200917026 九、發明說明： 【發明所屬之技術領域】 本發明係關於一種磁碟管理方法，尤其係關於一種透過位元 映像（bitmap)技術省略 RAID(Redundant Array of Inexpensive Disks ’廉價磁碟冗餘陣列)1設備之初始化同步過程之方法。 【先前技術】 目前，RAID(Redundant Array of Inexpensive Disks ,廉價磁碟 冗餘陣列)設備係透過RAID控制器(可為硬體或軟體)將N台硬碟 結合成虛擬單台大容量硬碟，以提供艾大的儲存容量、更快的存 取速度或資料冗餘功能。依照其資料組織方式之不同，RAK)設備 可以分為Linear(線性模式）、磁碟陣列〇(Rajd_〇)、从仍]、 、RAID-5和RAID-6以及由它200917026 IX. Description of the Invention: [Technical Field] The present invention relates to a disk management method, and more particularly to a method for omitting RAID (Redundant Array of Inexpensive Disks) through a bitmap technique ) 1 method of initializing the synchronization process of the device. [Prior Art] Currently, RAID (Redundant Array of Inexpensive Disks) devices combine N hard disks into virtual single large-capacity hard disks through a RAID controller (which can be hardware or software). Provides Ai's storage capacity, faster access speed or data redundancy. RAK) devices can be classified into Linear (Linear Mode), Disk Array (Rajd_〇), From], , RAID-5, and RAID-6, depending on how they are organized.
RAID-5、RAID-6 等。其中 raid-1、R 們擴展而來的RAID-10、RAID-50等， 200917026 設備100之原始資料硬碟12〇有一個鏡像硬碟14〇，因而原始資料 硬碟120之資料將被複製至鏡像硬碟14〇。讀取資料時，當原始資 料硬碟120出現問題，可以從其鏡像㈣14Q中讀取資料。這種 磁碟陣列可靠性高，但其有效容量減小到總容量之一半以下。 當RAID1設備剛被創建時’需要一個同步過程以初始化原始 資料。 然而習知技術之RAID1設備在創建時進行同步操作之做法存 在以下缺陷： 1、 當RAID1設備初鍵立的輕，絲酬和鏡像磁碟的 内容並不-致’需要透過讀取原始刺的倾，並寫人到其他的 RAH)1成員碟中，才能使RAID1設備之各成員碟之資料保持一 致，因此同步插作時間較長，且初始化同步過程所需之大量讀、 寫操作容易損壞硬碟。 2、 由於RAID1設備初次建立時，沒有外㈣料寫入，而讀 取未寫入的資料並無實際意義。 3、 RAID1設備同步時間較長，且同步操作中斷後必須從 RAID1設備資料區域之起始位置重新開始，這樣不僅浪費了時 間’而且降低了系統的整體效能，也容易對硬碟造成損傷。、 【發明内容】 為了解決上述習知技術中的問題與缺陷，本發明之目的在於 提供一種省略RAID1設備之初始化同步過程之方法，係透過位元 200917026 映像(bitmap)技術省略raID (Redundam八江町〇fRAID-5, RAID-6, etc. Among them, RAID-1, RAID-50, etc. extended by raid-1 and R, 200917026 The original data hard disk 12 of device 100 has a mirrored hard disk 14〇, so the data of the original data hard disk 120 will be copied to Mirror hard disk 14〇. When reading the data, when there is a problem with the original data hard disk 120, the data can be read from its image (4) 14Q. This disk array is highly reliable, but its effective capacity is reduced to less than one-and-a-half of the total capacity. When a RAID1 device has just been created, a synchronization process is required to initialize the original data. However, the conventional technology of RAID1 device synchronization operation has the following defects: 1. When the RAID1 device is initially light, the content of the silk disk and the image disk are not - cause 'need to read the original thorn And write to other RAH) 1 member discs, in order to make the data of each member of the RAID1 device consistent, so the synchronous insertion time is long, and the large number of read and write operations required to initialize the synchronization process are easy to damage the hard dish. 2. Since the RAID1 device was first established, there is no external (four) material to be written, and reading the unwritten data has no practical meaning. 3. The RAID1 device has a long synchronization time, and the synchronous operation must be restarted from the beginning of the RAID1 device data area. This not only wastes time and reduces the overall performance of the system, but also easily damages the hard disk. SUMMARY OF THE INVENTION In order to solve the problems and deficiencies in the above-mentioned prior art, an object of the present invention is to provide a method for omitting the initialization synchronization process of a RAID 1 device by omitting raID through the bit matrix technology 200917026 (Redundam) 〇〇f
