200.These parts connect with internal bus mutually.

Driver FCI/F205 is connected with FC dish group 310 by FC.Have again, in the present embodiment,, used fibre channel arbitrated loop (fibre channel arbitrationloop), but be not limited thereto, also can use other agreement, for example point-to-point, the silk weaving agreements such as (fabric) of FC as the agreement that is used for FC.

Driver ATAI/F206 is by ata bus, is connected with each ata disk device in being included in ata disk group 410.

The various programs that storage is carried out in CPU201 for console panel array 700 in the storer 202.Specifically, be the administration agent program 800 of the structure of the RAID control program 210 of work of console panel array 700 and management disk array 700.Also store various management information in the storer 202.Specifically, be the LU admin table 245 of the record information relevant with following part, described part comprises the logical storage (below be called " ATALU ") 420 of the driven management table 240 of the information that record and FC dish group 310 and ata disk group 410 are relevant, the logical storage of generation in FC dish group 310 (below be called " FCLU ") 320, generation in ata disk group 410.

RAID control program 210 further is made of several subroutines.Specifically, be the drive command release process of to FC dish group 310 or 410 issue an orders of ata disk group the time, carrying out 220, the driver management program of carrying out in order to manage FC dish group 310 and ata disk group 410 230 by CPU, the LU supervisory routine 225 carried out in order to manage the logical storage set in each dish group (below be called " LU ").

Administration agent program 800 further is made of several subroutines.Specifically, be Driver Information setting program 250, the Driver Information advising process of to office terminal 500 output driver information the time, carrying out 260 carried out when after receiving, setting the information relevant (below be called " Driver Information ") with the dish device from the input of office terminal 500, the array information setting program 270 of execution when after receiving, setting the information relevant (below be called " array information "), the array information advising process of to office terminal 500 output array information the time, carrying out 280 with disk array 700 from the input of office terminal 500.

Fig. 2 is the figure of an example of the content of expression driven management table 240.Driven management table 240 has each the dish device that has according to disk array 700 and registers the project of various information.Specifically, there is registration to give " driver No. " hurdle 240a of the number of dish device, " type of driver " hurdle 240b of the type of registration dish device, registration comprises " array structure " hurdle 240c of the RAID group (the dish device group of formation redundancy) of coil device and RAID grade thereof (below be called " array structure "), registration comprises " purposes " hurdle 240d of user's purposes that the RAID that coil device organizes (below be called " array purposes "), registration dish device is to start or " startup situation " hurdle 240e of the situation that stops, " cumulative time " hurdle 240f of the accumulative total working time of registration dish device, " lifetime settings " the hurdle 240g in the life-span of registration dish device, " residual life time " hurdle 240h of the residual life of registration dish device, be registered in and be judged as " shift automatically and specify " the hurdle 240i whether the dish device automatically transfers to data the information of other dish devices (below be called " shifting automatically ") when having reached the life-span.

The information of concrete registration expression FC, ATA etc. among " type of driver " hurdle 240b.The registration expression has generated the number and the information of representing the RAID grade of the sequence number of RAID group among " array structure " hurdle 240c.For example, the dish device of having registered " (1) RAID5 " is included in the dish device in the RAID group of RAID class 5 of first generation.Have again, in disk array 700, can generate a plurality of RAID groups, for example generate the RAID group of RAID1 and RAID5.Having, be included in dish device in the RAID group and can be the whole of dish device that disk array 700 had, also can be a part of.

Specifically registration form is shown the information of the DB of database purposes, the FS of expression file system purposes etc. in " purposes " hurdle 240d.In " startup situation " hurdle 240e, under the situation that the dish device has started, register ON, the information of registration expression OFF under the situation that the dish device has stopped.Storage representation is registered in the value of the difference of the time among " lifetime settings " hurdle 240g and " cumulative time " hurdle 240f in " residual life time " hurdle 240h.In " automatically shift specify " hurdle 240i, register 1 being set to shift automatically under the data conditions of storing in this dish device, under situation about not setting, register 0.

Under the situation under the situation of the kind of disk array control part 200 identification dish devices, in working time of instrumentation dish device, under the startup of console panel device and situation about stopping and when the dish device has reached the life-span, implementing under the situation about shifting automatically, in CPU201, carry out driver management program 230.

At disk array control part 200 according to input from office terminal 500, in driven management table 240 under the situation of the kind of assignment disk device, in driven management table 240, when the startup of assignment disk device and the situation that stops, carry out Driver Information setting program 250 under the situation in the life-span of assignment disk device and in driven management table 240 by CPU201.

Under the situation under the situation of the kind of 200 pairs of office terminals of disk array control part, 500 notice dish devices, in accumulative total working time of notice dish device,, carry out Driver Information advising process 260 by CPU201 under the situation of the startup of notice dish device and the situation that stops, under situation about because the dish device has reached the life-span data having been transferred to automatically under the situation of indication swapdisk device with in notice in other dish devices.

According to input, under the situation of setting array structure and under the situation of setting array purposes, carry out array information setting program 270 at disk array control part 200 by CPU201 from office terminal 500.

Under the situation of information relevant with the array purposes under the situation of 200 pairs of office terminals of disk array control part, the 500 notices information relevant, carry out array information advising process 280 by CPU201 with notice with array structure.

Fig. 3 is when being illustrated in the user of disk array 700 or supvr and carrying out the setting of disk array 700, the figure of an example of picture displayed in the efferent 520 of office terminal 500.Under this situation, the zone below in efferent 520, showing.That is, show input the dish device type type of driver setting regions 551, show input the dish device life-span driver lifetime settings zone 552, show driver by the condition of the startup of the dish device of user's appointment/stop start/stop setting regions 553, show the array structure of input array structure setting regions 571, show the array purposes setting regions 572 of the array purposes of input.This picture is seen on one side such as user, on one side by input part input data necessary, and confirms its content.

Fig. 4 has represented the user of disk array 700 or supvr when obtaining the information of disk array 700 by office terminal 500, an example of picture displayed in efferent 520.Under this situation, in efferent 520, show with lower area.

That is the type of driver viewing area 561 of the type of the dish device that had of indicating panel array 700,, the working time viewing area 562 of the accumulative total working time of the dish device that display panel array 700 is had, the Drive Status viewing area 563 of the situation of the startup of display panel device/stop, the exchange indicated number zone 564 of the exchange indication of display panel device, show because certain dish device has reached the life-span the automatic call announcement viewing area of data being transferred to automatically in other dish devices 565, the array purposes viewing area 582 of the array structure viewing area 581 of the array structure of display setting and the array purposes of display setting.By this picture is presented in the office terminal 500, supvr or user just can confirm the state of disk array 700.

