1263892 九、發明說明： 【發明所屬之技術領域】 本發明係有關於一種電腦資訊技術，特別是有關於一 種硬碟陣列重建程序中斷接續處理方法及系統，其可鹿用 於搭配至-硬碟陣列裝置，例如為—raid㈤㈣耐1263892 IX. Description of the Invention: [Technical Field] The present invention relates to a computer information technology, and more particularly to a method and system for interrupting and processing a hard disk array reconstruction program, which can be used for collocation to a hard disk Array device, for example -raid (five) (four) resistant
Array of Independent Dlsks)硬碟陣列裝置，用以對該 咖硬碟陣列裝置提供—重建程序中斷接續處理功能，藉 ^而讓該咖硬碟陣列裝置若於進行重建程序（rebuild) 時遭遇到何翻之f力巾斷狀況，料於後續進行復原 建私序日守，從先别中斷之處開始接續地進行未完成之 重建工作’而不必重新從頭開始進行整個的重建程 【先前技術】 陣列式硬碟系統（Redundant Array indepen^ 儲^ 1’ WRAID)為—種具備有多個實體硬碟單元的電腦資^ 數：二其通常係應用於搭接至網路飼服器，用以⑴ =古，龐大的電腦網路資料。由於RAID硬碟陣 /、備有夕個硬碟單元，因此可提供一 提:整體之存取效率，並亦可提供-多 子二二使侍電腦資料的儲存具有更高的妥善性。 元it y 用上’RAID硬碟陣列裝置中的多個硬碟單 為r之硬碟單元和備用之硬碟單元; 而備用之硬磾單：：：於：吊狀況下用來儲存電腦資料; 生損毀狀況== 中任何-個主用之硬㈣ P被利用來進打—重建程序（rebuild)， 18648 1263892 藉此而將發生損毀狀況之硬碟單元中原先 重建於備用之硬碟單元上，令飼服器可從 貝枓轉而 中讀取到發生損毀狀況之硬碟單元中原先儲存的石^單元 具體實施上，RAID硬碟 存的-貝料。於 (S叩erb1〇ck)的特定儲在厂±置#'^—名稱為超區塊 屬性及來儲放各個硬碟單元的相閛 之硬碟單元、何者為發生損浐 ^茱早兀和備用 建後的硬碟單元、等等。 彳、硬碟單兀、何者為重 然而於實務上，咖硬碟 =遭遇到不可預期之電力中斷狀況而;皮=中： 管理人員再重新啟動_重^=新回设及開機後，若系統 序會從頭開始進行敕個 2則重新啟動後的重建程 開始接續地進行未完成之^工作’而非從先前中斷之處 | # # % ^ 重建工作。由於此緣故，因此口 要重建過程發生電力申齡 U此，、 八二4 | > 中断狀况，則將使得已完成重建的邻 刀刖功盡棄。由於重建程 ， 系統管理效能。所知用之重建功能顯然會降低整體之 【發明内容】 鑒於以上所述先前技術之缺 是在於提供—種硬 本毛月之主要目的便 統，：】重1私序中斷接續處理方法及系 DtT '、 重建程序中斷接續處理功能，藉此而讓 況D貝Γ碟裝置若於進行重建程序時發生電力中斷狀 /、可於U進行復原性重建程序時，從先前中斷之處 18648 1263892 ::也進仃未兀成之重建工作，而不必重新從頭開始 進仃整個的重建程序。 本發明之另—目的在於提供一種硬碟陣列重建程序 1=方法及系統，其可增進整體之—系 本&明之硬碟陣射建料巾斷接續處理方法及李 先係广十來應用於搭配至一硬碟陣列裝置，例如二 ， ant Array of IndePendent Disks)硬碟陣列裝 ί二Γ請1D硬碟陣列裝置提供-重建程序中斷接 二，痒广，猎此而讓該RAID硬碟陣列裝置若於進行重 调：二5時遭遇到不可預期之電力中斷狀況，則 個的重建程序。冑作’而不必重新從頭開始進行整 本發明之硬碟_重建程料斷接續處理方法至少 匕3以下構成要件：⑴於該硬碟陣列裝置中的 :被用來進行重建程序時，逐步記錄下該硬碟單元中已: 之區塊的相關辨識資料，並將逐步記錄下來的資: 作為-組中斷點資料而儲放於一永久性 —、、八 硬碟陣列裝置發生電力中斷狀況時，可久=:該 保有該組中斷點資料；⑵若一…久性储存區 時遭遇到電力中斷狀況，則於:復建程序 性儲存區所儲放之中斷點資料:重亥:久 區塊及尚未完成重建之區塊―:::= 18648 1263892 .執行-復原性重建程序，藉以從尚未完成重建 接績地進行先前因中斷狀況而未完成之重建工作” ° 於實體架構上，本發明之硬碟陣列重建程序中斷接芦 二:ΐ =二含:⑷一中斷點記錄模組，其可於該“ 記錚下 1 單元被用來進行重建程序時，逐步 並將、豕：硬碟早兀中已完成重建之區塊的相關辨識資料， 步記錄下來的資料作為—組中斷點資料而儲放於一 .時久=該硬碟陣繼發生電力中斷狀況 了？ 3亥水久性儲存區保有該組中斷點資料.rh) 一士 =資料讀取模組’其可於該硬碟單元若於進行重 =遇到電力中斷狀況並接著回復電力之後，回應 點資料，藉以判別出已3=彔換組所記錄之中斷 區塊；以及⑷—ΐΐ:建之區塊及尚未完成重建之 模組所讀取出之中Si;二可依據該中斷點資物 以rw〜 辦』貝科來執行一復原性重建程序，藉 而未完成區塊開始接續地進行先前因中斷狀況 統的二建程序中斷接續處理方法及系 的相已重建完成之區塊 逐步記錄下來的資料作4 成之區塊的編號，並將 性儲存區，m 4 中斷點資料而儲放於—永久 況的硬磾單：：的该硬碟陣列裝置中的其它未發生損毀狀 發生電力儲存區，㈣ 了可用來作為一組中斷點資料，使得電 18648 8 1263892 中斷之處開始接續地進行未完成之重 ==必如先前作法般地需從頭開始進行整個的重 重建程序發生電力中斷狀況之後，更 而有效率地完成復原性之重建 體之網路系統管理效能。 J a退正 【實施方式】 以下即配合所附之圖式， 坪細揭路况明本發明之硬碟 >建私序中辦接績處理方法及系統之實施例。 理季：/二即，發明之硬碟陣列重建程序中斷接續處 式ίίϋ：^100所指之虛線框所包含之部分）的應用方 式及…架構的物件導向元件模型(〇bΜ—οηe則 ΓΓΓη 1)。如圖所示，本發明之硬碟陣列重建程 断接續處理系統1〇。於實際應用上係搭配至 ^ 置，例如為—R仙（R—Arrayoiindependent 工Sks)硬碟陣列裝置2Q ;亦即整合至該咖 硬物驅動單元％，且該硬碟陣列驅動單元30 作日:太:電腦平台1〇’例如為一網路词服器。於實際操 ^ ㈣之硬碟陣射建程序中斷接續處理系統100 來對該RAID硬碟陣列裝置2Q提供—重建程序中斷 、、另处理功能，藉此而讓該RAID硬碟陣列裝置2〇若於進 :重建程序(rebuild)時遭遇到不可預期之電力中斷狀 RAID則該電腦平台1〇重新回復電力及開機之後、或將 ^陣列裝置2{)拆移至其它未發生電力中斷的電腦 。於圖式中顯示）上時，令重新啟動之重建程序從先 18648 9 1263892 前中斷=處開始接續地進行未完成之重建工作。 置圖所示之實施例中，假設該腿硬碟陣列穿 ==硬碟單元21、22、23、24、25，其中： Μ則作為一備用3之為t用之硬碟單元，而硬碟單元 付- 硬碟早70 2 5 (註：第1圖所示之f祐如 一不範性地顯示RAID硬碟陣列裝置2〇具有5個硬碟單1 :際應用上，ID硬碟陣列裝置2。中的硬碟單 里可旎為更多而並無限制）。 ’、 如弟1圖所示，本發明之硬碟陣列重建程序 处理系、统1 00之實體架構的物件 伽c⑽卿entm。