Disks，廉價磁碟冗餘陣列）i設備之初始化同步過程。 本發明所提供之-種省略RAID1設備之初始化同步過程之方 法，係包含以下步驟： 創建RAID1設備，並於RAID1設備之複數個成員碟上分別 劃分出-塊大小侧的空間，用⑽存與上私成員碟上的複數 個資料區塊相對應之位元映像(bitmap);初始化上述各成員碟上的 位元映像之齡賴，簡此位元輯之值全部清零；啟動⑽加 設備，並將上述位元映像加載到内部記憶體或緩衝區記憶體中； 當對RAID1設備之資料區塊進行讀取操作時，讀取上述位元映像 中與資料區麟應之位元值’當位元映像中對應之位元值為〇時， 讀取操作依照讀取之資料均為〇處理，當位元映像中對應之位元 值為1時’舰RAID1設備之讀取操作流程進行處理；以及當對 RAID1設備之資料區塊進行寫人操作時，讀取位元映像中與資料 區塊對應之位元值，當位元映像帽應之位秘為q時，更新位 兀映像對應之位元值為1 ’並將位元映像資訊寫入狀仙設備中 的位元映像之儲存空財，且於位㈣像全部錢完畢後，從内 部記憶體或緩衝區記憶體中卸載上述位元映像，並在祕中標識 位元映像不存在’當位元映像巾對應之位元值為I _，則依照 RAID1設備之寫入操作流程進行處理。 200917026 其中，本發明之-種省略狀⑼設備之初始化同步過程之方 法，進一步包含以下步驟： 當對RAID1設備之資料區塊進行讀取_人㈣時，於讀取 位元映像中與資料區塊對應之位元值之前，首先判斷位元映像是 否存在’如果存在’則讀取位元映像中與資料區塊對應之位元值， 如果不存在備照RAID1設備之讀取或寫人操作流程進行處理。 此外，上述位TL映像之每一位元分別對應於以肋設備之每 -資料區塊，且位元映細二進制位元Q表示所對應之資料區塊 未被寫入過=諸，以二_位元丨表示所對應之資舰塊已被寫 入過資料。另外’上述位元映像之儲存空間的大小係透過以下公 式計算得到：位元映像之儲射間的大小+⑽叫設備之資料空 間的大小=卜(RAID 1設備之資料區塊之位元組數目χ8)。 綜上所述’本發明之優點在於： 本發明所提供之-種省略RAID1設備之初始化同步過程之方 法，係透過位元映像技術省略了 RAID1設備初次建立時之初始化 同步過程，因而使RAID1設備從創建開始即具有正常之效能，同 時省略了初始化同步過程所需之大量讀、寫操作及其所耗費的時 間，進一步提高了 RAID1設備之整體效能，保護了設備 中之硬碟。 【實施方式】 200917026 有關本發明之特徵與實作，茲配合圖示作較佳實施例詳細說 明如下。 請麥考「第2圖」，圖中表示了本發明之設備之資料 儲存結構之方塊圖，其中以3塊硬碟組成ra辽)！設備1〇〇來具體 說明，如「第2圖」所示，硬碟】、硬碟2、硬碟3分別由多個資 料區塊200、一個位元映像21〇及一個超級區塊22〇組成，其中 RAID1設備1〇〇之各資料區塊2〇〇之位元組（Byte)數目為4K， 且位元映像210在RAID 1設備之上述各成員碟上的儲存空間係位 於緊鄰超級區塊（Super Block) 220之前的位置，其中位元映像 210之儲存空間的大小透過以下公式計算得到：位元映像之儲存 空間的大小+RAID1設備之資料空間的大設備之資 料區塊之位元組數目χ8)(公式一）。假設RAID1設備之資料空間容 夏為1Τ(即’ 1〇24>α〇24Μ) ’每一個資料區塊之大小為4Κ ,則位 元映像210之儲存空間的大小為：1〇24χΐ〇24—(1〇24χ4χ8) = 32Μ。並且查詢位元映像之過程相對尺八仍丨中每個讀寫請求的轉 化要簡單的多，因此分配位元映像所導致的空間和時間上的代價 完全可以承受。 現在請參考「第3圖」’此圖為本發明之一種省略設 備之初始化同步過程之方法的方法流程圖，如圖所示，本發明之 一種省略RAID1設備之初始化同步過程之方法包含以下步驟： 12 200917026 創建RAID1設備(步驟300);在RAID1設備之每個成員碟上 分別劃分出一塊大小相同的空間，用以儲存與上述各成員碟上的 各個資料區塊相對應之位元映像(步驟301);初始化各成員碟上位 元映像之儲存空間，以將位元映像之值全部清零(步驟3犯）.啟動 RAID1設備，並將上述位元映像加載到内部記憶體或緩衝區記憶 體中(步驟303);當對raID1設備之資料區塊進行寫入操作時读 驟304);讀取位元映像中與上述資料區塊對應之位域(步驟 _);查詢位元映像中對應之位元值是否為Ο(步驟);如果 不是為〇，依照RAim設備之寫入操作流程進行處理(步驟3〇6); 如果位元映像中對應之位元值為G，更新此位元映料應之位元值 為！，並將位元映像資訊寫入議！設備中各成員碟相應的位元 映像之儲存空附(频则）；_位元映像是否全部更新完畢 (步驟3044) ’如果沒有全部更新，則繼續執行步驟氣·如果全部 則從内部記憶體或緩_雜體中卸载上述位元映 亚知、統中標識位元映像不存在(步驟);當對論^設 備之貧料區塊進行讀取操作時. 又 ^ 〇5)，_位元映像中對應之 位讀疋㈣物細）；如物，讀 均為〇處理(步驟3052)，而後繼 貝取之貝科 依昭RAm] 1 . 、執订乂驟30);如果不為0，則 '、、、1政備之練操作流程進行處理（步驟3⑹。 200917026 「―「弟4圖」為「第3圖」所示之方法中的—步驟分解流程圖。 如。「第4圖」所示，本發明之一種省略驗加設備之初始化同步 過耘之方法，可進一步包含以下步驟： 當將要對RAID1設備之資料區塊進行讀取或寫入操作時（步 驟二〇),’於判斷或讀取位元映像中與上述資料區塊對應之位元值 之刖’百先_位it映像是否存在（步驟伽），如果存在，則繼 續判斷或讀取侃映像中與上述資料區塊對應之位元值（步驟 )如果不存在，則依照設備之讀取或寫入操作流程進 行處理（步驟440)。 第5圖」為本發明中位元映像與資料區塊之映射關係圖， 如圖所不，位7C映像中分別對應於各個資料區塊之每一位元挪 之值係以一進制數字！或〇表示，其中當位元映像之位元㈣之 值為1時’則表示與其對應的資料區塊遞在讀、寫操作之前已 經被寫入過資料，而當位元映像之位元，之值為〇日夺，則表示 與其對應的資料區塊200a之前從未被寫入過資料。 現在將依據本發明中位元映像與資料區塊之映射關係以及本 發明之RAID1没備之資料儲存結構對本發明之省略狀腿設備 之初始化同步輕之方法進行詳細描述。其中，「第6圖」為本發 明中初始化位it映像之儲存空間的示意圖，如圖所示，當初次建 立議1設備並在設傷之每個成員碟上分別劃分出-塊 大小相同的用U儲存與各成員碟上的各個資料區塊相對應之位元 14 200917026 映像的儲存雜之後，對各成員碟上儲存位元映像2iq的空間進 行初始化，以使位元映像之位元52〇之值全部為〇。「第7圖」為 本發明中將位元映像加載到㈣記憶體或緩衝區記憶體之示音 圖，如圖所示，當位元映像之全部位元52〇之值清零後^位^ 映像加載到内部記憶體或緩衝區記憶體72〇中。「第8圖」為本發 明中對RAID1設備進行寫入操作之示意圖，如圖所示，當對⑽以 設備之資料區塊進行寫人操作時，首練據内部記憶體或缓衝區 記憶體720中位元映像之位元52()之值進行相應的操作（請參考 上述的方法之步驟304、步驟3041、步驟3042)，如果要寫入的資 料區塊在内部記憶體中或緩衝區記憶體72〇中對應的位元映像之 位元值為1時，代表此資料區塊為寫操作之前已經寫人過資料的 資料區塊·b ’於是触以职設備之原有的以操作流程進行 處理(請*考上述步驟306) ’而後，之前已經寫人過資料的資料區 塊200b將變為寫操作完成之後的資料區塊2〇〇c;如果位元映像對 應〜位it值為〇 ’則此資料區塊係為未被寫人過資料的資料區塊 200a’此時將更新位元映像龍之位元值為丨，然後再將位元映像 資訊寫入RAID1設備中的位元映像之儲存空間中(請參考上述步 驟3043)，由此未被寫入過資料的資料區塊2〇〇a也變為寫操作完 成之後的資料區塊2〇〇c。