201001165 九、發明說明： 【發明所屬之技術領域】 於於計算裝置的領域，更特別地，係關 於-^用以有效率地讓計算裝置根開機的系統與方法。 【先技術】 =叫。t b〇〇ting)是使用像是附接儲存陣列 attached storage array) ^ W # ^ # ^ £ ^ tl3，(VGlume)將計算|置開機的實施方式。該 tilt ί裝置由本地磁碟機(locaid—開機的之本地 所有貧料（包括作業系統、驅動程式、暫存播案、 應用私式、以及/或者虛擬記憶體交換空間）。 圖Γ示為用以根開機的一典型系統100,系統100 存裝置•每一計算裝請一般由在每 上=為:虛擬卷(virtual V()lume)〗05之附接儲存裝置收 的-專屬卷(dedicated ν〇1_)ι〇4開機。接著每 吏用其個別的專屬卷104 (透過個“ 虛挺卷105)來操作。當每一計算裝置1〇1在操作发 個別的專屬4 104讀取資料以及/或者寫入資料至盆個为、 的專屬卷1〇4 (透過個別的虛擬卷1〇5)，來自個別的 卷1〇4的資料一般會被儲存在附接儲存裝置102的快取 (Γ^1⑽i的—個別的區域’以便快速存取。第二圖所示 為母一計算裝置101與系統1〇〇的關係。每一計算 ^ ’、月b夠存取用於計算裝置之個別的卷1〇4與附接储 1〇2的快取1〇3的個別區域，儘管計算裝置1〇1實\ 認知其附接儲存裝置1〇2的個別虛擬卷1〇5。 π ，、 6 201001165 【發明内容】 一種用以根開機的系統與方法可包括複數個通訊地 連接至一附接儲存裝置的計算裝置。每一計算裝置可從該 附接儲存裝置的一唯讀基卷(read-only base volume)開機。 唯讀基卷可包括通用於該複數個計算裝置的資料，其中包 括但不限於一或更多個作業系統、驅動程式(driver)、暫存 檔案、應用程式，以及/或者虛擬記憶體交換空間。唯讀 基卷的常存取資料(frequently accessed data)可被儲存在附 接儲存裝置的快取的一部分。附接儲存裝置也可包括複數 個卷，母一個卷專屬於複數個計算裝置的其中之一，並為 在該唯璜基卷的寫入快照(write snapSh〇t)上之重定向。用 於每一計算裝置的個別的卷可包括用於一計算裝置的唯一 員目像疋唯一的注冊金鑰、虛擬記憶體交換空間，以及 /或暫存檔案。如果任何一個計算裝置對唯讀基卷的資料 進行嫒動，該變動可被指向並儲存於該計算裝置所用之個 別的卷中母卷的¥存取資料可被儲存於用於該個別的 卷之附接儲存裝置的快取的—部分。唯讀基卷、快取的該 邛刀以及個別的卷可以一虛擬卷(咖㈣的方式 =給每-賴數個計算裝置。#來自唯讀基卷的資料沒 = :樣置變動時，儲存系統可從唯讀基卷映射 =至用於母—該複數個計算裝置的_的虛擬卷， 可糾個別的計算裝置變動時，儲存“ ::: =射資料至用於每-該複數個計算裝置- 或更多個固態磁碟㈣“-〜、记隐體來儲存持續的資料的資料儲 7 201001165 存叙置），该固恶磁碟可被设定為一獨立磁碟冗餘陣列 (redundant array of independent disk, RAID)以及/¾ 者以 用於備援用途的一或更多個其他儲存驅動器（st〇mge drive) 被鏡射。該複數個卷的每一個可被儲存於一或更多個可被 設定為- RAID的固態磁碟以及/或者硬碟機。附接儲存 裝置可被操作以將通用於複數個計算裝置的資料加入至該 唯，基卷。因此，用來根關複數個計算裝置的唯讀基卷 可藉由將新的通用資料指定給該唯讀基卷而被更新。 本發明可改善從附接儲存裝置的卷根開機複數個計算 裝置所用的儲存空間、1/〇、以及快取的低效率。本發明藉 ，唯讀基卷儲存於-或更多個固態磁碟，可利用“』 效能。由於複數個計算裝置只有使用:個 硬碟機固態磁碟所提供的讀取效能和固 成本之略吨得伟。料，本發明 (deduphcat峨响，也可達到較簡單的系物i式 明僅;前的购與後續詳細的實施職 的圖式係合併入並組成=疋本發明。本說明書所附 【實施方^】圖不係相互用简釋本發_主要精神。 =將配合所附圖表詳細說明本發明。 對複數個計算裝置根開機_ 率的。在許多算:置使用，個別的卷是沒有效 的卷開機時，包人/田稷個计算裝置從附接儲存裝置 包含在個別的卷内的資料（包括但不限於作 201001165 業糸統、驅動程式、暫存檐牵、 擬記憶體交換空間等）可能^ :程式、以及/或者虛 可能合被、t眷掉。手相同。因此，儲存空間 J 被良費掉此外，複數個伺服哭从—加各目士 或更多個作業系統、驅動程式、二的母—個…有一 a / its h*· 9存描案、應用程式，以 及/或者纽錢體交換空__ 個伺服器的每-個來說，同構 J複本1=1此對I數 0 .Τ/ 门樣的貝料會從個別的卷個別地 5貝取所以，為了從個別的卷讀取重複個資料，就合有不 大需要的1/0(輪入/輪屮八+政後1U貝才十跣曰有不201001165 IX. Description of the Invention: [Technical Field of the Invention] In the field of computing devices, and more particularly, it relates to systems and methods for efficiently turning a computing device into a root. [First technique] = call. t b〇〇ting) is an implementation that uses the attached storage array) ^ W # ^ # ^ £ ^ tl3, (VGlume) to calculate | The tilt ί device consists of a local drive (locaid—all local poor (including operating system, driver, staging, application private, and/or virtual memory swap space). A typical system 100 for root booting, system 100 storage device • Each computing device is generally received by the attached storage device for each virtual disk (virtual V () lume) 05 - exclusive volume ( Dedicated ν〇1_) ι〇4 is turned on. Then each uses its own dedicated volume 104 (via a "virtual volume 105"). When each computing device 1〇1 is read in the operation of the exclusive 4 104 The data and/or the data to be written to the exclusive volume 1〇4 (through individual virtual volumes 1〇5), the data from the individual volumes 1〇4 will generally be stored in the