㈣至少ί含.（a)中 (^占^模組110;⑻一中斷點資料讀取模組12〇;以及 重建組13G。於具體實施上，本發明之硬碟陣列 =序中斷接續處理系統100例如可完 r w 將此_私式例如以軟體或勅體之附加模組 敕a入-on module)方式整合至該電腦平台的作業系統或 更碟陣列驅動單元30所採用之驅動程式，藉此而 i、所而之重建程序中斷接續處理功能。 的1!^己錄模組UG可於該RAID硬碟陣職置20中 之硬碟單元(21、22、23、或24)發生損毀狀況 =備用之硬碟單元25來進行重建程序時，被啟動來逐 =錄下該備用之硬碟單元25中已重建完成之區塊的相 =辨識資料（亦即每當備用之硬碟單元25上完成一個區塊 或-預定數量之區塊的重建工作時，即回應地將此些已完 ]〇 18648 1263892 =建的區塊的編號記錄下來），並將逐步記錄下來的資料 二資料而儲放於-永久性儲存區， ^ 十口 10、或是儲存至RAID硬碟陣列裝置2〇Array of Independent Dlsks) for providing the hard disk array device with a rebuild program interrupt processing function, which allows the hard disk array device to encounter a rebuild process. Turning over the force of the towel, it is expected that the subsequent restoration will be carried out, and the unfinished reconstruction work will be carried out from the beginning of the interruption. Without having to re-start the entire reconstruction process from the beginning [Prior Art] Array Redundant Array indepen^ (1) WRAID is a computer with multiple physical hard disk units: 2 is usually used to lap to the network feeder for (1) = Ancient, huge computer network data. Since the RAID hard disk array / has a hard disk unit, it can provide a whole: access efficiency, and can also provide - multi-second two to make the storage of computer data more appropriate. Yuan it y uses multiple hard disk units in the 'RAID hard disk array device' to be the hard disk unit of r and the spare hard disk unit; and the spare hard disk::: used to store computer data under the hanging condition ; damage condition == any of the main hard (4) P is used to play - rebuild, 18648 1263892 The hard drive unit that will be damaged in the original hard disk unit On the top, the feeding machine can read from the shellfish to the stone unit that was originally stored in the hard disk unit in which the damage occurred, and the storage of the hard disk stored in the RAID hard disk. The specific storage in the (S叩erb1〇ck) factory ± set # '^ - the name is the super block attribute and the corresponding hard disk unit to store each hard disk unit, which is the occurrence of damage ^ 茱 early 兀And the spare built-in hard drive unit, and so on.彳, hard disk single 兀, which is heavy but in practice, coffee hard disk = encountered unexpected power interruptions; skin = medium: management personnel restart _ heavy ^ = new reset and after boot, if the system The sequence will start from the beginning 2, then the rebuild process after restarting will continue to perform the unfinished work' instead of the previous break | # # % ^ Rebuild work. For this reason, the power generation period of the reconstruction process will be abandoned, and the interruption of the situation will cause the neighboring knives that have completed the reconstruction to be abandoned. Due to the redevelopment process, system management effectiveness. It is obvious that the reconstruction function used will reduce the overall content. [Invention content] In view of the above-mentioned prior art, the shortcoming of the prior art is to provide the main purpose of the hard-working month, :] heavy 1 private sequence interrupt connection processing method and system DtT ', the rebuild program interrupts the processing function, thereby