1292554 九、發明說明: 【發明所屬之技術領域】 本發明係有關於—種電腦資 二資:购元失效狀況回應處理方法及 ==?制介面、且該存取控制介面係二= 為-陣列切碑:飼服器群組)與一細存單元(例如 存單==ΐ)之間’用以對該電腦系統與該資料儲 ,處理功此’亦即可於該資料儲存單元發生暫時性的失 錄糊如為熱_叙餘、進行纟統 :,對此失效狀況進行一特殊之回應處理程序丄:寺而〕 亥ίΓ控制介面於此失效狀況發生時所接收到的存取指 々 胃因此而發生錯誤之處理結果。 【先前技術】 •陣列式硬碟裝置(Re—t Array of Independent : sks,RAID)為-種具備有二個或二個以上之實體硬碟的 電腦資料儲存裝置,用以儲存數量極為魔大的電腦網路資 枓,且其可例如透過-⑽奶⑼-…刪胸⑽Μ1292554 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a computer-funded capital: a method for processing a purchase status failure condition and a == system interface, and the access control interface system is two = Array cut: the feeding device group) and a thin storage unit (such as deposit slip ==ΐ) to use the computer system and the data storage, processing can be used in the data storage unit temporarily Sexual loss is as hot as _ _ Yu, 纟: a special response to this failure condition 丄: Temple and 亥 Γ Γ control interface received access finger when this failure occurs The result of the error is caused by the stomach. [Prior Art] • Re-t Array of Independent: sks (RAID) is a computer data storage device with two or more physical hard disks. Computer network resources, and it can be, for example, through -(10) milk (9)-...deletion chest (10)Μ
SmaI1 Computer System Interface)式之存取控制介面來搭 接至-或多台網路飼服器,藉此而讓此些網路飼服器可透 過该FC/iSCSI存取控制介面來對陣列式硬碟裝置進行資 料存取動作。 ' 於貫際應用上,網路系統中的陣列式硬碟裝置常有可 能因各種緣故,例如為進行熱插拔式之拔除動作時、進行 18377 5 1292554 系統組態變更時、耸望 丁 失效狀況下,若1 而發生暫時性的失效狀況。於此 t . 仔取硿制介面持續接收到伺服哭所傳逆、 =:指令,則由於此時陣列式硬碟裝置已;過 疒此些存取指令的處理動作產生錯誤結果而需由 ==面物出存取指令的_發出二:由 :二硬碟裝置未回復到正常連線狀態 因此會多消耗不二 =?;;將導致錯誤之 資訊處理效能。 以里―間,因此而降低整體之網路 【發明内容】 鑒於以上所述習知技 是在於提供一種資 “,本电明之主要目的便 統,其可於網路“"早兀失效狀況回應處理方法及系 _ ,, ± 糸、、先中的陣列式硬碟裝置發生暫時性的& 效狀況時,對後續之存 1曰^生的失 處理功能,_以#^ u 7 、一将殊之失效狀況回應 必要的處理 次g b而確保整體之網路資訊處理效能。 統係言”u貧料儲存單元失效狀況回應處理方法及系 介面H Γ怎用於搭配至—存取控制介面、且該存取控制 儲存:馬it於、電腦系統(例如為—伺服器群組)與一資料 —㈣式硬碟裝置)之間,用以對該電腦 存單元失效狀況_處理=_貝=取動作提供一資料儲 恭斗紅U 更功此,亦即可於該資料儲存單元 二能二^1Γ失效狀況(例如為熱插拔式之拔除、進行系統 、、且心变更、寻等)時,對此失效狀況進行一特殊之回應處理 18377 !292554 程序,以藉此而令 接收到的存取此人^取彳工制;丨面於此失效狀況發生時所 伯7个會因此而發生錯誤之處理結果。 本發明之資料俾左。口 — 統的特存早兀失效狀況回應處理方法及系 後續的存存單元發一 之資料儲在置_壬曰存至一預定之佇列暫存區來等待失效 次些存取指人回復到正常連線狀況之後再依序處理 而非如先前技術般地於發生失效狀況之後 仍對存取指令進行處後 消耗不必要的處理時門“f 處理、,、。果,因此不會 【實施方式】 守間而確保整體之網路資訊處理效能。 單元U &、j? °所附之圖式’詳細揭露本發明之資料儲存 早兀^狀_應處理方法及系統之實施例。 第1圖即顯示本發 :欠 — 理系統(如標號100所/之卢t儲存早元失效狀況回應處 構及豆物杜道A _斤‘之虛線框所包含之部分)的應用架 稱及/、物件^^向兀^彳半丨 . 、(object-oriented component mode!)的基本架構。如圖所示,本發明之資料儲 效狀況回應處理系統1〇 制介面3。、且該存取二;=广荅配至-存取控 統與至少—資料亀:係耗接於至少-電腦系 貝才Ή者存早兀20之間,用以 10與該資料儲存單元2〇夕η从-欠』丨各 ^ 儲存單元纽狀況㈣處㈣取動作提供一資料 口“慝理功旎,亦即可於該資料儲存單 元2。發生暫時性的失效狀況(例如 J存: 行系統組態變更、等等)時,對此失效狀況進卜特殊2 應處理程序,以藉此而使得該存取控制介面30於此失效狀 18377 7 1292554 况毛生日守所接收到的存取指令不會因此而發生錯誤之處理 結果。 於第1圖所示之應用實例中,存取控制介面3 〇例如 為耦接於3台伺服器與3台陣列式硬碟裝置(Redundant Array of Independent Disks,RAID)之間;但於具體實施 上,其可耦接之伺服器與陣列式硬碟裝置的數目並無限 制此外於具脰只加上,該存取控制介面3 〇所採用的介 面架構例如為FC (Flbre Channel)和iSCSI (Int⑽以冗幻, 其中 SCSI = Small Computer System Interface)結合之介面 架構;但亦可為採用其它型式之介面架構。 户理广斤示’本發明之資料儲存單元失效狀況回應 處里糸、、充100的模組化之物件導向元件模型 (object-onented component model) ^ ^ # ^ ^ =狀:編莫組11〇 ; (b)—存取指令符列模組⑽ 以及(C) 一存取指令取出模組130。 失效狀況備測模組1 1 〇可伯丨 7偵測各個貧料儲存單元20 疋否有么生一失效狀況時;若是,則回 待訊息;且可進而於該資 _ Μ ★出一佇列等 重新回複到下^ 早兀20發生失效狀況之後 重新口设到正常連線狀況時,回應地 —俊 訊息。於具體實施上 -回设正常 用-旗標來顯示各個二兄偵測⑷10例如可採 „ 貝枓儲存早兀20是否有有笋生妻1 狀況,亦即若此旗標的值為„ 2生失效 發生了失效狀、、⑴表不貝枓儲存單元20 王J天政狀况,反之’若此旗標的值 儲存單元20為正常連線狀況。 則表示資料 18377 8 1292554 麵指令佇列模組12〇具有一佇列暫 =上述之失效狀況細組n。所 :其 息而將該存取控制介面3〇 丁歹J寺得讯 況之後所接以 早兀20發生失效狀 121。 到的所有的存取指令均暫存至騎列暫存區 存取指令取出模組j 3〇可 組110所產生& i # 心I之失效狀況偵測模 組m中的件列暫 /存取才曰令传列模 .. 丄T依一預疋之順序方式 為先進先出⑺rst In Flrst 0ut,FI 序方式,例如 取出其中之夂伽六;貝序方式,來依序 八 口们存取指令,並將各個取出之存取> a# 給該存取控制介面3G,令該存取㈣介面m令傳送 料錢1_資料儲存單元⑼進行所需之資 請參閱第1圖,於眘 處於正常㈣h 若資料儲存單元20 、連線狀況’則存取控制介面3 0會依序卢理千㈣/ 統10所發送過來的各個存取指令 ^电細糸 對資料儲存單元20進 。