attached storage device 102. Take (Γ^1(10)i-individual area' for quick access. The second figure shows the relationship between the parent-computer device 101 and the system 1〇〇. Each calculation ^', month b is sufficient for access to the computing device Individual volumes 1〇4 and individual areas of the cache 1〇3 that are attached to the storage 1〇2, despite the computing device 1 〇1实\ Cognition of the individual virtual volume 1〇5 attached to the storage device 1〇2. π , , 6 201001165 SUMMARY OF THE INVENTION A system and method for root booting may include a plurality of communication connections to an attachment A computing device of the storage device. Each computing device can be powered from a read-only base volume of the attached storage device. The read-only base volume can include data common to the plurality of computing devices, including However, it is not limited to one or more operating systems, drivers, temporary files, applications, and/or virtual memory swap space. Frequently accessed data of read-only base volumes can be stored. Attached to a portion of the cache of the storage device. The attached storage device may also include a plurality of volumes, one of which is dedicated to one of the plurality of computing devices, and is a write snapshot of the read-only base volume (write Redirection on snapSh〇t). Individual volumes for each computing device may include a unique object for a computing device, a unique registration key, virtual memory swap space, and/or temporary storage If any of the computing devices incites the data of the read-only base volume, the change can be directed to and stored in the individual volume used by the computing device. The access data of the parent volume can be stored for the individual The cache-attached storage device's cache-part. The read-only base volume, the cached file, and the individual volumes can be a virtual volume (the way of the coffee (four) = for each of the several computing devices. # from The data of the read-only base volume is not =: When the sample is changed, the storage system can be stored from the read-only base volume map = to the virtual volume used for the mother-the plurality of computing devices, and can be stored when the individual computing device is changed. " ::: = Shooting data to use for each - the plurality of computing devices - or more solid-state disks (four) "- ~, remembering the hidden material to store the data of the persistent storage 7 201001165 Save the set), the solid The floppy disk can be set to a redundant array of independent disk (RAID) and one or more other storage drives (st〇mge drive) for backup use. Mirror shot. Each of the plurality of volumes can be stored in one or more solid state disks and/or hard disk drives that can be set to - RAID. The attached storage device is operable to add data common to the plurality of computing devices to the base, the base volume. Thus, a read-only base volume used to root a plurality of computing devices can be updated by assigning new generic data to the read-only base volume. The present invention can improve the storage space, 1/〇, and cache inefficiency used to power