allowing the D device to generate a power interruption when performing the rebuild procedure, and when the U is performing a restorative reconstruction procedure, from the previous interruption 18648 1263892: : It is also a rebuilding work that has not been completed, and it is not necessary to re-enter the entire reconstruction process from scratch. Another object of the present invention is to provide a hard disk array reconstruction program 1=method and system, which can improve the overall processing method of the hard disk array construction towel and the Li Xian system. For collocation to a hard disk array device, such as two, ant Array of IndePendent Disks) hard disk array device Γ 二Γ 1D hard disk array device provides - rebuild program interrupted two, itchy, hunting this and let the RAID hard disk If the array device is undergoing a re-tuning: an unexpected power interruption condition is encountered at 2:5, then a reconstruction procedure is performed.胄 ' 而 而 而 而 而 而 而 而 而 而 不必 不必 不必 不必 不必 不必 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建The relevant identification data of the block in the hard disk unit, and the gradually recorded resources: stored as a group-break point data in a permanent-, eight-disk array device power interruption condition , Long time =: This holds the breakpoint data of the group; (2) If a power interruption occurs during the ... long storage area, then: Break point of the storage in the procedural storage area: Heavy black: Long area Blocks and blocks that have not yet been rebuilt -:::= 18648 1263892 . Execution-recovery rebuild procedure to perform reconstruction work that was previously unsuccessful due to interruptions from the completion of the redevelopment. ° On the physical architecture, The invention of the hard disk array reconstruction program interrupts the connection of the second two: ΐ = two contains: (4) a break point recording module, which can be used in the "remembering unit 1 is used to carry out the reconstruction process, step by step and 豕: hard The disc has been finished Information related to the identification data block reconstruction, step down as a record - set breakpoints and data storage in a long time = one o'clock the hard disk array following a power interruption condition.? 3Hai water storage area retains the group break point information.rh) Ashi = data reading module 'which can be used in the hard disk unit if the weight is met = power interruption condition and then resume power, the response point The data is used to identify the interrupted block recorded by the group of 3=彔; and (4)—ΐΐ: the built block and the module that has not been reconstructed are read out of Si; Performing a restorative reconstruction procedure with rw~"Becae, the unfinished block begins to continue the block-breaking process of the previous two-build interrupted processing method and the phase-completed reconstruction of the system. The information obtained is numbered in the block of 40%, and the sexual storage area, m 4 break point data is stored in the permanent state: the other hard disk in the hard disk array device is not damaged. The power storage area, (d) can be used as a set of breakpoint data, so that