1293428 九、發明說明： 【發明所屬之技術領域】 本發明係關於-種獨立磁碟多重陣列⑽_映射儲存設 備，特別是一種映射儲存設備資料保護的方法。 又 【先前技術】 獨立_多重_⑽_目的在於將相_資料儲存 =碟機中不同的地方，並且改善儲存子祕的效能，其中映射 如：剛丨）使瓶少兩__磁碟機並且在兩者 ==目同的資料區塊，此為_中錯誤容忍度最慢的類 口為同日袖須將資料複翻兩個磁碟 向可靠度最簡單的方法。 而此為&供 映㈣__的方法是雙控制器配合 、雖存叹備的方式，請參考第！圖所示 儲存設備被一個控制器！所控制，如果該^猎由映射 使用（如電源損壞，機器當機）等情況 ^財2無法 給另一個控制哭9 丨. 才悮映射儲存設備則交 爲停止L法^ 考第2 ®所示’映射儲存設備會變 向該映射鱗，綠用者正在 控制器2完全控制=的貝科存放在一個暫存器10中，待另-個 者資料複製到映射館存==又備之後，再將暫存器W中的使用 因此，習知儲存方式的缺點是需要—暫存器H)，而且需要控 12934281293428 IX. Description of the Invention: [Technical Field] The present invention relates to an independent disk multiple array (10)_mapping storage device, and more particularly to a method for mapping data protection of a storage device. Also [previous technology] independent _ multiple _ (10) _ purpose is to save the phase _ data = different places in the disc player, and improve the performance of the storage sub-function, which maps such as: just 丨) make the bottle less two __ drive And in the data block where the two == the same, this is the slowest type of error tolerance in _ is the easiest way to reverse the data to the same time. And this is the method of & (4) __ is the dual controller to cooperate, although the way to sigh, please refer to the first! The figure shows that the storage device is a controller! Controlled, if the ^ hunting is used by the mapping (such as power supply damage, machine crash), etc. ^ 2 can not give another control to cry 9 . Only map storage device is turned into stop L method ^ test 2 ® The 'map storage device will change to the mapping scale, the green user is in the controller 2 full control = Beko is stored in a register 10, after the other data is copied to the mapping library == And then use the scratchpad W. Therefore, the shortcoming of the conventional storage method is that it needs to be - the scratchpad H), and it needs to control 1293428
制器的轉換過程，比較I 的時候，無法正常駐；：==統的性能。當控制器出現問題 都曝露了使用”料==過程與額外的暫存器 【發明内容】 本發細—種粒映射齡設備外， ΐ達一=新的方法’以針對使用者在使用儲』 、〜、。又的目的’該方法包括下列步驟：分配每一控 :制相對應的唯—儲存設備；再將複數_存設備建立 存設備;映射儲存設備將其狀態訊息反饋至每二 、軍妆田有=w控制器發生時，確定至少—個控制器正常 貝1持、’男接收使用者的資料，並記錄至映射儲存設備中； 約I啓，常控_ ’並獲取映射儲存設備的狀態訊息；最後 船狀㈣息找到資料最完整的儲存設備’將異常控制器所控 7目對應的唯-儲存設備進行同步更新並添加到映射儲存設 備0 以示^^關於本糾瓶町之實财紅說明係用 進一步^解=本|明之原理，並且提供本發明之專利巾請範圍更 【實施方式】 明如=關本發_特徵與實作，賊合®式作最佳實施例詳細說 1293428 (控制器〗控制鱗個控制器只控制自已的儲存設備 控_^ =自2 _存設借2、控制器3 映射儲存，備，丄1T 器的健存設備建立-個整體的 饋至每tir Γ輕映贿存賴_態訊息反 無法使用^====，，料―個控制器 器可以有多個，祕古I 傷的讀寫，這樣的控制 备 有㈣存設備生成—倾射儲存設備。 資气，、^都㈣自己儲存設備，並且知道整體映射設備的 =吏用者資料可以向這個映射設備進行複製。 異常情況導為例)發生 的原理，儲存便用根據上迷建立映射儲存設備 ，的時候就統-了 _備的大 %什麼都沒有改變。 "兄對使用者來 的目的，請參考第4圖= 備時，達物保護 制器相對應的唯； 存設備建立-映射触W4 ，再將魏個該儲 射館存設_狀”自==驟4G2);触射鱗設備將該映 任—異常控制器步驟侧）'·當有 生守確疋至少一個控制器正常運作，則持續 1293428 録及根據該狀態訊息備的狀態訊息（步驟 映射儲存設備（步驟侧）。&備進洲步更新並添加到該 睛參考第5圖所干，從& # 個儲存設備上的數據資_ θ 到母一個儲存設備上，每一 儲存設備的大小（步義為—個_上 __來映_存設備的t 的儲存空間將不會被使用，這樣伴：了每個=於麵存設備 者資枓&心 、樣保祖了母一個儲存設備上的使用 致性’接著判斷建立是否成 母個控制器都得到目前映射 1⑷如果疋’ 控制器上都是相同的，A :貞°，逆個魏在每一個 每-控制哭之斬⑽/、 叙備的所有的資訊係記錄在 立即的_給每-=制ί(用牛者驟資料佔用映射儲存設備的訊息會 儲存=備，則報告失敗和原因（步驟_。了映射 η申1參f第6圖所示，係為同步錯存設備資料更新的流程，牟 被修復之後，需要_是將控繼1啓動， -1獲取映射儲存設備的狀態訊息（步驟_後，檢查該= 1293428 器1資料是扣整(步驟6G2)， 到資料最完整的儲存設_f_二:該狀態訊息找 所控制之儲存設傷1的 “（4 將控制器1 儲存設備1添加_輸:（==糊）’域再將 復到原來的狀態，作々 5)，攻樣系統可以恢 、言料/ 者來説不會有任何的影響。 在-i正方ί可以延伸到多個控㈣的問題發生，只要存 巾’系統就可以正常 器修復完畢之後可以隨時添加到映射:=問靖的控制 達到各個控制器所控制的儲存 =中’通過同步操作 分方便，咖者輸_=^同，_護系統十 備資====== ::=Γ板，兩個磁碟設備，每個主機板二::： 2= 資訊。將所有的磁碟設備建立成—個映射儲 :備’ _齡設_資峨寫人聽置财，制者資料可 以通過網卡寫_映射儲存設備上。 碟報供果魅機板發生了問題無法正常的卫作，它所管理的磁 r會出現問題，這個時候映射儲存設備變爲降階狀態，正 吊主機板的配置標會發生變化，記錄下另一個主機板損壞的情 映射儲存设備仍舊可以正常的使用，使用者的資料被記錄在 1293428 那個=有問題的磁碟設備上，對使用者來説什麼都沒有發生。 板合板恢復正常工作之後，再次㈣社機板，該主機 做二二個正常主_配置槽’然後對映射儲存設備 置卜將塊磁碟上的資料相同，然後修改兩個配 變化。 仃貝科傳輪，對使用者來説沒有 不其他⑽存設财料絲的賴也是相同的， 備^=備的同步過程’因爲映射健存設備自身有比較完 來實現映射機制，在資料備不依减射儲存設備 需要依靠自編侧^執:復之後朗步過程 行，對使用者的讀寫沒有辟第m作1在後端進 二‘接著將資料寫人每—個狀態正常控制 驟：否顺告失敗和原寫入完畢(步 器的二=Γ通：：=_的過㈣看控制 -’通過該控制―備:::==: 沒~置(步驟•這樣可_資1=:: 1293428 月〕=制為儲存設備沒有的資料寫入(步驟_，接著判斷資料 =入=成功（步驟9()5)，如果成功，在寫人資料的結尾做標示， 9(m 4料寫人完畢（步驟9G6) ’否則報告失敗和原因（步驟 士。