up a plurality of computing devices from the root of the attached storage device. According to the present invention, the read-only base volume is stored in one or more solid-state disks, and the """ performance can be utilized. Since a plurality of computing devices use only the read performance and the solid cost provided by the hard disk drive solid state disk. In the present invention (deduphcat squeaking, it is also possible to achieve a simpler formula i formula; the previous purchase and subsequent detailed implementation of the schema are incorporated and composed = 疋 this invention. The accompanying [implementation party ^] diagrams attached to the manual are not mutually exclusive. The main spirit is explained. The invention will be described in detail in conjunction with the attached chart. The root of the plurality of computing devices is turned on. In many calculations: Individual volumes are inefficient when the volume is turned on, and the data is included in the individual volumes from the attached storage device (including but not limited to 201001165 industry, driver, temporary storage) , the memory of the memory exchange, etc.) may be: program, and / or virtual may be combined, t 眷 off. The same hand. Therefore, the storage space J is expensive. In addition, a plurality of servos cry from - plus the various men or More operating systems, drivers, The mother-a... has a / its h*· 9 deposit description, application, and / or Newton exchange space __ for each server, isomorphic J replica 1 = 1 this pair I The number of 0. Τ / door type of bait will be taken from individual rolls individually 5, so in order to read duplicate data from individual volumes, there is a small need for 1 / 0 (round / rim eight + After the political administration, 1U is only ten or not.
別ΐ卷個別的切能會制在附接儲存裝置内的有限快 =工間（以及/或者—可通訊地連接複數 ^之-至附接儲存裝置的控彻）。因此，複數個伺服器的 母-個可能會碰到更多的快取未中(eachemiss)，導致為了 要與儲存於—蚊多_存_⑼個關卷互動要花更 多的I/O，進而降低了效能。 /第三圖所示為根據本發明的一實施例之用以根開機的 系統300，系統300包括複數個計算裝置3〇1可通訊地連 接至一附接儲存裝置302。複數個計算裝置3〇1可包含任 種數位计异裝置，包括但不限於，個人桌上型電腦、個 人膝上型電細、伺服器電腦、以及/或者虛擬終端機 (dummy terminal)。每一計算裝置3〇1可從附接儲存裝置 302上的一唯讀基卷305來開機，該附接儲存裝置邛^上 的一唯讀基卷305係呈現給每一計算裝置3〇1為一用於該 計算裝置之虛擬卷(virtual v〇lume)307。唯讀基卷3〇5可包 括通用於複數個計算裝置301的資料，包括但不限於一= 更多個作業系統、驅動程式、暫存檔案、應用程式、'以及 /或者虛擬記憶體交換空間等。唯讀基卷的常存取資料可 9 201001165 被儲存於附接儲存裝置302的快取3〇3的一部分3〇4 Γ的該部分3〇4可儲縣讀基卷中常被複數個 计^裒置101的任一個（透過個別的虛擬卷3〇7) 貧料，因此快取303的該部分304在整個快取中所佔卜 例’應可比如果用到個別的卷（會要求快取的個別 所佔的比例要多’而且在—給鱗間内，在快取中用二 一該複數料算裝置3G3的資财叹多，以翻較少: ，取未中。附接儲存裝置3Q2也可同樣包括複數個卷、， 母一個卷專屬該複數個計算裝置3〇1的1 =為在該唯讀基卷奶的寫人快⑽ 2疋向。用於每-計算裝置3()1的個別的卷规可包括 =於該计异裝i 301的唯-項目，像是唯一的註冊金 虛，記憶體交換”、以及，或㈣韻案。 計算裝置301對唯讀基卷305的資料進錄 " :虛擬卷叫該變動可被指向並儲存於:七:： 用的個別的卷306巾。每一卷3〇6的 01所 於用於個別的卷3〇6之附接儲存裝置 3〇被= •當來自唯讀基卷3晴料沒有為個別的; 數個計算裝置3〇1的個別的虛擬Ϊ 來自唯§買基卷305的資料右為個另丨丨» 田 時’儲存系統可從個別的卷遍映射、資^至裝置變動 數個計算裝置3〇1的個別的虛擬卷。"用於母一該複 複數個計异農置3〇1可句合 卷305可被儲存於—或更多個固態砰°唯讀基 來儲存持續的資料的資料儲存裝置)。