Fig. 5 is the figure of the content of the LU admin table 245 that uses in order to manage FCLU320 and ATA LU420 of expression one routine disk array control part 200.In LU admin table 245,, there is the project of the following information of registration at each LU for disk array control part 200 each LU of management.

The identifier that registration distributes arbitrarily each LU in " LU No. " hurdle 245a.Whether registration expression has distributed to LU the information of main frame 100 (below, be called " using " under the situation of having distributed, be not called " using " also under the situation of not distributing) " having " or " nothing " in " host assignment situation " hurdle 245b.In " SCSI ID " hurdle 245c, LU to be distributed under the situation of main frame 100, the SCSI id number of the logical storage devices that comprises this LU is distributed in registration." LUN " among the hurdle 245d, LU is being distributed under the situation of main frame 100, the registration main frame 100 for this LU of access necessary SCSI logical block number.

The memory capacity of LU has been distributed in registration in " capacity " hurdle 245e.Registration is used to represent that LU is the information of FCLU320 or ATA LU420 type in " LU type " hurdle 245f.The residual life time of registration LU in " LU residual life time " hurdle 245g.Specifically, disk array control part 200 is obtained from driven management table 240 and is registered in the residual life time of respectively coiling device " driver No. tabulation " hurdle 245h, that constitute LU, the value that registration is wherein minimum.

The registration expression constitutes the information of the dish device of LU in " driver No. tabulation " hurdle 245h, as the guide look of driver No..Registration form is shown as the information of the evaluation of estimate of the benchmark of selecting into LU in " evaluation of estimate " hurdle 245i.Narrations later on such as computing method about evaluation of estimate.Have, under the situation of the LU that has used, registration can be registered in maximal value in this hurdle (in the example at this figure be+99999) in " evaluation of estimate " hurdle 245i again.

Below, the situation of constructing of carrying out driven management table 240, LU admin table 245 and RAID group according to user or gerentocratic indication about disk array control part 200 describes.

When disk array 700 had been connected power supply, disk array control part 200 was carried out drivers management programs 230, implemented detecting the dish device that is connected with driver ATA I/F206 with driver FC I/F205.As an example of detecting of dish device, the example that uses the ModeSelect order is arranged.Specifically, at first, 200 pairs of disk array control parts respectively coil device issue ModeSelect order.Received the dish device of ModeSelect order, the page information that is stored in this dish device has been sent to disk array control part 200.In page information, comprising the information relevant with the type of this dish device.According to the page information of obtaining from the dish device, the kind of the dish device that 200 identifications of disk array control part detect.

In addition, disk array control part 200 is carried out Driver Information setting programs 250, about detected whole dish devices, with its information registering in " the driver No. " and " type of driver " project of driven management table 240.

The kind that input part that also can access management terminal 500 comes the input disc device is arranged again.Under this situation, disk array control part 200 is based on 500 information of collecting from the office terminal, with information registering in driven management table 240.

After having detected or registered the type of dish device, disk array control part 200 carries out the setting of array structure and array purposes based on user's indication.

At first, one receives the user obtains indication with the information to " driver No. " " type of driver " of input part 510 input of office terminal 500, disk array control part 200 is just carried out Driver Information advising process 260, obtain requested information from driven management table 240, by managing I/F207, send office terminal 500 to.Office terminal 500 is presented at the information that receives in efferent 520 on the part of the demonstration " driver No. " of the type of driver viewing area 561 that shows and " type of driver ".Have again, under the situation of the type of user definition dish device, also can omit this processing.

Afterwards, the user is based on the information that shows in the type of driver viewing area 561 that shows in the efferent 520 of office terminal 500, the selected device that coils arbitrarily uses array structure setting regions 571, uses selected dish device to construct the indication of RAID group from input part 510 inputs.In addition, the user imports the purposes of the RAID group of constructing.Disk array control part 200 is carried out the array information setting programs, will organize by the RAID that managing I/F207 receives and the array information of purposes etc., is registered in " array structure " " purposes " hurdle of driven management table 240.

The user is based on the type of the dish device of demonstration in the type of driver display frame 561, construct RAID group, but can be not be set at the array purposes that requires high reliability and high performance array purposes, for example data base-oriented with reliability and the low ata disk device of performance this moment.Specifically,, specified by the user under the situation of " purposes ", judged whether the life-span of dish device is fit to its purposes in information based on life-span of dish device described later.Then, according to this judged result, the user is warned.

For example, suppose the ata disk device is set under the situation of high reliability and high performance array purposes, office terminal 500 can be issued by 520 couples of users of efferent and be noted and warning.In addition, in by type of driver display frame 561 under the situation of display panel device, can be with different diagrams, with can the different dish device of identifying species.

In addition, when initial generation RAID organizes, the life-span of not considering to coil device, and when regenerating the RAID group once more, also consider to add the method for the residual life of dish device as the judgment standard of purposes setting.

Have again, one receives that the user obtains the information of the information of " array structure " " purposes " by input part 510 input of office terminal 500 in order to indication, disk array control part 200 is just by carrying out the array information advising process, from driven management table 240 retrieval from the office terminal 500 by managing I/F207 information requested, and transmit the information that retrieves.Office terminal 500 after the collection information, the information that demonstration obtains in the array structure viewing area 581 of efferent 520 and array purposes viewing area 582.

Below, the life-span management of the dish device that carries out about disk array control part 200 describes.As previously mentioned, disk array control part 200 is by carrying out driver management program 230, the automatic transfer of the management of the work cumulative time of console panel device, the startup of plan/stop and data.

At this, " life-span " of described dish device is meant, the manufacturing plant of dish device serves as the time that the basis is set with mean time between failures that coils device and standards such as recommending the duration of runs, or the disk array supvr serves as the time that the basis is set with the time of coiling the setting of device manufacturing plant.At this, be defined as " being used to prevent the time that operates as benchmark fault, that become preventive maintenance that reason causes for a long time " with the dish device.Thereby,, need after the transfer of having implemented the storage data, wait the exchange of implementing to new dish device by the Service Technician about having surpassed the dish device in " life-span ".

In order to carry out life-span management, at first, disk array 700 carries out the lifetime settings of each dish device and shifts automatically and specify.This processing also can be carried out when carrying out the setting of various tables according to user's indication simultaneously by disk array control part 200.Perhaps, also can carry out before.Followingly carry out lifetime settings of each dish device, automatically shift and specify.

One receive user's 500 input part, 510 inputs from the office terminal, indication obtains the indication of the information of " driver No. ", " type of driver ", disk array control part 200 just transmits by managing I/F207 to office terminal 500 information requested.Received the office terminal 500 after the information, in the array type viewing area 561 of efferent 520, shown " driver No. " and the information of " type of driver " about receiving.