1第9圖」為本發明中對raidi設備進 行寫入操作後的位元映像和資料區塊之狀態示意圖，如圖所示， 當對RAID1設備之寫操作處理完成後，在内部記憶體或緩衝區記 200917026 憶體料的位元映像之位元520之们分別對應資料區塊中寫 料區塊2GGc及寫操作之前已被寫人過的資料區 塊雇’而位元映像之位元520之值0則對應未被寫入過 區=a。「第10圖」為本發明中對咖設備進行讀取操作之 圖所示’進行讀操作處辦，錢要查師判斷位元 3〇5中牛對應之位元520之值是否為0 (請參考上述的方法之步驟 3,如果為。，則證明此位元映像之位元所對應！ 塊係為一作之所未被寫人過的資料區塊職，讀取剩乍 ::取:資卿處理(請參考上述―，即按照圖中 庫之資料=處理，如果為1 ’則證明此位元映像之位元所對 ^、…係為％操作之前已被寫入過的資料區塊2_，讀取 :、=照_設備之原有的讀取操作流程 參Disks, a redundant array of inexpensive disks) The initialization synchronization process of i devices. The method for omitting the initialization synchronization process of the RAID1 device includes the following steps: Creating a RAID1 device, and separately dividing the space on the block size side of the plurality of member disks of the RAID1 device, and using (10) The bit map corresponding to the plurality of data blocks on the private member disc; initializing the age of the bit map on each member disc, and the value of the bit set is all cleared; starting (10) adding the device And loading the above-mentioned bit map into the internal memory or the buffer memory; when reading the data block of the RAID1 device, reading the bit value in the bit map and the data area When the corresponding bit value in the bit map is 〇, the read operation is processed according to the read data. When the corresponding bit value in the bit map is 1, the read operation flow of the ship RAID1 device is performed. Processing; and when performing a write operation on the data block of the RAID1 device, reading the bit value corresponding to the data block in the bit map, and updating the bit map when the bit image cap is the secret block q Corresponding bit value 1 'writes the bit map information into the bit map of the device, and saves the bit map from the internal memory or the buffer memory after the bit (4) is completed. In the secret, the identifier bit map does not exist. 'When the bit value corresponding to the bit map towel is I _ , the processing is performed according to the write operation flow of the RAID 1 device. 200917026 The method for initializing the synchronization process of the device (9) of the present invention further includes the following steps: when reading the data block of the RAID1 device, the data block is read in the bitmap image. Before the bit value corresponding to the block, first determine whether the bit image exists 'if present', then read the bit value corresponding to the data block in the bit map, if there is no read or write operation of the standby RAID1 device The process is processed. In addition, each bit of the bit TL image corresponds to each data block of the rib device, and the bit map binary bit Q indicates that the corresponding data block has not been written. The _bit 丨 indicates that the corresponding asset block has been written. In addition, the size of the storage space of the above-mentioned bit map is calculated by the following formula: the size of the storage space of the bit map + (10) the size of the data space of the device = Bu (the byte of the data block of the RAID 1 device) The number is χ8). In summary, the present invention has the advantages that: the method for omitting the initialization synchronization process of the RAID 1 device is omitted by the bit mapping technology, and the initialization synchronization process when the RAID 1 device is first