表甩驷糸統10可 資_存單元2。處於暫時性的失效狀況時 待…, 測到此狀況而回應地發出-娜 待u存取指令㈣模組i 2 人 τ列寺 =將存取控制介面30於輸存單;; =所接收到的所有的存取指令均暫存至作;=狀 當發生失效狀況 之貧料儲存單元20重新再回復到 正 18377 1292554 常連線狀況時,其即可 發出—連線回復正常1自=效狀況偵測模組no回應地 取指令取出槎㈣广子取指令取出模組⑽,令存 順序方式,來取出1由 、 例如為先進先出之 存取指令依序傳送二存取指令,並將各個取出之 面30執行該存取 工制介面30’令該存取控制介 料儲存單元2。進行復到正常連線狀爾 況由於本發明之資_存單元失效狀· 後,將所有元-…效狀況之 失效之資料儲存單元2〇 —曰子 仔列暫存區來等待 依序處理此些存取指入新回復到正常連線狀況之後再 狀況之後仍對存取指曰:進:=先前技術般地於發生失效 因此不會消耗不必要的=而造成錯誤之處理結果, 理效能。 才3而確保整體之網路資訊處 制介面、且該存取控料面其可應用於搭配至一存取控 少一資料儲存單元之門 ”、馬妾於至少一電腦系統與至 單元之間的資料存取動作提/ί 4糸統與該資料儲存 應處理功能,·且复特點在^、—貧料儲存單元失效狀況回 況後,將所有後續的存:::::::存單元發生失效狀 區來等待失效之資料儲存單^存至—預定之仲列暫存 重新回復到正常連線狀況之 18377 1292554 =依序處理次些存取指令,而非如先前技術般地於發生 ^双狀況之後仍對存取指令進行處理而造成錯誤之處理結 却考口此:昔消耗不必要的處理時間而確保整體之網路資 二处理效A。本發明因此較先前技術具有更佳之進步性及 貫用性。 =上料僅為本發明之較佳實施例而已,並非用以限 :瘙::。貝技術内容的範圍。本發明之實質技術内容 我地Μ於下述之申請專利範圍中。若任何他人 術實體或方法與下述之申請專利範圍 是r種等效之變更,均將被視為涵蓋於本發 月之申3月專利乾圍之中。 【圖式簡單說明】 儲二1 r為一系統架構示意圖,其中顯示本發明之資料 ^子早兀效狀況回應處理系統的應用架構及其物件導 元件模型的基本架構。 、σ 【主要元件符號說明】 10 電腦系統(伺服器) 20 30 100 110 120 121 資料儲存單元(RAID) 存取控制介面(FC/iSCSI) 本發明之資料儲存單元失效狀況回應處理系統 失效狀況偵測模組 存取指令彳宁列模組 佇列暫存區 存取指令取出模組 18377 11 130SmaI1 Computer System Interface) access control interface to connect to - or multiple network feeders, so that these network feeders can be arrayed through the FC/iSCSI access control interface The hard disk device performs data access operations. For continuous applications, array hard disk devices in network systems are often subject to various reasons, such as when performing hot-swappable extraction operations, when performing 18377 5 1292554 system configuration changes, In the case of a temporary failure condition if 1 occurs. In this case, the 硿 interface is continuously received by the servo crying, and the =: command, because the array hard disk device has already passed; the processing result of the access commands has an error result and needs to be = = IE outgoing access command _ outgoing two: by: two hard disk device does not return to the normal connection state, so it will consume more than two =?;; will lead to the wrong information processing performance. In order to reduce the overall network [invention] In view of the above-mentioned conventional technology is to provide a kind of capital, the main purpose of this electric power is that it can be used in the network "" early failure status In response to the processing method and the system _ , , ± 糸 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , A special failure condition is responded to by the necessary processing of the gb to ensure the overall network information processing performance. The system is used to match the access control interface, and the access control storage: the computer system (for example, the server group) Between the group and the data-(four)-type hard disk device, it is used to provide a data storage unit failure status _ processing = _ be = take action to provide a data storage gongdouhong U more, this can also be used in the data storage unit A special response to the failure condition, 18377 !292554, is used to make a failure condition (for example, hot plug removal, system removal, heart change, seek, etc.) The received access to this person ^ takes the work system; after the failure condition occurs, 7 cases will result in an error. The information of the present invention is left. The status response processing method and the subsequent storage unit send a data stored in a predetermined queue to wait for the failure of the access to return to the normal connection status and then sequentially Processing rather than as prior art After the birth of the access failure condition