The user is based on the information that shows in the type of driver viewing area 561, the kind of identification dish device, input part 510 by office terminal 500, the life-span of assignment disk device with whether need automatic transfer. by managing I/F207 received with life-span of dish device and dish device whether need to shift automatically disk array control part 200 after the relevant information, carry out Driver Information setting program 252, the information that registration receives in " lifetime settings " and " shift automatically and specify " hurdle of driven management table 240.Have again, also can not allow the user specify the information of " lifetime settings " and " shift automatically and specify ", and determine singlely, be set in the driven management table 240 by the kind of disk array control part 200 according to the dish device.For example, just long if consider FC dish device from being predicted as the life-span at first, be set in the driven management table 240.

Fig. 6 is the process flow diagram of the order of the driver life-span management that carries out of indicating panel array control part 200.

At first, disk array control part 200 is by carrying out driver management program 230, the working time of respectively coiling device that beginning instrumentation disk array 700 is had, with cumulative calculation as a result in " cumulative time " of driven management table 240.Have again, under the situation of instrumentation working time,, carry out the instrumentation (step 1003) of working time based on the clock that disk array 700 has.Afterwards, disk array control part 200 is the value of " cumulative time " and " lifetime settings " of driven management table 240 relatively, judges whether the cumulative time has reached the life-span of setting.Have, timing relatively can be to handle nonsynchronous arbitrary timing (step 1004) with other again.

Do not reach in the cumulative time under the situation in life-span, turn back to step 1003.Reached in the cumulative time under the situation in the life-span of setting, judged in " shift automatically and specify " hurdle 240i of dish device whether registering 1 (step 1005).Under the situation of the appointment of not shifting automatically (value of registration is 0), disk array control part 200 is by carrying out Driver Information advising process 260, the dish device that will reach the life-span of setting the cumulative time is notified to office terminal 500 as the dish device that should exchange by managing I/F207.

Office terminal 500 shows that in the exchange indicated number zone 564 of efferent 520 cumulative time reached the dish device in life-span, as the dish device (step 1006) that should exchange.Have at the dish device that has reached the life-span about the cumulative time under the situation of appointment of automatic transfer (value of registration is 1), disk array control part 200 is by carrying out driver management program 230 and RAID control program 210, reach data in the dish device in life-span with being stored in the cumulative time, sent the dish device that becomes automatic transfer destination ground to.Have again, wait by the user and specified the dish device that becomes automatic transfer destination ground in advance.Under this situation, also can specify the automatic transfer destination domain device (step 1010) of a dish device as a plurality of dish devices.

After the data transmission was over, disk array control part 200 was notified and becomes the relevant information (step 1011) of dish device on automatic transfer destination ground by managing I/F270 to office terminal 500.One announcement information, office terminal 500 just show in the automatic call announcement viewing area 565 of efferent 520 and the relevant information of dish device that becomes automatic transfer destination ground.Driver management program 230 is removed " cumulative time " (step 1012) of the driven management table 240 relevant with the driver after automatic the transfer, turns back to step 1003.

Have, in the present embodiment, the dish device that has reached the life-span about the cumulative time shifts automatically again, but also can shift automatically etc. about the dish device that the residual life time is lower than certain threshold value.Under this situation, disk array control part 200 is confirmed registered value in " residual life time " 240h, and special appointment is lower than the dish device of threshold value.

On the other hand, for the increase of accumulative total working time of suppressing to coil device, the startup of the work of disk array control part 200 console panel devices/stop.The life-span management of representing among this control and Fig. 6 carries out independently.

Disk array control part 200 is based on predefined opportunity, coils the startup of device/stop.Make the dish device opportunity that stops/starting can be made as the time that stops fully at night for example etc. from user's access.Perhaps, the short situation of certain threshold value of residual life time ratio that also certain can be coiled device is made as the opportunity of the startup of dish device/stop.

In addition, also can the user startup/stop by the input part 510 direct given disc devices of office terminal 500.Office terminal 500 is by managing I/F207, with the startup of dish device/stop relevant indication to give disk array control part 200.Disk array control part 200 is based on the indication that receives, the startup of console panel device/stop.Afterwards, the disk array control part 200 startup/stop conditions that will coil device is registered in " the startup situation " of driven management table 240.Have again, by to driver issue Start/StopUnit order, carry out the startup of driver/stop.

Have again, the user utilizes the information of input part 510 indications " startup situation " " cumulative time " of office terminal 500 to obtain under startup situation that obtains driver and the situation of accumulative total working time. and disk array control part 200 transmits office terminal 500 specified message by managing I/F207.Office terminal 500 shows the startup situation of the dish device that receives and the accumulative total working time of dish device in the Drive Status viewing area 563 of efferent 520 and working time viewing area 562.

The user also can be based on the accumulative total working time or the residual life time of the driver that shows, the exchange of enforcement dish device in advance.

Below, the situation that has the dish device issue an order of distinct interface about 200 pairs of disk array control parts describes.

For example, ata disk device 401 can not be handled the order with a plurality of labels.Therefore, disk array control part 200 is carried out drive command release process 220 when having received indication from main frame 100 and write the order of data etc., and decision should be to dish device issued command.Specifically, disk array control part 200 is to the issue an order of dish device the time, " type of driver " with reference to driven management table 230, the issue destination of recognition command is FC dish device 301 or ata disk device 401, under the situation that is the ata disk device, do not issue order, and issue single order with a plurality of labels.

According to present embodiment, can in same disk array 700, manage the reliability different dish device of life-span for example that is mixed with dish device self.In addition, can consider to constitute the reliability of the dish device of RAID group, decide the purposes of LU.In addition, just can in same disk array, manage the different dish device of control mode.

Fig. 7 is the figure that expression has been suitable for second embodiment of computer system of the present invention.Below, only just narrate with the difference of first embodiment.In Fig. 7, disk array 700 ' has the FC actuator housing 300 that is used to take in FC dish group 310 and is used to take in the ATA actuator housing 400 of ata disk group 410.Be accommodated in the FC dish device 301 in the FC actuator housing 300, the port bypass circuit 330 by FC actuator housing 300 is had is connected with ATA actuator housing 400 with driver FC I/F205.In addition, be connected with FC loop 600 between FC actuator housing 300 and the ATA actuator housing 400, specifically connect with optical cable.

The ata disk device 401 that is accommodated in the ATA actuator housing 400 can not directly be connected with FC loop 600.Therefore, ATA actuator housing 400 has FC-ATA conversion I/F430, is used for conversion FC interface and ata interface.Thereby ata disk device 401 is connected with FC actuator housing 300 just by FC-ATA conversion I/F430 and FC loop 600.