== 10 201001165 一或更多個被設定為—獨立磁 of independent disk, RAID)之儲存驅:哭歹I a卿 驅動器可被設定為—沒有實施備援的：·心= 被勿割儲存於複數個儲存驅動器内 可^ 於-或更多個儲存驅動器, 土卷Q5可储存 ^丄 邊储存驅動器以一戎f容徊甘 他儲存驅動器鏡射，以作為備 個其 被儲存独其賴存驅絲鏡㈣—或更乡 中，複數個許瞀获罢2 η 1 ~τ· * $ 4子’驅黃? 4Don't slap individual cuts to make a limited speed in the attached storage device (and/or - communicatively connect multiples - to the control of the attached storage device). Therefore, the mother-of-servers of multiple servers may encounter more caches (eachemiss), resulting in more I/Os in order to interact with the storage--mosquito__9) , which in turn reduces performance. The third diagram shows a system 300 for rooting in accordance with an embodiment of the present invention. System 300 includes a plurality of computing devices 3.1 connected communicably to an attached storage device 302. The plurality of computing devices 3.1 may include any of the digital counting devices including, but not limited to, a personal desktop computer, a personal laptop computer, a server computer, and/or a dummy terminal. Each computing device 3.1 can be powered up from a read-only base 305 attached to the storage device 302, and a read-only base 305 on the attached storage device is presented to each computing device 3〇1 Is a virtual volume (virtual v〇lume) 307 for the computing device. The read-only base volume 3〇5 may include data common to a plurality of computing devices 301, including but not limited to one = more operating systems, drivers, temporary files, applications, and/or virtual memory swap space. Wait. The frequently accessed data of the read-only base volume can be stored in the cache 3 of the cache storage device 302. The portion of the cache 3 is a part of the 3 〇 4 〇 〇 读 可 常 常 常 常 常 常 常 常 常 常 常 ^ ^ ^ Any one of the devices 101 (through individual virtual volumes 3〇7) is poor, so the portion 304 of the cache 303 is occupied in the entire cache. 'It should be comparable if an individual volume is used (a cache will be required) The proportion of individual occupations is much more than that. In the case of the scales, the resources of the 3G3 are counted in the cache. In order to reduce the amount of money, the number of the 3G3 is too small. It is also possible to include a plurality of volumes, one of which is exclusive to the plurality of computing devices 3〇1 = 1 is the writer of the read-only base volume (10) 2 。. For each per-computing device 3 () The individual volume gauges of 1 may include = only items in the metering i 301, such as the only registered gold deficiency, memory swapping, and/or (4) rhyme. The computing device 301 pairs the read-only base volume 305 The data entry " : The virtual volume called the change can be pointed to and stored in: Seven:: The individual volume used is 306 towels. Each volume of 3〇6 of 01 is The attached storage device 3 of the individual volume 3〇6 is = • when the raw material from the read-only base volume 3 is not individual; the individual virtual files of several computing devices 3〇1 are from the only purchase base roll 305 The information on the right is an alternative. The "Tianshi" storage system can change the individual virtual volumes of several computing devices 3〇1 from individual volume mappings and devices to the device.异 农 〇 305 305 305 305 305 can be stored in - or more than one solid 砰 ° read base to store continuous data storage device). == 10 201001165 one or more is set to —Independent magnetic of independent disk (RAID) storage drive: crying I A Qing drive can be set to - no backup support: · heart = not stored in multiple storage drives can be - or more A storage drive, the soil roll Q5 can store the storage drive to store the drive mirror, as a preparation for the storage of the mirror (4) - or more, a number of Xu Wei won 2 η 1 ~τ· * $ 4 子 '驱黄? 4