配置槽之標示資料區塊的資料結構包括儲存設備Id、建立 日寸間、寫人資料的大小、所屬控制器的Id以及其他信息等等。 本發明所提之方法具有以下優點： h沒有轉換控制器的過程，大大提高了系統的可靠性，和安全 性，系統不會在控制器轉換失敗的時候而癱瘓。 2.不使用暫存H ’提高了減的性能和使用者資料的安全性， 即使在單個控制ϋ發生損毀的時候，制者的㈣依舊可以 向這個映射儲存設備進行複製，沒有了中間的緩衝減少了記 憶體的消耗。 3·使得整體系統轉起來比較枝，當控繼發生問題的時 候’不树認制者資料贿細難絲，就可 對出現問題的控制器進行維修。 軸本發日肋前述之較佳實補揭露如上，然其並非用以限 在不脫離本發明之精神和範圍内，所為之更動與潤飾， =杨明之專利保護範圍。本發明所界定之保護範 考所附之申請專利範圍。 / 【圖式簡要說明】 第1圖係為習知雙控制器配合映射儲存設備的示意圖； 第2圖係為習知雙控制器配合映射儲存設備的控制器的轉換 11 1293428 過程示意圖； 第3Β圖係為本發明之控制器失效的示意圖； 、 第4圖係為本發明之方法流程圖； 帛5 _為本發财立映射贿設備的流程圖； • L6:為本發明之同步儲存設備資料更新的流程圖； 圖係為本發明之實施例示意圖； 第8圖係為本發明之實施之哭次 第9圖係為本發明之實施例之^‘恢復=的流程圖；及 【圖式符號說明】 _資料的流程圖。 10...................暫存器 存::4°1分配每-控制器只控制相對應的唯—儲 # 備步驟4°2將複數個該儲存設備建立一映射儲存設 態==::r該映射儲存設備的狀 二=.1...#有任—織繼_，確定至少- 至外Γ w作’則持續接收使用者的資料，並記錚 至4映射儲存設備中 σ己錄 步驟405·.··...·重新啓動該異常控制器，並獲取該映射儲 12 存設備的狀態訊息 存設備:㈣異息找_ 的唯一 步驟5。;:1:;個控制器上物備的大小 步驟5〇3 —的錯存設请 --I：，；：：：：：：- 立即的映射給每 佔用映射物備的訊息會 步:5°6........報告失敗二 IIZ控制器1獲取映射儲存設備的狀態訊息 取········檢查該控制器1資料是否完整 4Γ03Π根據錄態訊息找物最完整的儲存 a 又備進仃同步操作 04········將控制器1所控制之儲存設備1的資料進 行補充完整 v秣605........將儲存設備1添加到映射儲存設備 步驟801··..··..查看控制器的狀態 步驟802 ········獲取網絡層傳輸的資料 v私803........將貧料寫入每一個狀態正常控制器的儲 存設備 $ 步驟 804·..· ....判斷資料寫入是否成功 步驟 805···· 入完畢 ....在寫入資料的結尾做標示，證明該資料寫 步骤806 ·… ....報告失敗和原因 步驟 901··.. ....查看控制器的狀態 步驟 902···. ....通過配置檔找到沒有損壞的控制器 1293428 步驟903 ··.·····通過該控制器上儲存設備上的標示資料 找到當前控制器上沒有的資料位置 步驟904 ····.···將目前控制器儲存設備沒有的資料寫入 步驟 905···· ....判斷資料寫入是否成功 步驟906 ·… 入完畢 ....在寫入資料的結尾做標示，證明該資料寫 步驟907.… ....報告失敗和原因The conversion process of the controller, when comparing I, can not be normal; :== system performance. When the controller has problems, it is exposed to use "material == process and additional register" [invention content] This hair is fine-grain mapping age equipment, ΐ达一 = new method 'to use the storage for the user 』, 〜, 。的目的' The method includes the following steps: assigning each control: the corresponding only storage device; then storing the device as a storage device; When the military makeup field has a =w controller, it is determined that at least one controller is normally held, and the information of the male receiving user is recorded in the mapping storage device; about I start, always control _ 'and obtain the mapping The status information of the storage device; finally, the ship-like (four) information finds the most complete storage device of the data. The synchronous storage device corresponding to the 7-mesh device controlled by the abnormal controller is synchronously updated and added to the mapping storage device 0 to show The real money of the bottle is explained by the principle of further solution = the principle of this method, and the scope of the patented towel of the present invention is provided. [Embodiment] Mingru = Guanbenfa _ characteristics and implementation, the best implementation of the thief Detailed example 1293428 (Controller) control scale controller only controls its own storage device control _^ = from 2 _ storage 2, controller 3 mapping storage, backup, 丄 1T device storage device establishment - an overall feed to Every tir Γ 映 贿 贿 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Pouring storage equipment. 资气,、^ (4) own storage equipment, and know that the overall mapping equipment = user data can be copied to this mapping device. The abnormal situation is taken as an example), the principle of storage, based on the storage When the fan establishes the mapping storage device, it will be unified - the large % of the preparation has not changed. " Brother for the purpose of the user, please refer to Figure 4 = When the standby device, the corresponding protection device only ; storage device establishment - mapping touch W4, then Wei will store the storage hall _ shape" from == step 4G2); touch the scale device will be mapped - abnormal controller step side) '· when there is a living If at least one controller is working