Fig. 8 is the figure of the structure of expression port bypass circuit 330.The port bypass circuit 330 of this figure has 5 ports.In addition, port bypass circuit 330 has the selector switch identical with port number 331, like this, presses port and implements the loop tie lines.Utilize selector switch 331,, just constitute signal path,, just form not the signal path that has passed through internal circuit via port if port does not connect via port if be connected the state of tie lines with port.

Fig. 9 is the figure of the structure of expression FC-ATA conversion I/F430.FC-ATA conversion I/F430 is used for being connected with FCI/F the interface of ata disk device.FC-ATA conversion I/F430 have with the interface FC I/F431 of FC, with ATA interface ATA I/F438, storer 441, carry out the program that is positioned at storer 441 CPU440, be connected the internal bus 442 of FC I/F430, ATA I/F438, storer 441 and CPU440.

In storer 441, taking in the commands buffer 439 of carrying out adhoc buffer, the FC data convey program 432 that described impact damper is used to keep and regenerates and carry out when receiving the order of FC, the FC data convey program of with the FC transceive data time, carrying out 433, the order conversion program of when the order with FC is transformed to the order that gives ATA, carrying out 434, carry out the data buffer 435 of the buffering of the data between FC and the ATA, the ata command release process of carrying out during order after ATA side issue conversion 436, ATA data convey program of carrying out during with the ATA transceive data 437 and order.

CPU440 carries out FC order reception program 432, receives the order from FC, carries out FC data convey program 433, and the data of carrying out 435 of FC I/F431 and data buffers transmit.In addition, CPU440 fill order conversion program 434 is transformed to the order that gives ATA with the order of FC, carries out ata command release process 436, the order after ATA side issue conversion.In addition, CPU220 carries out ATA data convey program 437, and the data of carrying out 435 of ATA I/F438 and data buffers transmit.

In the present embodiment,, just can pass through FC, disk array control part 200 ' is connected with ata disk device 401 by using FC-ATA conversion I/F430.FC-ATA conversion I/F430 will be transformed to the order that the ata disk device is used from the order of disk array control part 200 ' reception.In addition, also utilize FC-ATA conversion I/F430, the data that conversion is received and dispatched between disk array control part 200 ' and ata disk device.

According to present embodiment, can be that fibre channel arbitrated loop connects the ata disk device with the interface that is connected FC dish device.

Figure 10 is the figure that expression has been suitable for the 3rd embodiment of computer system of the present invention.Below narrate difference with first embodiment.The computer system of present embodiment also has other disk arrays 701B that is connected with disk array 701A, the principal computer 100 that is connected with disk array 701B, is storage managing server 10000 with network 20000 with the computing machine that disk array 701A is connected with 701B by management except that the structure of first embodiment.Two disk arrays are connected 30000 by the main frame FC I/F204 that each disk array had with remote copy and interconnect.Have again, consider for example optical-fibre channel and ESCON etc. in the method for remote copy connection 30000, but not special in the present embodiment difference.In addition, the managing I/F207 that is had with each disk array 701 of use in being connected between the network 20000 in each disk array 701 and management.

In addition, the disk array 701 of present embodiment has increased newly in the storer 202 in disk array control part 2001 and has had LU to admin table 246 except that the structure of first embodiment.Have again, in the RAID control program, newly added the program module that the data that the data that are used for controlling between the disk array 701B that is connected with disk array 701A duplicate the program module of (below be called " remote copy ") and be used to manage disk array 701A or B are duplicated (below be called " duplicating (レ プ リ ケ-シ ョ Application in the device) in the device ").Have again, about the implementation method of remote copy, have synchronously (data duplicate finish before, do not write finishing of request to the principal computer report data) and asynchronously (do not wait finishing that pending data duplicates, and write being over of request to the principal computer report data), but the not special difference of present embodiment.

Figure 11 is the figure of an example of expression LU admin table 246. Disk array 701A and 701B use registered LU admin table 246 in each storer 202, the LU that management is used in the identical data between holding tray array in the remote copy is right, or the LU of the identical data that is positioned at disk array 701 that generates in duplicating in the holding device to (below be called " LU to ").

LU has the project of " LU is to No. ", " copy source LU information ", " duplicating destination LU information " and " to state " to admin table 246.

Registration is to the identifier of LU to any distribution in " LU is to No. " hurdle 246a.In the present embodiment, do not distinguish LU in 701 of disk arrays and the disk array 701 to distribution marker.

" copy source LU information " project has 3 the next projects of following explanation.Registration has given the LU No. LU centering of certain identifier, that distribute to the LU that stored raw data (below be called " copy source LU ") in " LU No. " hurdle 246b.To be used to discern LU right to the LU that is based on remote copy in registration in " to attribute " hurdle 246c, also be based on the right information of LU of duplicating in the device, specifically, the value of registration expression " long-range " under the situation of remote copy, the value of registration expression " this locality " under the situation of in device, duplicating.The type of the dish device of copy source LU is taken in registration in " LU type " hurdle 246d.

Project has 3 the next projects of following explanation " to duplicate destination LU information ".The identifier of the disk array 701 under the LU paired with copy source LU (below be called " duplicating destination LU ") is distributed in registration in " device ID " hurdle 246e.For example, under the situation of remote copy, the identifier of the disk array 701 of remote copy destination is distributed in registration, and under the situation of duplicating in device, the identifier of the disk array 701 with copy source LU is distributed in registration.The LU No. that duplicates destination LU has been distributed in registration in " LU No. " hurdle 246f.The type that the dish device of destination LU is duplicated in the registration maintenance in " LU type " hurdle 246g.

The right current state of registration LU in " to state " hurdle 246h.For example, the registration expression obtains the data synchronization that is stored in each right line centering of LU, the state of the content unanimity of Cun Chu data (below be called " Pair state ") separately, perhaps, in the value of LU to the state of not obtaining data sync (below be called " Split state ").

The function of duplicating in disk array 701 operative installationss is carried out following work.For example, in the time arbitrarily, the LU that will be in the Pair state is to changing to the Split state.In this wise, just in duplicating destination LU, preserve the data (such processing be called " obtain take out a little ") of LU under the random time to being had.Afterwards, by main frame 100 from duplicating destination LU sense data, be written to other memory storages for example in magnetic tape equipment etc., just can access and obtain the backup that is stored in the data of LU centering in the time of taking out a little. in addition, also the destination LU that duplicates that has obtained after taking out a little himself can be preserved, as the backup of data.

Have again, the LU that also can make up remote copy to device in the LU that duplicates right.For example, carrying out between disk array 701A and the B under the situation of remote copy, duplicating destination LU, be made as the copy source LU that duplicates in the performed device of disk array 701B what disk array 701B had.Then, also can become in the inside of disk array 701B the LU that duplicates in the device right duplicate destination LU.At this, generation duplicate destination LU as mentioned above, by being made as the split state, can be used in the backup of data.