:二可透過該-或更多個儲存驅動哭 f取，基卷305 (透過個別的虛擬卷3。7)，而非-1 夕個八他儲存驅動器。舉例來說，唯讀基卷3Q5可被儲 ，二或更多個輯磁碟’該固態磁碟[或更多個硬碟機 (一種數位地儲存編碼資料於呈右磁 揮發儲存裝置）鏡射，而複數個計算裝置則可透^= 或更多個固態磁碟存取唯讀基卷奶（透過個別的虛擬卷 3〇7)，而非該一或更多個硬碟機。複數個卷3〇6的每—個 可被儲存於-或更多個固態磁碟以及/或者硬碟機内 數個卷306的每一個可被儲存於一或更多個被設定為一 RAID的儲存驅動器内。儲存驅動器可被設定為—沒有實 施備援的RAID，其中資料會被分割儲存於複數個儲存驅 動器内。附接儲存裝置302可包含一或更多個網路附接儲 存(network attached storage，NAS)裝置以及/或者—戋更 多個儲存區域網路(storage area network，SAN)裝置。附接 儲存裝置302可包含一或更多個RAID儲存裝置。 第四圖所示為每一個計算裝置301所看到的系统 3〇〇。每—個計算裝置3〇1僅能存取唯讀基卷3〇5、用來 儲存來自唯讀基卷305的常存取資料之快取3〇3的部分 11 201001165 304^、個別的卷306以及用來儲存來自個別的卷的常存 取，料之快取303的部分304，不過計算裝置3〇1實際上 只涊知其個別的附接儲存裝置3〇2的個別的虛擬卷。 附接儲存裝置302可被操作以將通用於複數個計算裝 ί =的資!ί加入唯讀基卷305。在此方式下，用來根開 ^個5十异I置的唯讀基卷3〇5可藉由將新的通用資 料才曰定唯讀基卷305而被更新。 、 一儘官本發明已經說明與敘述了複數個計算裝置，其由 : = = =至而計算裝 的計算裝置的通用的唯夂;= = =丄，然而要注意的是不只—種複數個物置，也； :通用於該個別的複數個計算装置之個別的唯讀 飼服器與十個微軟Win J 固Linux 個附接儲存f ^ T. 均可通祕_至同一 通用於該十個ij· 服器可從附接儲存裝置中 個微—服開機，而十 —飼服器的—通用於該十個 所需======開機複數個計算裝置 將唯讀基卷3〇5儲存於一或’低效率。本發明可藉由 :碟供的讀取效能。由於；由:態 用-個唯讀基卷305，相較 十：ξ置加只有使 轉效能和固.態磁碟-般較高的成本之提衡供的 12 201001165 另外，本發明並不需要在ι/ο路徑中使用重覆刪除的程式 碼(deduplication code) ’也可達到較簡單的系統3〇〇之^ 施。 、 第五圖所示為根據本發明的一實施例之複數個計算妒 置之根開機的方法。在步驟501，由一被儲存於至少一^ ，存驅動益的附接儲存裝置的一唯讀基卷根開機複數個 算裝置，該唯讀基卷包含該等計算裝置通用的至少一個作 業系統或至少一個應用程式的其中至少一個，該至少一個 儲存驅動器可包含至少一個固態磁碟。該至少一個儲存驅 動态可包含被设疋為一 RAID之複數個儲存驅動器。複數 個計算裝置可包含複數個無磁碟的計算裝置。附接儲存裳 置I包含i接的RAID裝置。在步驟5〇2，從該複數個 计，裝置接收用於該唯讀基卷的寫入。在步驟5⑽，將用 於，邊複數個計异裝置之寫人重定向至該附接儲存 的複數個卷的其中—個，該附接儲存裝置的複數個卷^ 在該唯，、基卷的寫人快照上之重定向，該附接儲存裝置二 複數個卷的每一個係專屬於該等計算裝置的其中之一 =定於該等計算裝置的其中之一的資料。該複數個卷的 母一個可被儲存於至少—個硬碟機。該複數個卷的每 =儲存於被設定為—RAID之魏個硬碟财。特定於 該專稷數個計算裝⑽其巾之—的#料可包含至少一個註 冊金餘、至少—個虛擬記紐交換空間、或至少 在 標案的至少其中之一。 =六圖所示為根據本發明的—替代實施例之複數個計 根開機的方法。在步驟601 ’由-被儲存於至少 储存驅動器謂接儲存裝㈣—唯讀絲㈣機複數 13 201001165 個作業^ # 5"唯讀基卷包含該料算裝置通用的至少一 一個儲;一個應用程式的其中至少—個，該至少 存驅動器可包含至少一個固態磁碟。該至少-個儲 複數個ϋ含被設定為—RAn^複數個儲存驅動器。 儲存裝id:包含複數個無磁碟的計算裝置。該附接 少-個附加二：：接的raid裝置。在步驟6〇2，以至 驟6〇3，It存=動器鏡射該至少一個儲存驅動器。在步 入。在步驟:裝置触綠該唯讀基卷的寫 定向至該附接於母—該複數個計算裝置之寫入重 存芽詈的、裝置的複數個卷的其中一個，該附接儲: Two can use the - or more storage drivers to cry, the base volume 305 (through individual virtual volumes 3. 7), rather than -1 ‧ eight eight storage drives. For example, the read-only base volume 3Q5 can be stored, two or more sets of disks 'the solid-state disk [or more hard disk drives (a digitally stored coded data in the right magnetic volatilization storage device) mirror Shots, and a plurality of computing devices can access the read-only base volume (through individual virtual volumes 3〇7) through the ^= or more solid-state disks instead of the one or more hard drives. Each of the plurality of volumes 3〇6 may be stored in one or more solid state disks and/or each of the plurality of volumes 306 in the hard disk drive may be stored in one or more of which are set to be a RAID. Store in the drive. The storage drive can be set to — RAID without backup, where the data is split and stored in multiple storage drives. Attached storage device 302 can include one or more network attached storage (NAS) devices and/or - more storage area network (SAN) devices. Attachment storage device 302 can include one or more RAID storage devices. The fourth diagram shows the system 3 seen by each computing device 301. Each of the computing devices 3〇1 can only access the read-only base volume 3〇5, and is used to store the cached 3〇3 portion of the frequently accessed data from the read-only base volume 305. 