properly, it will continue to record 1293428 according to the status message. The status message of the backup (step mapping storage device (step side). & backup into the step update and add to the eye reference to Figure 5, from the &# storage device data _ θ to the mother one storage On the device, the size of each storage device (step is - _ _ _ _ _ _ storage device t storage space will not be used, so with: each = in the face of the equipment owner 枓 & The heart, the sample protects the mother's use of a storage device, and then determines whether the establishment of the parent controller is the current mapping 1 (4) if the controller is the same, A: 贞 °, reverse Wei In each of the per-control crying (10) /, all the information is recorded in the immediate _ give each -= system ί (the message is stored in the map storage device will be stored = backup, then the report fails And the reason (step _. The map η申一参f is shown in Figure 6, which is the process of updating the data of the synchronously-disconnected device. After the 牟 is repaired, it is necessary to start the control 1 and -1 to obtain the status information of the mapping storage device (step _ After that, check the = 1293428 device 1 data is deducted (step 6G2), to the most complete storage of the data _f_ two: the status message to find the control of the storage damage 1 " (4 will controller 1 storage device 1 add _ lose: (== paste) 'domain will be restored to the original state, as 々 5), the sample system can be restored, the material / the person will not have any effect. In -i square ί can The problem that extends to multiple controls (four) occurs, as long as the system can be added to the map at any time after the normal device is repaired: = the control of Jing Jing reaches the storage controlled by each controller = in the convenient operation by synchronous operation. Coffee loses _=^ with, _ protect system ten reserve ====== ::= seesaw, two disk devices, each motherboard two::: 2= information. All disk devices are built Into a map storage: preparation ' _ age set _ 峨 峨 峨 峨 听 听 , , , , , , , , , , , , , , , , , , , , On the disc, the problem is that the problem is not normal, and the magnetic r managed by it will have problems. At this time, the mapping storage device becomes a reduced-order state, and the configuration label of the hanging motherboard changes. Recording another motherboard damage information storage device can still be used normally, the user's data is recorded on 1293428 that = problem disk device, nothing happens to the user. Plywood recovery After normal work, again (four) social machine board, the host does two or two normal main _ configuration slot 'and then map the storage device to the same data on the block disk, and then modify the two matching changes. For the user, there is no other (10) the same is true for the storage of the financial material. The synchronization process of the backup preparation device is used because the mapping storage device itself has a comparison to implement the mapping mechanism. The device needs to rely on the self-editing side to perform: after the complex stepping process line, the user's reading and writing does not open the first m to 1 in the back end into the second 'and then write the data every one state normal control No failure and original writing completed (Step 2 = Γ通::=_过过(四)看控制-'Through the control--:::==: No~Set (Steps • This can be _ 1 =:: 1293428 month]=Write data that is not stored in the storage device (step _, then judge data = enter = success (step 9 () 5), if successful, mark at the end of the writer's data, 9 (m 4 material writing is completed (step 9G6) 'Other reports failure and reason (steps. The configuration data of the configuration slot data block includes the storage device Id, the establishment date, the size of the writer data, the ID of the controller) And other information, etc. The method of the present invention has the following advantages: h There is no process of switching controllers, which greatly improves the reliability and security of the system, and the system does not falter when the controller fails to convert. 