the 18648 8 1263892 breaks at the beginning of the unfinished weight == must be the same as the previous practice to carry out the entire re-rebuild process from the beginning of the power Interrupt After the situation, and more efficient completion of the network system of the body of the reconstruction of restoring management efficiency. J a retreat [Embodiment] The following is a description of the hard disk of the present invention in conjunction with the attached drawings. The season: / 2, the invention of the hard disk array reconstruction program interrupts the connection type ίίϋ: ^100 refers to the part of the dotted line frame) and the object-oriented component model of the architecture (〇bΜ-οηe ΓΓΓη 1). As shown, the hard disk array reconstruction process of the present invention terminates the processing system. In practical applications, it is matched to a device, for example, a R-Arrayoiindependent Sks hard disk array device 2Q; that is, integrated into the coffee hard drive unit %, and the hard disk array driving unit 30 is used as a day. : Too: The computer platform 1〇' is for example a network word server. In the actual operation (4), the hard disk array construction program interrupts the connection processing system 100 to provide the RAID hard disk array device 2Q with a rebuild program interrupt, and another processing function, thereby allowing the RAID hard disk array device 2 to Yu Jin: Unexpected power interruption in the rebuild process. The computer platform restarts the power and power on, or removes the array device 2{) to other computers that have not experienced power interruption. When shown in the figure, the restarting rebuild procedure starts the unfinished reconstruction work from the first interruption of 18648 9 1263892. In the embodiment shown in the figure, it is assumed that the leg hard disk array wears the == hard disk unit 21, 22, 23, 24, 25, wherein: Μ is used as a spare 3 hard disk unit for t, and hard Disc unit pay - hard disk early 70 2 5 (Note: Figure 1 shows the f you like a non-standard display RAID hard disk array device 2 〇 has 5 hard disk single 1: application, ID hard disk array The hard disk in device 2. can be more and not limited. As shown in Figure 1, the hard disk array reconstruction program of the present invention processes the object of the physical structure of the system 100 cc (10) ent. (4) at least ί. (a) (^ 占 ^ module 110; (8) a break point data reading module 12 〇; and reconstruction group 13G. In a specific implementation, the hard disk array of the present invention = sequential interrupt connection processing The system 100 can, for example, be integrated into the operating system of the computer platform or the driver used by the disk array driving unit 30 by means of a private module such as a software or an add-on module. In this way, the reconstruction program interrupts the connection processing function. The 1!^ recorded module UG can be used when the hard disk unit (21, 22, 23, or 24) of the RAID hard disk array 20 has a damaged state = the spare hard disk unit 25 for the reconstruction process. It is activated to record the phase=identification data of the reconstructed block in the spare hard disk unit 25 (that is, each time a spare block or a predetermined number of blocks is completed on the spare hard disk unit 25) When rebuilding the work, it is necessary to record the number of the block that has been completed] 〇18648 1263892 = the block is built, and the data recorded in the step-by-step data is stored in the - permanent storage area, ^ ten 10 Or save to a RAID hard disk array device 2〇