The user can independent assortment FCLU320 and ATALU420, sets copy source LU and duplicates destination (below be called " LU to ").Particularly, the user can pass through storage managing server 10000 described later, sets the right distribution of LU of having considered each LU feature.For example, the LU type as feature, can be distributed the optical-fibre channel LU320 of the FC dish device that has used the high-performance high reliability to copy source LU, distribute the ATALU420 that has used ata disk device at a low price duplicating destination LU.In addition, can with the LU residual life time as feature, will have few dish device LU of residual life time not as object for allocation, perhaps be set at only be used in duplicate in the device duplicate destination etc.

Below, describe about storage managing server 10000.Figure 12 is the figure of the structure of expression storage managing server 10000.

Storage managing server 10000 has and is used for the managing I/F14000, efferent 15000 and the input part 16000 that are connected with network 20000 with CPU11000, primary memory 12000, storage part 13000, management.These parts interconnect by the communication path 17000 of bus etc.Storing the program of carrying out in the storage part 13000 and be storage management program 13100, storage in CPU11000 promptly stores archival memory 13200 from the zone of the information that disk array 701 is collected, keeps storage management program 13100 as device tabulation 13300, the LU assessment item table 13400 of the guide look of the disk array 700 of management object be used to carry out the LU residual life threshold value table 13500 of the residual life management of LU.

Have again, in storage archival memory 13200 termly the driven management table 240, LU admin table 245 and the LU that are had of catch tray array 701A and B to admin table 246 in canned data, generate duplicating of these tables.In addition, by based on user's indication, carry out storage management program 13100 by storage managing server 10000, generating apparatus tabulation 13300, LU assessment item table 13400 and LU residual life threshold value table 13500.

Figure 13 is the figure of an example of indication device tabulation 13300.For the disk array of confirming by storage managing server 10000 self should manage 701, reference and the needed information of communicating by letter of disk array 701, operative installations tabulation 13300.The identifier that the disk array 701 that registration is managed storage managing server 10000 in " object apparatus No. " hurdle 13300a distributes arbitrarily.The intrinsic identifier of disk array 701 is distributed in registration in " device ID " hurdle 13300b.Registration storage managing server 10000 and the information that disk array 701 signal posts need are for example used under the situation of using IP network in the network 20000 in management in " IP address " hurdle 13300c, register and the relevant information in IP address of distributing to disk array 700.

Figure 14 is an example of expression LU assessment item table 13400.The project that should consider when in LU assessment item table 13400, registering the right LU of selected LU and the evaluation content in this project.When storage managing server 10000 meets the LU of user's indication in retrieval, with each LU of project evaluation that is registered in the LU assessment item table 13400, with the evaluation content of appointment in this table, be registered in the project of evaluation of estimate 245i of the duplicating of LU admin table 245 (below be called " LU admin table 245 " ').Afterwards, based on this evaluation result, storage managing server 10000 is selected suitable LU, is prompted to the user.

Assessment item when the LU that the registered user can select in " assessment item " hurdle 13400a specifies.In " evaluation object LU admin table hurdle " hurdle 13400b, be registered in the project that becomes in the assessment item in the LU of the evaluation object admin table 245, for example LU type 245f.When in " initial conditions is read again " hurdle 13400c, being registered in the evaluation of carrying out each LU of registration in the LU admin table 245 ', whether need to read again the condition of user's input.When the content of registering in the project of the LU admin table 245 ' of the object that the condition that is stored in user's input in when setting up " condition compensation (ペ Na Le テ イ) " hurdle 13400d and quilt are estimated is consistent, with the value of the value addition of " evaluation of estimate " hurdle 245i of LU admin table 245 '.When the content of registering in the condition that is registered in user's input in " when condition is false compensation " hurdle 13400e and the hurdle of the LU admin table 245 ' of the object of being estimated is inconsistent, with the value of the value addition of " evaluation of estimate " hurdle 245i of LU admin table 245 '.Have again, narrate the evaluation method of concrete LU later on.

Figure 19 is the figure of an example of expression LU residual life threshold value table 13500.The residual life of the LU that LU residual life threshold value table 13500 is had at each disk array 701 of storage managing server 10000 management the time use.Specifically, the project that has the following information of registration.

The identifier that registration distributes arbitrarily respectively the threshold value after a plurality of settings in " threshold value No. " hurdle 13500a.The identifier of the disk array 701 of the object that becomes the residual life management is distributed in registration in " device ID " hurdle 13500b.Registration form is shown as the information of LU among the LU that the disk array 701 for management object had, the actual management residual life in " monitored object LU tabulation " hurdle 13500c.The information of registration expression and the corresponding threshold value of identifier in " residual life threshold value " hurdle 13500d.

In the present embodiment, about the LU that registers as management object, storage managing server 10000 is confirmed the residual life times, if the residual life time of confirming is lower than preset threshold, just exports warning from storage managing server 10000 to the user.According to circumstances, promote LU to shift to the user and set, and shift indication automatically to the issue of dish device.Have again, under the low situation of the importance of the data in being stored in LU, not with the residual life data as judgment standard, but also can will constitute the life time of dish device of LU as judgment standard.But, the urgent exchange that needs to coil device under this situation.

In the present embodiment, the user when using storage managing server 10000 to generate the LU of the disk array 701 that is mixed with the different dish device of input and output I/F, parameters such as Performance And Reliability that can consideration dish device.Below, this processing is called the LU allocation process.Have, do not limit breakpoint, the processing that illustrates in following flow process is that storage managing server 10000 is by carrying out the processing that storage management program 13100 is implemented.

Figure 15 is the process flow diagram of the process of expression LU allocation process 2000.At first, storage managing server 10000 is according to the indication from the user, and judgement is that the LU that will become copy source distributes to main frame 100 and handles, and still carries out the processing (step 2100) of duplicating destination LU that distributes in duplicating in remote copy and the device.

Be judged as in step 2100 under the situation of distribution of copy source LU, storage managing server 10000 carries out main frame provides LU allocation process.Be described in detail this processing (step 2200) later on.Afterwards, whether storage managing server 10000 duplicates the distribution of destination LU to user's query.Specifically, the demonstration of whether handling to efferent 15000 output inquiries (step 2300).

Indicated by the user in step 2100 and duplicated under the situation that destination LU distributes and indicated by the user in step 2300 and duplicated under the situation that destination LU distributes, storage managing server 10000 duplicates destination LU allocation process.Narrate (step 2400) in detail about this processing later on.Afterwards, storage managing server 10000 finishes the LU allocation process.Be designated as in step 2300 under the situation that need not duplicate destination LU allocation process, storage managing server 10000 finishes the LU allocation process immediately.