established is omitted, thereby making the RAID 1 device It has normal performance since its creation, and omits the large number of read and write operations required to initialize the synchronization process and the time it takes, further improving the overall performance of the RAID1 device and protecting the hard disk in the device. [Embodiment] 200917026 The features and implementations of the present invention are described in detail with reference to the accompanying drawings. Please refer to the "Figure 2" of Mai Khao, which shows the block diagram of the data storage structure of the device of the present invention, which consists of 3 hard disks: The device 1 is specifically described. As shown in FIG. 2, the hard disk, the hard disk 2, and the hard disk 3 are respectively composed of a plurality of data blocks 200, a bit map 21, and a super block 22. The composition, wherein the number of bytes of each data block 2 of the RAID 1 device is 4K, and the storage space of the bit map 210 on the member disks of the RAID 1 device is located in the immediate vicinity of the super zone. The position before the Super Block 220, wherein the size of the storage space of the bit map 210 is calculated by the following formula: the size of the storage space of the bit map + the bit of the data block of the large device of the data space of the RAID 1 device The number of groups is χ8) (Formula 1). Assume that the data space of the RAID1 device is 1Τ (that is, '1〇24>α〇24Μ). 'The size of each data block is 4Κ, then the size of the storage space of the bit map 210 is: 1〇24χΐ〇24— (1〇24χ4χ8) = 32Μ. And the process of querying the bit map is much simpler than the size of each read/write request, so the space and time cost of allocating the bit map is completely affordable. Referring now to FIG. 3, this figure is a flowchart of a method for omitting a method for initializing a synchronization process of a device. As shown in the figure, a method for omitting an initialization synchronization process of a RAID 1 device includes the following steps. : 12 200917026 Create a RAID1 device (step 300); each of the member disks of the RAID 1 device is divided into a space of the same size for storing a bitmap corresponding to each data block on each member disk ( Step 301): Initialize the storage space of the bit map on each member disc to clear all the values of the bit map (step 3). Start the RAID1 device and load the above bitmap image into internal memory or buffer memory. In the body (step 303); when performing a write operation on the data block of the raID1 device, step 304); reading the bit field corresponding to the data block in the bit map (step _); querying the bit map Whether the corresponding bit value is Ο (step); if not, processing according to the RAIM device write operation flow (step 3〇6); if the corresponding bit value in the bit map is G, update this bit yuan Bit value of the material to be! And write the bitmap information to the discussion! The storage of the corresponding bit map of each member disc in the device is attached (frequency); whether the _bit map has been completely updated (step 3044) 'If not all the updates are continued, the step gas is continued. If all are from the internal memory Or unloading the above-mentioned bit meta-map, the meta-image of the system does not exist (step); when reading the poor block of the device ^. ^ 5), _ bit The corresponding position in the meta-image reads (4) the material is fine; if it is, the reading is all 〇 processing (step 3052), and the subsequent 取 取 贝 贝 RA RA RA RA RA RA RA RA RA 、 、 、 、 、 、 、 、 、 ; ; ; ; ; ; ; ; ; ; ; ; ; ; Then, the processing flow of the ',,, and 1 political preparations is processed (step 3 (6). 