匕未發生損毁狀況的硬磾單元22、23、24、W N 為最佳之，… 早 23 24 25’但以後者 2〇拆移至因為此作法可將_硬碟陣列裝置 示）來谁二 生電力中斷的電腦平台（未於圖式中顯 丁）末進订後原性重建程序，令並 斷點資料。若為計 買取到中 22 …子至其它未發生損毀狀況的硬碟單元 斷 25 ,則於具體實施上，如第2圖所示，此中 已重係將她錄下之中斷點資料(即 單 ▲、編唬）寫入至未發生損毀狀況的硬碟 規矿所^3、24、25上的一特定之儲存區，例如為raid 規耗所定義之超區塊(superblock)儲存區40。 心：：：=°可於該_硬碟陣⑽置 開機之後，回電力中斷狀況並接著重新 之中斷點記輸且12。二己奸：求事件201而讀取上述 已完成重建之區塊及尚斷點資料，藉以判別出 兀成重建之區塊。於呈體每# 二所若中斷點記錄模組110係將中斷點資料儲放至二 圖所不之未發生損毀狀況之硬碟單元22、23、24、 =:存區40,則中斷點資料讀取模組 令 中斷點資料。 免儲存£4〇中項取出所需之 重建模組130可依據上述之中斷點資料讀取模組12〇 18648 11 1263892 二讀取出之中斷點f料而令硬碟陣列驅動單元3 q執行一 Ή生之重建私序，藉以從尚未完成重建之區塊開始接續 j進行先前因中斷狀況而未完成之重建工作。舉例來說， 若中斷點資料顯示已完成重建的最後一個區塊的編號為 31:則此復原性重建程序即可首先從編號為犯的區塊開始 進仃重建工作。於具體實施上，此重建模組^別所執行之 重建料需包括—初始之快取狀態及寫入缓衝區狀態判斷 t驟’藉此而判斷該RAID硬碟陣列裝置2〇目前的快取記 2體和寫人緩衝區使用狀態是否為開啟狀態；若是，則於 貝丁執仃重建耘序岫將其使用狀態均切換成關閉狀態，藉 此而確保重建之貧料可確實地被寫入至硬碟陣列穿 中被用來重建的備用硬碟單元25;並於重建程序^ 成後，再將快取記憶體和寫人緩衝區❹㈣ 前的操作模式狀態（即尚未執行該重建程序之前的狀態）。 於以下之應用實例中，假設raid硬碟陣列裝置20 ::::單元21、22，，為主用之硬碟單元，而硬碟 生r貝i為備用之硬碟單元；且假設主用之硬碟單元21 j J貝毀狀況而令硬碟陣列驅動單元3。對備用之硬碟 =、㈣進行一重建程序，但該重建程序的進行過程中卻因 為延：到不可預期之電力中斷狀況而被迫中止。 木、2叫茶閱第1圖和第2圖，於上述之假設狀況下’ 二之重建程序開始進行之後，每當備用之硬碟單元烈 上兀成一個區堍或_搞a 曰 點_組110即可二地::二=:工作時，中斷 口 4地立即ό己錄下此些已完成重建的 18648 1263892 區 第 h的相關辨識資料，例如為此些區塊的編號；並進而如 :圖所示般地將記錄下來的區塊編號資料作為中斷點資 “儲存至各個其它未發生損毀狀況的硬碟單元&、、 :4:25上的超區塊儲存區4〇。若此初始之重建程序未發生 书力中斷狀況而順利完成，則超區塊儲存區4q中所 中斷點資料將於完成之後被消除掉；反之1發生不可預 教電力中斷狀況，則發生電力中斷狀況之前已完成重建 的編號資料即可保存於各個其它未發生損毀狀況的 石”單兀22、23、24、25上的超區塊儲存區4〇。 當電腦平台10回復電力及重新開機後（或嶋硬碟 p幻衣置20被拆移至其它未發生電力中斷的電腦平台）， 則本發明之硬碟陣列重建程序中斷接續處理I統⑽中的 中斷點資料讀取模組12G即可回應—重建程序重啟要求事 1 2%而令硬斜列驅動單元_其它未發生損毁狀況的 、”單兀22、23、24、25上的超區塊儲存區4〇上讀取上 述之中斷點記錄模組12〇所記錄之中斷點資料（亦即已完 成重建之區塊的編號資料)’藉以判別出已完成重建之區塊 =尚未完成重建之區塊，並將尚未完成重建之區塊的編號 傳送給重建模組130 ’令重建模組13〇回應地啟動一復原 性重建程序來重建其餘尚未完成重建之所有的區塊H 際執行重建程序之前，重建模組130會首先執行一初始之 快取狀態及寫入缓衝區狀態判斷步驟，藉此而判斷該隱 ，石^列裝置20目前的快取記憶體和寫入緩衝區使用狀 悲是否為開啟狀態；若是’則將其使用狀態均切換成關閉 18648 ]3 1263892 狀態，藉此而確保重建之資料可確實地被寫入至raid硬 碟陣列裝置2G中被用來重建的備用硬碟單元25 ;並於重 建程序完成之後，再將快取記憶體和寫入緩衝區使用狀態 回復成先前的操作模式狀態（即尚未執行該重建程序之前 :狀^ ) ° &貝際執仃重建程序時，假設中斷點資料顯示前 次已完成重建的最後-個區塊的編號為31，則此復原性重 建私序即可首先從編號為32的區塊開始進行重建工作。 於上述之復原性重建程序中，中斷點記錄模組^ 〇亦 同樣地將持續執行中斷點記錄功能，藉以於此復原性重建 過過鞋中右又發生電力中斷狀況，則可用來啟動再一次之 2=重建程序；依此類推，直至所有的區塊均完成重建 工作為止。 。而’之本叙明提供了 一種新穎之硬碟陣列重建程 序中斷接續處理方法及李蛴 系、冼可格配至一硬碟陣列裝置， 用乂提仏一重建程序中斷接續匕· 建過程中可逐步記錄下# %’ 在於重 ，、已重建元成之區塊的相關辨識資 料’例如為已重涂6 >广1 、 4之區塊的編號，並將逐步記錄下來 的貝料作為一組中斷里上咨沐、丨 ^ ^ 畊”、、占貝枓而儲放於一永久性儲存區，例磾The hard unit 22, 23, 24, WN that has not been damaged is the best, ... 23 24 25 'but the latter 2 〇 moved to the _ hard disk array device because of this method) The computer platform for power interruption (not shown in the figure) is the original reconstruction program, and the data is broken. In the case of the purchase of 22 pieces to the other hard disk unit breaks 25, in the specific implementation, as shown in Figure 2, the break point data recorded by her has been re-emphasized (ie Single ▲, edited) to a specific storage area on the hard disk distribution site ^3, 24, 25 