201001165 304^, individual volumes 306 and portion 304 of the cache 303 for storing frequent accesses from individual volumes, but the computing device 〇1 actually only knows the individual virtual volumes of its individual attached storage devices 〇2. The attached storage device 302 can be operated to add a plurality of computing devices ί 加入 to the read-only base volume 305. In this manner, the read-only base volume 3 〇 5 used to root the 十 异 I can be updated by setting the new general-purpose data to the read-only base 305. The present invention has been described and described with respect to a plurality of computing devices, which are: = = = to calculate the general purpose of the computing device; = = = 丄; however, it is noted that not only a plurality of Object, also;: The individual read-only feeders that are commonly used for the individual computing devices and the ten Microsoft Win J-solid Linux attached storage f ^ T. can be used for the same ten The ij. server can be turned on from the attached storage device, and the ten-feeding device is used for the ten required ====== booting a plurality of computing devices will read the base volume 3〇 5 stored in one or 'inefficient. The invention can be used by: the read performance of the disc. Because: the state uses - a read-only base volume 305, compared to the ten: ξ set plus only the transfer efficiency and the solid state disk - the higher cost of the supply 12 201001165 In addition, the present invention is not You need to use the deduplication code in the ι/ο path to achieve a simpler system. The fifth figure shows a method for rooting a plurality of computing devices in accordance with an embodiment of the present invention. In step 501, a plurality of computing devices are booted by a read-only base volume stored in at least one of the attached storage devices, the read-only base volume including at least one operating system common to the computing devices. Or at least one of the at least one application, the at least one storage drive may comprise at least one solid state disk. The at least one storage drive dynamic can include a plurality of storage drives that are configured as a RAID. A plurality of computing devices can include a plurality of computing devices without magnetic disks. The attached storage device I contains a RAID device connected to it. At step 5, 2, from the plurality of devices, the device receives a write for the read-only base volume. In step 5 (10), the writer for the plurality of different devices is redirected to one of the plurality of volumes that are attached to the storage, and the plurality of volumes attached to the storage device are in the Redirecting on the Snapshot of the Writer, each of the plurality