200917026 "" "Dia 4" is the step-by-step decomposition flowchart in the method shown in "3rd figure". For example, "4th. As shown in the figure, the method for omitting the initialization synchronization of the verification device of the present invention may further include the following steps: when the data block of the RAID 1 device is to be read or written (step 2), 'To determine or read the bit value corresponding to the above data block in the bit map刖 'Hundreds of _ bit it image exists (step gamma), if it exists, continue to judge or read the bit value corresponding to the above data block in the 侃 image (step), if it does not exist, read according to the device Or write to the operation flow for processing (step 440). Figure 5 is a mapping diagram of the bit map and the data block in the present invention, as shown in the figure, the bit 7C image corresponds to each of the data blocks. The value of a meta-forward is represented by a single digit ! or 〇, where when the bit (4) of the bit map is 1, it means that the corresponding data block has been written before the read and write operations. After the data, when the bit of the bit map is the value of the day, it means that the data block 200a corresponding thereto has never been written before. Now the bitmap and the data block according to the present invention will be used. The mapping relationship and the data storage structure of the RAID 1 of the present invention are described in detail for the method of initializing and synchronizing the ellipsis leg device of the present invention. The "figure 6" is the storage space for initializing the bit image of the present invention. Intention, as shown in the figure, when the device 1 is established for the first time and the member blocks of the same size are respectively divided into U-storage bits corresponding to the respective data blocks on each member disc. 200917026 After the image is stored, the space of the bit map 2iq stored on each member disk is initialized so that the value of the bit map 52〇 of the bit map is all 〇. "FIG. 7" is a bit in the present invention. The image is loaded into the (4) memory or buffer memory, as shown in the figure, when the value of all the bits of the bit map is cleared, the image is loaded into the internal memory or the buffer memory. 72〇. "Figure 8" is a schematic diagram of the write operation of the RAID1 device in the present invention. As shown in the figure, when the (10) write operation is performed on the data block of the device, the first exercise is based on internal memory or The value of the bit 52 () of the bit map in the buffer memory 720 is correspondingly operated (refer to step 304, step 3041, step 3042 of the above method), if the data block to be written is internally stored The corresponding bit in the volume or buffer memory 72〇 When the bit value of the image is 1, it means that the data block is the data block that has written the data before the write operation. b 'The original operation of the device is processed by the operation process (please * test the above steps) 306) 'And then, the