Figure 16 is expression is provided the detailed process of LU allocation process by the main frame of step 2200 execution of Figure 15 a process flow diagram.Storage managing server 10000 receives the necessary condition of copy source LU (below be called " parameter ") from the user.At this, described parameter be except that the object apparatus, the memory capacity of LU, main frame FC I/F, the LU such as SCSI ID, LUN that generate LU generate the necessary essential information, about the information (step 2210) of the assessment item in " project " hurdle that is registered in LU assessment item table 13400.

Afterwards, the parameter of storage managing server 10000 to give by the user, with the candidate of suitable LU as copy source LU, the distribution candidate LU retrieval process of selecting.Be described in detail this processing (step 2220) later on.Afterwards, represent the candidate of the LU that selects to the user, ask user's affirmation (step 2230).

In step 2230, admit that the user LU candidate of pointing out is under the situation of copy source LU, storage managing server 10000 update stored in the LU admin table 245 in the disk array 701 with admitted LU and be its duplicate 245 '.Specifically, at first, storage managing server 10000 is about being stored in the LU admin table 245 ' in the storage archival memory 13200, " host assignment situation " hurdle 245b of admitted LU is changed to " having ", the value of the parameter that registration gives from the user in " SCSI ID " hurdle 245c and " LUN " hurdle 245d.Afterwards, storage managing server 10000 is indicated to the disk array 701 with admitted LU, and feasible update content with LU admin table 245 ' is reflected in the LU admin table 245 (step 2240).

Afterwards, 200 indications of 10000 pairs of disk array control parts of storage managing server will reflect that the LU admin table 245 that upgrades is basis execution LU management, and End Host provides LU allocation process (step 2250).

In step 2230, to have rejected the user under the situation of the LU that points out, storage managing server 10000 is ended main frames LU allocation process (step 2260) is provided.Have again, for the robotization of seeking to handle, the LU that also can be not select to user prompt, and carry out the processing set as copy source LU.

Figure 17 is the process flow diagram that is illustrated in the detailed process of representing in the step 2240 of Figure 15 of duplicating destination LU allocation process.

Storage managing server 10000 receives and is used for selecting the parameter of duplicating destination LU from the user. and at this, described parameter is the information (step 2410) that illustrates in above-mentioned copy source LU.

Afterwards, the parameter of storage managing server 10000 to be given by the user advances to select to become the distribution of the candidate of duplicating destination LU candidate LU retrieval process.Be described in detail this processing (step 2420) later on.Afterwards, storage managing server 10000 is shown in the candidate of selecting in the step 2420 of duplicating destination LU to subscriber's meter, and request confirms to use the LU (step 2430) that selects.

In step 2430, admit that the user LU that points out duplicates under the situation of destination LU, storage managing server 10000 upgrade LU to admin table 246 and LU admin table 245 and separately duplicate 246 ', 245 '.In addition, storage managing server 10000 is indicated to the disk array 701 of the selecteed LU of storage, makes that the content after will upgrading is reflected in each table.Specifically, to admin table 246 ', generate new LU to item, the information of duplicating destination LU of registered user's designated parameters and selection about LU about the destination LU that duplicates that selects.In addition, about LU admin table 245 ', " host assignment situation " hurdle 245b of the LU that selects is changed to " having ", the value (step 2440) of registration parameter in " SCSI ID " hurdle 245c and " LUN " hurdle 245d.

Afterwards, 10000 pairs of storage managing servers have duplicating destination LU and becoming the disk array control part 200 of the disk array 701 of the right copy source of line of selection, indication is carried out remote copy based on the line after upgrading to LU admin table 246 and is handled or install interior replication processes, finishes to duplicate destination LU allocation process (step 2450).

In step 2430, under the situation of duplicating destination LU of having rejected prompting, storage managing server 10000 ends to duplicate destination LU allocation process, finishes to duplicate destination LU allocation process (step 2460).Have, for the robotization of seeking to handle, also the LU that can not select to user prompt duplicates the processing that destination LU sets and carry out conduct again.

Figure 18 is the process flow diagram of the detailed process of the distribution candidate LU retrieval process of expression in the step 2420 of the expression step 2220 of Figure 16 and Figure 17.

At first, LU admin table 245 ' the middle extraction of storage managing server 10000 storage from storage archival memory 13200 is not used in the LU of main frame 100.Specifically, the LU that is registering " having " in " host assignment situation " hurdle 245b of LU admin table 245 ' is judged as and uses, and registers maximal value (in Fig. 5 be+99999) in evaluation of estimate.The LU that is registering " nothing " is not judged as and uses, and evaluation of estimate is changed to 0 (step 3020).

Then, storage managing server 10000 gives evaluation of estimate to projects of registration in each LU assessment item table 13400 to the whole not use LU that extracts out in step 3020.Narrate concrete example (step 3030) later on.

Afterwards, storage managing server 10000 output is registered in the LU of value minimum among whole LU in the LU admin table 245 ', " evaluation of estimate " hurdle 245i, as the copy source LU or the candidate (step 3040) of duplicating destination LU.

Have again, when the renewal of 10000 pairs of each disk arrays of storage managing server, 701 indication LU admin tables 245 (step 2240,2440), upgrade the evaluation of estimate that is registered in the LU admin table 245 ' simultaneously.

Below, the concrete example of 4 steps 3030 is described.

As first example, consider the evaluation the when user is appointed as " FC " with " type ".Storage managing server 10000 is by confirming " evaluation object LU admin table " hurdle 13400b, and the hurdle of the LU admin table 245 ' that affirmation should compare with user's designated value " FC " is " LU type " hurdle.Afterwards, storage managing server 10000 is selected to extract out does not use among the LU one, confirms the content with " LU type " hurdle of the corresponding LU admin table 245 ' of this LU.Information in being registered in this " LU type " hurdle is under the situation of " FC ", because term harmonization with user's appointment, so storage managing server 10000 according to be registered in " type " corresponding items when setting up " condition compensation " hurdle 13400d in value, in this uses the hurdle of evaluation of estimate 245i of LU admin table 245 ' of LU, do not add 0.Information in being registered in this " LU type " hurdle is under the situation of " ATA ", because it is inconsistent with the condition of user's input, so storage managing server 10000 according to be registered in " type " corresponding items " when condition is false compensation " hurdle 13400e in value, in this uses the hurdle of evaluation of estimate 245i of LU admin table 245 ' of LU, do not add+100.