after the instruction is still at the door consume unnecessary processing "f ,,, process. Therefore, it does not [Implementation] Guardianship to ensure the overall network information processing performance. The unit U &, j? ° attached to the drawings </ RTI> detailing the data storage of the present invention, the embodiment of the method and system. Figure 1 shows the application of the present invention: the system of the under-system (such as the part of the dotted line box of the 100-year-old t-stored early-stage failure status response and the bean-dot road A _ jin's) And /, the object ^ ^ to 兀 ^ 彳 彳 彳 , (object-oriented component mode!) the basic structure. As shown in the figure, the data storage status response processing system 1 of the present invention controls the interface 3. And the access two; = wide distribution to - access control system and at least - data 亀: is consumed by at least - the computer system is only between 20, for 10 and the data storage unit 2〇夕η从-欠』丨^ The storage unit Newton condition (4) (4) Take action to provide a data port : line system configuration change, etc.), this special condition should be processed for this failure condition, so that the access control interface 30 is received in this failure state 18377 7 1292554 The access command does not cause an error processing result. In the application example shown in FIG. 1, the access control interface 3 is, for example, coupled to three servers and three array hard disk devices (Redundant Array). Between the Independent Disks (RAID); however, the number of servers and arrayed hard disk devices that can be coupled is not limited. In addition, the access control interface is used. Interface architecture such as FC (Flbre Channel) and iSCSI (Int(10) for redundancy Fantasy, SCSI = Small Computer System Interface), but it can also be used in other types of interface architecture. Grouped object-onented component model ^ ^ # ^ ^ = shape: edit group 11〇; (b) - access instruction array module (10) and (C) an access instruction fetch module Group 130. Failure status test module 1 1 〇可丨 7 detects each of the poor storage units 20 疋 whether there is a failure condition; if yes, then the message is returned; and the information can be further _ Μ ★ Re-respond to the next line after the failure status is changed to the normal connection status, respond to the local-jun message. In the specific implementation - set the normal use - flag to display each of the two brothers Detection (4) 10, for example, can be used to store 兀 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否Tianzheng situation, and vice versa if the value of this flag is stored Element 20 is the normal connection status. It means that the data 18377 8 1292554 surface command queue module 12 has a queue temporary = the above-mentioned failure condition fine group n. The information is controlled by the access control interface 3 After the 歹J Temple has received the message, it will be terminated by the early 20th. All the access commands are temporarily stored in the riding staging area access command fetch module j 3 〇 group 110 generated & i #心I's failure condition detection module m in the list of temporary / access orders to pass the module. 