data block 200b that has previously written the data will become the data block 2〇〇c after the write operation is completed; if the bit map corresponds to the ~ bit it value is 〇' then the data block system The data block 200a' for the unwritten data area will update the bit map dragon bit value 丨, and then write the bit map information into the storage space of the bit map in the RAID1 device (please Referring to the above step 3043), the data block 2〇〇a from which the data has not been written also becomes the data block 2〇〇c after the completion of the write operation. 1 ninth diagram is a schematic diagram of the state of the bit map and the data block after the write operation of the raid device in the present invention, as shown in the figure, after the write operation of the RAID 1 device is completed, in the internal memory or Buffer record 200917026 Recalling the bit map of the material bit 520, respectively, corresponding to the data block 2GGc in the data block and the data block that has been written before the write operation, and the bit of the bit map A value of 520 of 0 corresponds to the unwritten area = a. "Fig. 10" is a read operation operation for the reading operation of the coffee machine in the present invention. The money is checked by the inspector to determine whether the value of the bit 520 corresponding to the cow in the bit 3〇5 is 0 ( Please refer to step 3 of the above method. If it is, it proves that the bit of this bit map corresponds to! The block is a data block that has not been written by the user. Read the remaining:: Take: Capital processing (please refer to the above -, that is, according to the data in the library in the figure = processing, if it is 1 ', it proves that the bit of the bit map is ^, ... is the data area that has been written before the % operation Block 2_, read:, = photo_the original read operation flow of the device
輪的〃_麵㈣料J ㈣像之^侧她軸雜細鐘中卸載 元、2不忍圖，如圖所示’當位元映像全部更新完畢後，位 内h卜各個娜塊相對應之位元52G之值全部為1，於是從 =體，記憶體72。中卸載上述位元映像，並在系統 L錢像不存在(請參考上述步驟3045)。 限定發明以前述之較佳實财式揭露如上，然其並非用以 之申。本領域之技術Μ應當意識到在微離本發明所附 專利範圍所揭示之本發明之範圍和精神之情況下，所為之 16 200917026 =動與my均屬本發明之專利保護範圍之内。關於本發明所界 定之保護範圍請參考所附之申請專利範圍。 【圖式簡單說明】 第1圖為習知技術之RAID!設備之資料儲存結構之方塊圖； 第2圖為本發明之以肋設備之資料儲存結構之方塊圖； 第3圖為本發明之一種省略以肋設備之初始化同步過程之 方法的方法流程圖； 第4圖為「第3圖」所示之方法中的—步驟分解流程圖； 第5圖為本發明中位元映像與資料區塊之映射關係圖； 第6圖為本發明中初始化位元映像之儲存空間的 示意 第 圖為本發日种將位元映像加載_部記憶體或 憶體之示意圖； ° 第8圖為本發明中對RAID1設備進行寫入操作之示音圖. / 9圖為本發财對狀皿設備進行寫人操作後驗元 和資料區塊之狀態示意圖； 第10圖為本發明中對RALD1設備進行讀取操作之示音 以及 w ° ’ 第11圖為本發明中從内部記憶體或緩衝區記憶體中卸載位元 映像之示意圖。 凡 【主要元件符號說明】 100 RAID 1 設備 200917026 120 原始資料硬碟 140 鏡像硬碟 200 資料區塊 210 後元映像 220 超級區塊 ·_ 2〇〇a 未寫入之資料區塊 2⑻b 已寫入之資料區塊 200c 寫操作完後之資料區塊 520 位元 720 内部記憶體或緩衝區記憶體 步驟300 創建RAID1 步驟301 劃分RAID1，以在RAID1每個成員碟上分配出 相同大小的空間存放位元映像 步驟302 初始化各成員碟上的位元映像儲存空間，以將位 元映像之值全部清零 步％ 3〇3啟動以皿並將位元映像加載到内部記憶體或緩 衝區記憶體中 步驟304 寫操作 步驟3 041讀取位元映像中與資料區塊對應之位元值 ..步驟3042查詢位元映像對應之值是否為〇 200917026 並將位元映像資訊寫 入各 步驟3043更新位元映像值為j， 成員碟相應之位元映像中 步驟3044位元映像是否全部更新 步驟3045從内部記憶體或緩衝區記 並在系統中標識位元映像不存在 憶體中卸載位元映像， 作 步驟305 讀操作 步驟3051 判斷位元映像對應之值是否為〇 步驟3052 讀取操作依照讀取之資料均為〇處理 步驟306 按RAID1 備之§胃取或寫人操作流程進行處理 步驟410 將要對RAID1設備之聽區塊進行讀取或寫入操 步驟420 判斷位元映像是否存在 步驟430 判斷或讀取位元映像中與資料區塊對應之位元值 步驟440 依照RAID1設備之讀取或寫入操作流程進行處理 19The 〃_面(4) material of the wheel J (4) is like the side of the side of the shaft, the unloading element in the miscellaneous clock, 2 can not bear the picture, as shown in the figure, 'When all the bitmaps are updated, the corresponding blocks in the position are