Storage managing server 10000 is judged as above processing, promptly judge each project whether with the term harmonization of user's appointment, according to its result, with the regulation value as evaluation of estimate, carry out the processing with the evaluation of estimate hurdle 245i addition of each LU admin table 245 ' repeatedly, up to the suitable evaluation that is documented in the whole projects in the LU assessment item table 13400 of the whole LU of use that extract out.The result of this processing does not all use the value of " evaluation of estimate " hurdle 245i of LU admin table 245 ' of LU certain.In this concrete example, the type selecting that will coil device by the user is " FC ", and storage managing server 10000 just can provide the LU that is made of FC dish device as candidate.

As second example, consider the evaluation the when user generates the LU that is used to back up utilization.Keep in the utilization state of backup of data in a plurality of epoch, consider in the LU of backup usefulness, to use LU that the dish device by cheapness constitutes and by since the residual life time shortly be difficult to be used in the LU that the dish device of ordinary business practice in using constitutes.For such situation, in this example, definition is as the assessment item of so-called " LU utilization " that Figure 14 put down in writing.This assessment item is appointed as " LU utilization " under the situation of " backup " the user, be storage managing server 10000 " LU type " preferentially is provided is " ATA ", and " LU residual life time " is the project of the LU of " being less than or equal to 1000 hours " as the such evaluation LU of candidate.Have, evaluation procedure is identical with first example again.Utilize this assessment item, storage managing server 10000 can be further be suitable for backing up the candidate of the LU of use to user prompt.

As the 3rd example, consider to be used to generate the evaluation of the corresponding LU of reliability that is asked with the user.Under such situation, in the assessment item of so-called " reliability " put down in writing as Figure 14, be defined as evaluation " LU residual life time ".This assessment item is appointed as " reliability " under the situation of " height " the user, is that " LU residual life time " preferentially is provided is the project of the LU of " more than or equal to 30000 hours " as such each LU of evaluation of candidate to storage managing server 10000.Utilize this assessment item, storage managing server 10000 just can use predefined metewand to estimate the residual life time, and suitable LU candidate is provided.

As the 4th example, consider to generate the corresponding LU of the quantity of duplicating the destination that sets with the user.For a plurality of epoch remain on the data of narrating in second example, there is the function of duplicating in remote copy of utilizing and the device to generate many situations of duplicating the destination.Under such situation, " LU type " estimated in definition in as the assessment item of Figure 14 so-called " duplicating destination quantity ".Specify under many situations of duplicating the destination the user, this assessment item is the cheap ATA LU of storage managing server 10000 preferential retrievals are prompted to such each LU of evaluation of user as candidate a project.Utilize this assessment item, storage managing server 10000 is based on the quantity of duplicating the destination of user's appointment, uses predefined metewand to estimate the LU type, and the suitable destination LU candidate of duplicating is provided.

In the present embodiment, storage managing server 10000 is realized the life-span management by the LU grade in the disk array 701 that is mixed with the different dish device of input and output I/F.Below this processing being called the LU life-span management handles.Have, the particular determination breakpoint is not respectively handled storage management program 13100 by the CPU11000 execution of storage managing server 10000, carries out each and handles again.

Figure 20 is the process flow diagram of the process of expression LU residual life management processing.

Storage managing server 10000 uses information in storage archival memory 13200 regular update, disk array 701A and B LU admin table 245 ' separately, monitors the LU residual life time termly.Specifically, storage managing server 10000 is about each specific LU among " device ID " hurdle 13500b of LU residual life threshold value table 13500 and " object LU tabulation " hurdle 13500c, will compare with preset threshold (value of registering among corresponding with each LU " residual life threshold value " hurdle 13500d) among each LU with the value registered among " LU residual life time " hurdle 245g of the corresponding LU admin table 245 ' of each LU.Have, utilize the setting situation of " monitored object LU tabulation " hurdle 13500c, can be suitable under the situation of a plurality of threshold values a LU, the value that is suitable for " residual life threshold value " hurdle 13500d is maximum threshold value (step 4010).

Result relatively was lower than under the situation that is registered in the value among " residual life threshold value " hurdle 13500d in the residual life time of each LU, and storage managing server 10000 is judged as the required LU (step 4020) of warning output with this LU.

Also under the situation that does not have the LU that becomes the warning object in step 4020, storage managing server 10000 turns back to the processing of step 4010, continues the management of LU residual life.

Exist in step 4020 under the situation of the LU that becomes the warning object, storage managing server 10000 is about the required LU of this warning output, and the specific residual life time is lower than the dish device of threshold value, confirms whether to have specified automatic transfer about this dish device.Specifically, be lower than whole dish device of putting down in writing among " driver No. tabulation " hurdle 245h of the LU admin table 245 ' under the LU of threshold value for the residual life time, investigation driven management table 240 ' " residual life time " hurdle 240h, the specific residual life time is lower than the dish device of threshold value, in addition, confirm " shift automatically and specify " hurdle 249 (step 4030).

In step 4030, in constituting whole dish devices of LU that the residual life time is lower than threshold value, to have specified under the situation of automatic transfer, storage managing server 10000 is only exported warning.At this, also do not begin under the situation of automatic jump operation at disc array device, indication is shifted automatically.The concrete example of content by way of caution exemplifies " monitoring constantly ", " the device ID of disk array ", " LU No. ", " the residual life time is lower than the driver No. of threshold value ", " related copy source or duplicate device ID and the LU No. of destination LU ".

Have again, when output " related copy source or duplicate device ID and the LU No. of destination LU ", use the LU that remains in the storage archival memory 13200 information admin table 246.By output " related copy source or duplicate device ID and the LU No. of destination LU ", the user be identified as immediately for the LU that warns object to other copy source or duplicate the possibility that destination LU influences, according to circumstances, can implement to be accompanied by the manual change (step 4040) right above the LU of residual life.

In step 4030, one all under the situation of not specifying transfer automatically in constituting the dish device of LU that the residual life time is lower than threshold value, storage managing server 10000 is when exporting warning to the user, and output promotes to shift the information of appointment or driver exchange indication.The concrete example of content except that the content of step 4040 record, also exemplifies " the driver No. that becomes transfer destination ground candidate " by way of caution.

When decision " becomes the driver No. of transfer destination ground candidate ", storage managing server 10000 uses the driven management table 240 that remains in the storage archival memory 13200 ' information, the not use dish device that decision retrieval identical type of driver also useless is carried out the array structure setting.Utilize the output of " the driver No. that becomes transfer destination ground candidate ", the user can alleviate the burden that retrieval is used to shift the transfer destination ground LU of appointment.Have again, also can pass through the disk array automatic transfer destination ground of decision dish device automatically, the output (step 4050) of omitting " the driver No. that becomes transfer destination ground candidate ".