without damage, such as the superblock storage area defined by the raid consumption . Heart:::=° can be returned to the power interruption condition after the _ hard disk array (10) is turned on, and then the interrupt is recorded again and 12 is lost. The second person: seeking the event 201 and reading the above-mentioned reconstructed block and the breakpoint data, in order to identify the block of the reconstruction. In the presentation of each of the two if the breakpoint recording module 110 is to store the breakpoint data to the hard disk unit 22, 23, 24, =: storage area 40 of the second map, the break point is The data reading module makes the break point data. The rebuilding module 130 required for the removal of the item 4 can be executed by the hard disk array driving unit 3 q according to the above-mentioned interruption point data reading module 12〇18648 11 1263892 A re-establishment of the private sequence, in order to continue from the block that has not yet completed the reconstruction, to carry out the reconstruction work that was not completed due to the interruption. For example, if the breakpoint data shows that the last block of the completed reconstruction has the number 31: then the restorative reconstruction program can start the reconstruction work from the block numbered. In a specific implementation, the reconstruction module executed by the reconstruction module needs to include an initial cache state and a write buffer state determination t to determine the current cache of the RAID disk array device 2 Check whether the use state of the 2 body and the write buffer is on; if it is, then change the use state to the closed state after the bedding reconstruction process, thereby ensuring that the reconstructed poor material can be written reliably. The hard disk unit 25 is used to rebuild the spare hard disk unit 25; and after the rebuild program is completed, the operation mode state of the cache memory and the write buffer buffer (4) is performed (that is, the reconstruction program has not been executed yet). Previous state). In the following application examples, it is assumed that the raid hard disk array device 20 :::: units 21, 22, is the hard disk unit for the main use, and the hard disk is the spare hard disk unit; The hard disk unit 21 j J destroys the hard disk array drive unit 3. A rebuild procedure is performed on the spare hard disk =, (4), but the reconstruction process is delayed due to an unpredictable power interruption. Wood, 2, tea, read Figure 1 and Figure 2, after the above-mentioned hypothetical situation, after the second reconstruction process begins, whenever the spare hard disk unit is smashed into a zone or _ a point _ Group 110 can be two places:: two =: At work, the interrupt port 4 immediately records the relevant identification data of the h-th of the completed 18648 1263892 area, such as the number of the block; and further For example, as shown in the figure, the recorded block number data is stored as a break point to the other hard disk unit &, : 4:25 super block storage area 4〇. If the initial reconstruction procedure does not occur due to the book power interruption condition, the data of the interruption point