of volumes of the attached storage device is specific to one of the computing devices = data destined for one of the computing devices. The parent of the plurality of volumes can be stored on at least one hard disk drive. Each of the plurality of volumes is stored in a hard disk that is set to -RAID. The # material specific to the plurality of computing devices (10) may include at least one registration amount, at least one virtual token exchange space, or at least one of the standards. = Six figures show a plurality of methods of rooting up in accordance with an alternative embodiment of the present invention. In step 601 'by- is stored in at least the storage drive, the storage device (four) - read only silk (four) machine plural 13 201001165 jobs ^ # 5 " read-only base volume contains at least one storage common to the computing device; At least one of the applications, the at least one storage drive may include at least one solid state disk. The at least one of the plurality of storages is set to -RAn^ a plurality of storage drives. Storage id: A computing device containing a plurality of non-magnetic disks. The attachment is less - an additional two:: the raid device is connected. At step 〇2, and at step 〇3, the ITO device mirrors the at least one storage drive. In step. At step: the device touches green and the write of the read-only base volume is directed to one of a plurality of volumes of the device attached to the parent-the plurality of computing devices that write the duplicate buds, the attached store
算裝置的其中、置的複數個卷的每一個係專屬於該等計 :資料:定於該等計算裝置的其中之-機。該複數母—個可被儲存於至少一個硬碟 複數個硬碟機ί =：：個可被儲存於被歧為—RAID之 的資料可包含至，丨、j疋於該等複數個計算裝置的其中之一 換空問、,少個註冊金鑰、至少一個虛擬記憶體交 ^少—個暫存财的至少其中之-。 計算發::另-替代實施例之獅 少—個儲存驅動哭沾法。在步驟701，由一被儲存於至 數個計曾Μ ^附接儲树置的—唯讀基卷根開機複 -個作ΐ她;少==等計算裝置通用的至少 少一個式的其中至少—個，該至 儲存驅動器可包=至少-個固態磁碟。該至少-個 器。複數個計算裝置^^:RAID之複數個儲存驅動 匕3複數個無磁碟的計算裝置。附 14 201001165 接儲存裝置可包含—附麵RAID裝置 該複數個計算裝置接收用於該唯 ^驟702 = 70=用於每—該複數個計算裝置之寫人重定向至該^接 复數個卷的其中一個，該附接儲存裝置的複數 個卷係為在該唯讀基卷的寫人快照之重定向， 裝置的複數個卷的每—個係專屬於該等計算裝置: 一亚儲存特定於該計算裝置的其中之一 —中 -二卷的每 定於該等複數個計算裝置的其中之一的資料可=::: ==錄、至少—個虛擬記憶體交換空間、 暫存播案的至少其中之_。在步驟7G 個 至唯讀基卷。 ’遇用貝抖加入 在本發明中’所揭示的方法可以指 可讀取的軟體來實施。此外，應可了解的：j:由-裝置 示的特定次序或步驟的階層係示範方法的2例。：：所揭 欠ί或步驟的階層可根據設計偏好，ί法中 新排列’㈣仍屬於本發明的料。所附的 嗅仃重 的次序來說明各個步驟的元件，而這些=从示範 限在所示的特定次序或階層。—亚不一定要偈 伴隨=由::.解本發明之揭露〜 部的實質效果Τ，料_可的林犧牲其全 排列作出多種的變更’所描述的 、^式、結構、 申請專利範圍係能夠涵蓋且包含該等㈣了明性的， 15 201001165 【圖式簡單說明】 藉由參考該等隨附圖式，可由本技術領域中具有通常技藝 者更佳地瞭解本發明所揭露的許多特點： 第一圖所示為用於根開機的一典型系統； 第二圖所示為從複數個計算裝置的其中之一之觀點來看第 一圖的系統； 第三圖所示為根據本發明的一實施例之用以根開機的系 統； 第四圖所示為根據本發明的一實施例，從複數個計算裝置 的其中之一之觀點來看第三圖的系統； 第五圖的流程圖所示為根據本發明的一實施例之用以根開 機的方法； 第六圖的流程圖所示為根據本發明的一替代實施例之用以 根開機的方法；以及 第七圖的流程圖所示為根據本發明的另一替代實施例之用 以根開機的方法。 【主要元件符號說明】 100系統 101計算裝置 102附接儲存裝置 103快取 104專屬卷 105虛擬卷 300系統 16 201001165 301計算裝置 302附接儲存裝置 303快取 304快取的部分 305唯讀基卷 306個別的卷 307虛擬卷Each of the plurality of volumes of the computing device is specific to the meter: data: the machine that is scheduled for the computing device. The plurality of mothers can be stored on at least one hard disk and the plurality of hard disk drives ί =:: can be stored in the data that is classified as - RAID can be included, 丨, j 疋 in the plurality of computing devices One of them changes the space, and