corresponding to each other. The value of bit 52G is all 1, so from = body, memory 72. Unload the above bitmap image and the image in the system L does not exist (please refer to step 3045 above). The invention is disclosed above in the preferred form of the foregoing, and is not intended to be used. It is to be understood that the scope of the invention and the scope of the invention as disclosed in the appended claims are intended to be within the scope of the invention. Please refer to the attached patent application for the scope of protection defined by the present invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a data storage structure of a rake device of the prior art; FIG. 2 is a block diagram of a data storage structure of a rib device of the present invention; A flow chart of a method for omitting a method of initializing a synchronization process of a rib device; FIG. 4 is a flow chart of step-by-step decomposition in the method shown in FIG. 3; FIG. 5 is a bitmap image and data area of the present invention. Block diagram of the block; Figure 6 is a schematic diagram of the storage space of the initialized bit map in the present invention. FIG. 8 is a schematic diagram of loading a bit map into a memory or a memory; In the invention, a voice diagram of a write operation to a RAID1 device is provided. FIG. 9 is a schematic diagram showing the state of the element and the data block after the writing operation of the fortune device; FIG. 10 is a view of the RALD1 device in the present invention. The sound of the read operation and w ° ' Fig. 11 is a schematic diagram of the unloading of the bit map from the internal memory or the buffer memory in the present invention. [Main component symbol description] 100 RAID 1 device 200917026 120 Raw data hard disk 140 Mirror hard disk 200 Data block 210 Post-image 220 Super block _ 2〇〇a Unwritten data block 2 (8) b Written Data block 200c Data block after writing operation 520 bit 720 Internal memory or buffer memory Step 300 Create RAID1 Step 301 Divide RAID1 to allocate the same size space storage space on each member disk of RAID1 The meta-image step 302 initializes the bit map storage space on each member disc to clear all the values of the bit map, and the bit map is loaded into the internal memory or the buffer memory. Step 304: Write operation step 3 041 reads the bit value corresponding to the data block in the bit map. Step 3042 queries whether the value corresponding to the bit map is 〇200917026 and writes the bit map information to each step 3043 update bit. The meta-image value is j, and the step 3044 bit map in the corresponding bit map of the member disc is all updated. Step 3045 records from the internal memory or buffer and identifies the bit in the system. The image does not exist in the memory image of the unloading bit map. In step 305, the read operation step 3051 determines whether the value corresponding to the bit map is 〇. Step 3052. The read operation is performed according to the read data. Step 306 According to RAID1 The process of taking or writing a human operation process is performed. Step 410: Read or write to the listening block of the RAID 1 device. Step 420: Determine whether the bit map exists. Step 430: Determine or read the bit corresponding to the data block in the bit map. The value step 440 is processed in accordance with the read or write operation flow of the RAID 1 device.