in the super block storage area 4q will be eliminated after completion; otherwise, the power interruption occurs when the unpredictable power interruption condition occurs. The numbered data that has been reconstructed before the condition can be stored in the super-block storage area on each of the other stone-free units 22, 23, 24, and 25 that have not been damaged. When the computer platform 10 recovers power and restarts (or the hard disk device 5 is removed to another computer platform where power interruption has not occurred), the hard disk array reconstruction program of the present invention interrupts the connection processing (10). The interrupt point data reading module 12G can respond - the rebuild program restarts the request 1 2% and makes the hard oblique column drive unit _ other super-zones on the 22, 23, 24, 25 without damage The block storage area 4 reads the interruption point data recorded by the interruption point recording module 12 (that is, the number data of the block that has been reconstructed) to identify the block that has been reconstructed = the reconstruction has not been completed. The block, and the number of the block that has not been reconstructed is transmitted to the reconstruction module 130', so that the reconstruction module 13 responsively initiates a restorative reconstruction process to reconstruct all the blocks of the remaining unfinished reconstructions. Before the program, the reconstruction module 130 first performs an initial cache state and a write buffer state determination step, thereby determining the implicit cache, the current cache memory and the write buffer of the device 20 Sad No is on; if it is, then its usage status is switched to off 18648]3 1263892 state, thereby ensuring that the reconstructed data can be reliably written to the spare hard disk used for reconstruction in the raid hard disk array device 2G. The disc unit 25; and after the rebuild procedure is completed, the cache memory and the write buffer use state are restored to the previous operation mode state (ie, before the rebuild procedure has been executed: shape ^) ° & When rebuilding the program, assuming that the breakpoint data shows that the last block of the previous completed reconstruction has the number 31, the restorative reconstruction private sequence can start the reconstruction work from the block numbered 32. In the restorative reconstruction program, the interruption point recording module will also continue to perform the interruption point recording function, so that the restorative reconstruction has passed the right and the power interruption condition occurs, and can be used to start again 2= Rebuild the program; and so on, until all the blocks have completed the reconstruction work. And the 'this description provides a novel hard disk array reconstruction program interrupted connection processing The method and the Li Wei system and the 冼 格 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 配 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建 重建The relevant identification data 'for example, the number of blocks that have been repainted 6 > Guang 1 and 4, and the stepped record of the shell material as a group of interruptions in the consultation, 丨 ^ ^ 耕", and accounted for Stored in a permanent storage area, for example