丄T in a pre-ordered way for the first in first out (7) rst In Flrst 0ut, FI order mode, For example, the gamma hexagram is taken out; the sequel mode is used to sequentially access the instructions, and each of the extracted accesses >a# is given to the access control interface 3G, so that the access (four) interface m is used to transmit the material. Money 1_Data Storage Unit (9) Please refer to Figure 1 for the required resources. Yu Shen is in normal (4)h. If the data storage unit 20 and the connection status are 'the access control interface 3 will be followed by the order (4) / system 10 Each of the accessed access commands is sent to the data storage unit 20. The meter system 10 can be stored in the unit 2. Waiting for a temporary failure condition, detecting this condition and respondingly issuing - Na to u access command (4) Module i 2 person τ column temple = access control interface 30 on the deposit slip;; = received All the access commands are temporarily stored; = when the poor condition storage unit 20 of the failure condition is restored to the positive 18377 1292554 normal connection condition, it can be issued - the connection returns to the normal 1 self-effect status The detecting module no responds to the fetching command to remove the 四 (4) 广子 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取 取Each of the removed faces 30 executes the access interface 30' to cause the access control media storage unit 2. After returning to the normal connection state, due to the failure of the resource-storage unit of the present invention, all the data storage units 2 曰 曰 仔 列 暂 暂 暂 暂 暂 暂 来 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待 等待These accesses refer to the new replies to the normal connection status and then the access indicators after the status: In: = previously technically failed, so it does not consume unnecessary = and the result of the error processing, efficacy. Only 3 to ensure the overall network information processing interface, and the access control surface can be applied to the door of an access control and less data storage unit", the horse is at least one computer system and to the unit The data access action between the two / 糸 4 system and the data storage should be processed, and the complex features in the ^, - poor material storage unit failure status, all subsequent storage::::::: The memory unit has a failure zone to wait for the invalid data storage list to be stored. - The predetermined secondary column is temporarily restored to the normal connection state. 18377 1292554 = The next access command is processed sequentially instead of the prior art. After the occurrence of the double condition, the processing of the access instruction still causes the error to be processed. However, the unnecessary processing time is consumed to ensure the overall network processing efficiency A. The present invention therefore has a higher performance than the prior art. Further improvement and versatility. The loading is only a preferred embodiment of the present invention, and is not intended to limit the scope of the technical content of the present invention. The technical content of the present invention is as follows. In the scope of the patent application. Any other entity or method and the following patent application scope are equivalent to r changes, which will be considered to be covered in the patent for the month of March of this month. [Simplified diagram] It is a schematic diagram of a system architecture, which shows the application architecture of the invention and the basic architecture of the object component model of the object. σ [Key component symbol description] 10 Computer system (server) 20 30 100 110 120 121 Data Storage Unit (RAID) Access Control Interface (FC/iSCSI) Data Storage Unit Failure Status Response Processing System Failure Status Detection Module Access Commands Access instruction fetch module 18377 11 130