According to present embodiment, when generation is mixed with the LU of disk array of the input and output I/F disc driver different with reliability, can consider that the Performance And Reliability of disc driver generates LU based on the characteristic of user's indication.In addition, under the situation of the function of in disk array has remote copy and device, duplicating, duplicate destination LU, also can consider that the Performance And Reliability of disc driver automatically determines based on the characteristic of user's indication about this.In addition, concerning the user, can enough logical volumes be the grade of LU, the life-span management of the dish device in the disk array that is difficult to realize.

Have again, the storage managing server 10000 of present embodiment, even disk array 701 is the such structures of second embodiment, Performance And Reliability that also can consideration dish device carries out the generation of LU and the life-span management of LU grade.

Have, in the present embodiment, the control part 200 of the disk array 700 in also can first embodiment is carried out the life-span management of the LU unit that storage managing servers 10000 carry out again.Under this situation, can be the LU residual life also with the parameter modification in the life-span management of the dish device unit that in first embodiment, carried out.Then, can detect the dish device that constitutes this LU, confirm to have or not the processing of automatic transfer etc. respectively about it having found that the LU residual life is lower than under the situation of LU of certain threshold value.

According to the present invention, just can in same disk array, be mixed with the reliability dish device different and manage with life-span or residual life.In addition, can consider the reliability of the dish device of forming array, user's purposes of decision volume.In addition, can consider the performance of the dish device of forming array, user's purposes of decision volume.In addition, can be mixed with the different dish device of control mode, in same disk array, manage.

In addition, having to be the effect that connects the ata disk device in the fibre channel arbitrated loop at the interface that connects optical-fibre channel dish device.

In addition, can provide a kind of Performance And Reliability of having considered the dish device LU automatic generation and by the life-span management of LU grade, alleviated the administrative burden of user's memory device system.

Claims (16)

1. A storage system (700,700 ', 701A), it is characterized in that comprising:

   The a plurality of dish devices (301,401) that comprise the first dish device (301) and the second dish device (401), and

   The control part (200) that is connected with described a plurality of dish devices, wherein

   The predetermined life-span (240g) of the described first dish device (301) is different with the predetermined life-span of the described second dish device (401),

   Described control part (200) is used to preserve the accumulation effective time (240f) of each described dish device, receive described a plurality of dish device (301, the described predetermined life-span (240g) of each dish device 401) also preserves the described predetermined life-span, utilize the described predetermined life-span and the described accumulation of each described dish device to manage described storage system (700 effective time, 700 ', 701A).
2. 2. storage system as claimed in claim 1 is characterized in that:

   It is predetermined that the described predetermined life-span (240g) of each the dish device in described a plurality of dish devices (301,401) is based in described a plurality of dish device repeatedly averaging time between the fault of each dish device.
3. 3. storage system as claimed in claim 1 is characterized in that:

   It is predetermined that the described predetermined life-span (240g) of each the dish device in described a plurality of dish devices (301,401), the specification of each the dish device in the described a plurality of dish devices that provided by manufacturer was provided.
4. 4. storage system as claimed in claim 1 is characterized in that:

   The described first dish device (301) has first interface, and the described second dish device (401) has second interface,

   The described predetermined life-span (240g) of the described first dish device (301) is longer than the described predetermined life-span (240g) of the described second dish device (401).
5. 5. storage system as claimed in claim 4 is characterized in that:

   Described control part (200) is arranged to the dish device that utilizes same type and forms the RAID group, makes the dish device (301) of winning form at least one group the one RAID group, and the second dish device (401) forms at least one group the 2nd RAID group.
6. 6. storage system as claimed in claim 1 is characterized in that:

   Described control part (200) is arranged to and utilizes its predetermined life-span (240g) to form the RAID group for the dish device of par.
7. 7. storage system as claimed in claim 6 is characterized in that:

   Described control part (200) is arranged to and receives described storage system (700,700 ', the purposes of RAID 701A) group, and whether satisfy based on the described predetermined life-span (240g) of each the dish device that forms described RAID group the purposes that receives required condition judgement and export warning.
8. 8. storage system as claimed in claim 7 is characterized in that:

   The described purposes of RAID group represents to be stored in the type of the data in the described RAID group.
9. 9. storage system as claimed in claim 8 is characterized in that:

   The described predetermined life-span (240g) of the described first dish device (301) is longer than the described predetermined life-span (240g) of the described second dish device (401),

   When above-mentioned control part (200) receives database purposes by the RAID group of second dish device (401) configuration, described control part be arranged to based on the described second dish device do not satisfy described database purposes required condition judgement and export warning.
10. 10. storage system as claimed in claim 1 is characterized in that:

    Described control part (200) is arranged to carries out following operation:

    Management residue pot life (240h), described residue pot life are described predetermined life-span (240g) and described accumulation effective time (240f) poor of each dish device in described a plurality of dish devices (301,401), and

    Select logical block, described logical block will be stored the Backup Data of the data of storing in another logical block according to the described residue pot life (240h) of each the dish device that disposes described logical block.
11. 11. storage system as claimed in claim 1 is characterized in that:

    Described control part (200) is arranged to the described a plurality of dish devices (301 of output, 401) the described predetermined life-span (240g) and described accumulation effective time (240f) of each dish device in make display device (520) can show the expected life of each dish device in described a plurality of dish device.
12. 12. storage system as claimed in claim 1 is characterized in that:

    Described control part (200) is arranged to: by controlling beginning and stopping the operation that each coils device, control described accumulation effective time (240f) of each dish device in described a plurality of dish devices (301,401).
13. 13. storage system as claimed in claim 1 is characterized in that:

    Described control part (200) is arranged to: according to predetermined life-span of described predetermined life-span (240g) of a dish device and described accumulation effective time (240f) and another dish device and accumulation effective time, the data of storing in the described dish device are transferred to described another coil in device.
14. 14. storage system as claimed in claim 13 is characterized in that:

    Described control part (200) is arranged to: when the described accumulation of dish device when effective time, (240f) was equal to or greater than the described predetermined life-span (240g) of described dish device, the data of storing in the described dish device are transferred to another coil in device.
15. 15. storage system as claimed in claim 14 is characterized in that:

    The described first dish device (301) is the FC dish device with FC interface (205),

    The described second dish device (401) is the ata disk device with ata interface (206),

    The described predetermined life-span that the described predetermined life-span (240g) of described FC dish device is longer than described ata disk device.
16. 16. storage system as claimed in claim 1 is characterized in that:

    Described control part (200) is arranged to carries out following operation:

    Management residue pot life (240h), described residue pot life are described predetermined life-span (240g) and described accumulation effective time (240f) poor of each dish device in described a plurality of dish devices (301,401), and

    Its residue pot life (240h) is shorter than the data of storing in the dish device of the schedule time transfers to another dish device.

Images