如為该硬碟陣列裝詈+ J …μ AA 置中的各個其它未發生損毁狀況的硬碟 早兀上的一超區塊儲存區，藉此 中斷狀況時可用來作兔^ 心斤1生电力 後可從先前中斷之#門私于电刀口復之 蚵之處開始接續地進行未成 而不必如先前作沬加A 士 里史往斤， 此β點可;/又而從頭開始進行整個的重建程序。 此知·點可於重建程序菸 序毛生-电力中斷狀況之後’更為快速而 18648 14 1263892 =t地元成復原性之重建程序，因此可增進整體之網路 糸、、先g理效能。本發明因此較先前技術具有更佳之進步性 及實用性。 ^以上所述僅為本發明之較佳實施例而已，並非用以限 定本發明之實質技術内容的範圍。本發明之實質技術内容 係廣義㈣義於下述之中請專利範圍中。若任何他人所士 =之技術實體或方法與下述之中請專利範圍所定義者為= 王相同、或是為一種等效之變更 明之申請專利範圍之中。 ”被視為涵盖於本發 【圖式簡單說明】 第1圖為-系統架構示意圖，其中顯示本發明之硬 陣列重建程序中斷接續處理系 ’、 的物件導向元件模型；彳相應用方式及其實體架構 圍马一貢料示意圖 ^^ T翔不本發明之硬碟陳 【主要元件符號說明】 1 〇 電腦平台 J建程序中斷接續處理系統所應用之咖硬碟陣列 ❿中的各個硬碟單元上的—超區塊儲存區。 又 ZU 硬碟陣列裝置（RAID) 21 硬碟單元（主用之硬碟單元） 22 硬碟單元（主用之硬碟單元） 23 硬碟單元（主用之硬碟單元） 24 硬碟單元（主用之硬碟單元） 硬碟單元（備用之硬碟單元） 18648 15 25 1263892 30 硬碟陣列驅動單元 40 超區塊儲存區（superblock) 100 本發明之硬碟陣列重建程序中斷接續處理系統 110 中斷點記錄模組 120 中斷點資料讀取模組 130 重建模組 201 重建程序重啟要求事件For example, the hard disk array is mounted on a hard disk of the other hard disk that is not damaged, and the interrupted condition can be used as a rabbit. After the power is turned off, it can be carried out continuously from the place where the previous interruption was made to the electric knife. It is not necessary to add A to Shishishi as before, and this β point can be; Rebuild the program. This knowledge point can be re-established after the rebuilding of the program's smoke-and-power interruption condition and 18648 14 1263892 = t is a regenerative reconstruction procedure, thus improving the overall network performance and efficiency. The present invention therefore has better advancement and utility than the prior art. The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the technical scope of the present invention. The technical content of the present invention is broadly defined in the following patent scope. If any other person's technical entity or method is the same as the following, the scope of the patent is defined as the same as the king, or an equivalent change to the scope of the patent application. "It is considered to be covered in the present invention [Simplified description of the drawings] Figure 1 is a schematic diagram of the system architecture, which shows the object-oriented component model of the hard-array reconstruction program interrupted processing system of the present invention; The physical structure of the Ma Yi tribute diagram ^ ^ T Xiang not the invention of the hard disk Chen [main symbol description] 1 〇 computer platform J construction program interrupt processing system used in the hard disk array ❿ each hard disk unit Upper - Super Block Storage Area. Also ZU Hard Disk Array Unit (RAID) 21 Hard Disk Unit (Main Hard Disk Unit) 22 Hard Disk Unit (Main Hard Disk Unit) 23 Hard Disk Unit (Mainly Used) Hard disk unit) 24 hard disk unit (master hard disk unit) hard disk unit (spare hard disk unit) 18648 15 25 1263892 30 hard disk array drive unit 40 super block storage area (superblock) 100 hard of the invention Disk array reconstruction program interrupt connection processing system 110 interruption point recording module 120 interruption point data reading module 130 reconstruction module 201 reconstruction program restart request event
16 18648 (§)16 18648 (§)