there is one less registration key, at least one virtual memory, and at least one of the temporary deposits. Calculation hair:: Another - alternative embodiment of the lion less - a storage drive crying method. In step 701, at least one of the computing devices is used by a computing device that is stored in a number of counts and attached storage trees. At least one, the storage drive can be packaged = at least one solid state disk. The at least one. A plurality of computing devices ^^: a plurality of storage drives of RAID 匕3 a plurality of computing devices without magnetic disks. Attachment 14 201001165 The storage device may comprise - a surface RAID device, the plurality of computing devices receiving for the 702 = 70 = for each of the plurality of computing devices, the writer redirects to the plurality of volumes One of the plurality of volumes attached to the storage device is a redirection of a write snapshot of the read-only base volume, each of the plurality of volumes of the device being dedicated to the computing device: Each of the plurality of computing devices of the computing device may be =::: == recorded, at least one virtual memory swap space, temporary broadcast At least one of the cases. In step 7G, the read-only base volume. The method disclosed in the present invention may be implemented by means of a readable software. In addition, it should be understood that: j: 2 instances of a specific order or step hierarchy shown by the device. :: The hierarchy of the steps or steps that can be revealed can be based on design preferences, and the new arrangement in the method is still a material of the present invention. The accompanying order of smelling is used to illustrate the elements of the various steps, and these = are limited to the particular order or hierarchy shown. -Asia does not have to be accompanied by = by::. to solve the invention of the invention ~ the substantive effect of the Ministry 料, _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ </ RTI> </ RTI> </ RTI> </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; Features: The first figure shows a typical system for root booting; the second figure shows the system of the first figure from the point of view of one of the plurality of computing devices; A system for rooting an embodiment of the invention; the fourth figure shows a system of the third figure from the perspective of one of a plurality of computing devices, according to an embodiment of the present invention; The flowchart shows a method for root booting according to an embodiment of the present invention; the flowchart of the sixth diagram shows a method for root booting according to an alternative embodiment of the present invention; The flow chart is shown as Invention with another alternative embodiment of the method of root booting embodiment. [Main component symbol description] 100 system 101 computing device 102 attached storage device 103 cache 104 dedicated volume 105 virtual volume 300 system 16 201001165 301 computing device 302 attached storage device 303 cache 304 cached portion 305 read-only base volume 306 individual volume 307 virtual volumes