2. Do not use the temporary storage H' to improve the performance of the reduction and the security of the user data, even when a single control failure occurs, the manufacturer's (4) can still copy to the mapping storage device, without the intermediate buffer Reduced memory consumption. 3. Make the whole system turn to compare branches. When the problem occurs after the control is made, the controllers that have problems can be repaired. The above-mentioned preferred embodiment of the present invention is not limited to the spirit and scope of the present invention, and is modified and retouched, = the scope of patent protection of Yang Ming. The scope of the patent application attached to the protection model defined by the present invention. / [Simplified description of the drawings] Figure 1 is a schematic diagram of a conventional dual controller with a mapping storage device; Figure 2 is a schematic diagram of a conventional dual controller with a controller for mapping a storage device 11 1293428 process; The diagram is a schematic diagram of the failure of the controller of the present invention; and FIG. 4 is a flowchart of the method of the present invention; 帛5 _ a flow chart of the bribe device of the present invention; • L6: the synchronous storage device of the present invention FIG. 8 is a schematic diagram of an embodiment of the present invention; FIG. 8 is a flowchart of the embodiment of the present invention, which is a flowchart of the recovery of the embodiment of the present invention; Symbol Description] _ data flow chart. 10...................storage store:: 4°1 allocation per-controller only controls the corresponding only-storage #备步4°2 will be plural The storage device establishes a mapping storage setting ==::r. The mapping storage device has the shape==1.##有任-织继_, determines at least-to the outer Γw as 'and continues to receive the user The data is recorded in the 4 mapping storage device. σ················································································ The only step 5 of _. ;:1:; Size of the device on the controller Step 5〇3 — The error setting --I:,;::::::- Immediate mapping to the message prepared for each occupied map: 5°6........Report failure 2 IIZ controller 1 obtains the status message of the mapping storage device.········Check whether the controller 1 data is complete 4Γ03Π Find the most according to the recorded information Complete storage a Further preparation for synchronization operation 04········Complete the data of the storage device 1 controlled by the controller 1 v秣605........Add the storage device 1 Go to the mapping storage device Step 801··..·.. View the status of the controller Step 802 ········Get the data transmitted by the network layer v Private 803........Write the poor material Enter the storage device of each state normal controller. Step 804·..·.. Determine whether the data is written successfully. Step 805···· Enter.... Mark the end of the written data to prove the Data writing step 806 ..... report failure and cause step 901 ··.. .... view the state of the controller step 902 ..... ... to find the controller 129 without damage through the configuration file 3428 Step 903 ·················································································· Write step 905····.. determine whether the data is written successfully. Step 906 ·... Entering.... Marking at the end of the written data, proving that the data is written in step 907..... Failure and cause