M432075 - ， . 第100220008號專利申請案 * * 101年3月13日修正替換頁 五、新型說明： 【新型所屬之技術領域】 本創作係關於一種應用於雲端計算之虛擬機(Virtual Machine ;簡稱VM)控管技術’尤係關於一種結合非同步 通訊機制之虛擬機集群(cluster)控管技術。 【先前技術】 雲端計算是一種透過網路連接而使用相關軟體的服 務模式，可使客戶端在進入雲端計算系統後獲取需要的資 鲁源。 現今許多雲端計算系統的部署皆建置於虛擬機集 群，具體而言，透過虛擬化方法能將計算資源虛擬化，進 而切割硬體的計算和儲存能力。而在此雲端架構中，如何 透過有效且穩固的機制監控與管理虛擬機集群，即成為雲 端計算系統優劣的關鍵因素。 目前市場上推出的虛擬機管理平台（VMHypervis〇r) φ 係能針對虛擬機的新建、佈署、管理等方面進行控管，然 而，此虛擬機管理平台卻具有不利於開放式的雲端環境的 缺點。其次，虛擬機之管理過程都需要一定的回應時間， 而差異化高的混合虛擬架構更增加控管機制的不確定性， 如採用同步化（synchronized)之通訊模式，將會影響整體 機制的效能以及可靠性。再者，針對虛擬機進行相關的部 署時，往往需要登入虛擬機取得相應的使用權限，但此舉 部今易讓使用者對虛擬機的安全性及資料隱私產生一定程 度的疑慮。另外，於現有的雲端計算系統中，針對虛擬機 3 M432075 第100220008號專利申請案 101年3月13 '日修正替換頁 集群的控管處理，並沒有提供一個有效的控管備援機制， 因此一旦發生故障，則會造成重要的控制資訊遺失，並迫 使虛擬機集群產生斷訊之情況。 综上所述，如何提供一種嶄新的控管方案，以克服現 有針對虛擬機集群的控管技術存在的種種缺陷，即為各界 急欲解決的課題。 【新型内容】 有鑒於上述習知技術之種種缺點，本創作之主要目的 之一，係在於提供一種能克服同步通訊機制之缺失、提供 備援機制、及能充分應用於雲端計算之控管裝置及控管系 統。 為達上述目的及其它目的，本創作提供一種應用於雲 端計算之控管裝置，用於控管雲端計算系統中由複數虛擬 機組成的虛擬機集群，該應用於雲端計算之控管裝置係包 括：收發單元，係用於接收由網路服務平台所輸入之控制 指令訊息，並將該控制指令訊息傳送至指令發送緩衝區之 接收佇列；觸發單元，係用於接收與傳送該接收佇列中之 控制指令訊息；以及驅動單元區，係用於接收觸發單元所 傳送之控制指令訊息，而傳送控制指令訊息至虛擬機集群 中之虛擬機之至少一者，以驅動該虛擬機執行特定的處理 流程，以及接收該虛擬機於執行完成後所回傳之處理結果 訊息，並將該處理結果訊息傳送至指令回傳緩衝區之回傳 佇列，進而將該處理結果訊息經由該收發單元回傳至該網 路服務平台。 M432075 * * 第100220008號專利申請案 1〇1年3月13日修正替換頁 其次’本創作還提供一種應用於雲端計算之控管系 統’係包括：第—控管裝置，係具有用於接收由網路服務 平台所輪入之控制指令訊息並將該控制指令訊息傳送至指 令發送緩衝區之接收佇列的第一收發單元、用於接收與傳 送該接收佇列中之控制指令訊息的 第一觸發單元、用於接 收该第一觸發單元所傳送之控制指令訊息而傳送該控制指 令訊息至虛擬機群中之至少一者以驅動虛擬機執行特定的 處理流程及接收該虛擬機於執行完成後所回傳之處理結果 鲁訊息並將該處理結果訊息傳送至指令回傳緩衝區之回傳佇 列進而將該處理結果訊息經由該第一收發單元回傳至該網 路服務平台的第一驅動單元區，其中，該第一控管裝置另 具有第一共享儲存單元；第二控管裝置，係具有用於接收 由網路服務平台所輸入之控制指令訊息並將該控制指令訊 息傳送至指令發送緩衝區之接收仔列的第二收發單元、用 於接收與傳送該接收佇列中之控制指令訊息的第二觸發單 Φ 元、用於接收該第二觸發單元所傳送之控制指令訊息而傳 送該控制指令訊息至該虛擬機群中之至少一者以驅動該虛 擬機執行特定的處理流程及接收該虛擬機於執行完成後所 回傳之處理結果訊息並將該處理結果訊息傳送至指令回傳 緩衝區之回傳佇列進而將該處理結果訊息經由該第二收發 單元回傳至該網路服務平台的第二驅動單元區，其中，該 第一控管裝置另具有第二共享儲存單元；更新模組’係於 偵測到該第一收發單元接收到該控制指令訊息或該處理 結果訊息時，將該第一收發單元接收之該控制指令訊息 5 M432075 - 第100220008號專利申請案 101年3月13 &修正替換頁 或該處理結果訊息更新至該第一共享儲存單元；同步模 組，係於偵測到該更新模組已將該控制指令訊息或該處理 結果訊息更新至該第一共享儲存單元時，將該第一共享儲 存單元中儲存的控制指令訊息或處理結果訊息同步備份至 該第二共享儲存單元；以及切換模組，係於偵測到該第一 控管裝置之工作狀態發生異常時，啟動該第二控管裝置以 取代該第一控管裝置以執行控管任務。 相較於習知技術，本創作的控管裝置係結合非同步訊 息傳遞機制，以克服習知技術因使用同步訊息傳遞機制所 造成之缺失與瓶頸。此外，藉由本創作提供的控管系統， 還能於視為主機(master)的控管裝置發生故障或異常時，令 視為從機（slave)的其它控管裝置進行無缝接管，而因視為 主機的控管裝置及視為從機的其它控管裝置會同步備份資 料，所以不但能防止資料遺失，還可有效提升整體之可靠 度。 【實施方式】 以下藉由特定的具體實施形態說明本創作之技術内 容，熟悉此技藝之人士可由本說明書所揭示之内容輕易地 瞭解本創作之其他優點及功效，亦可藉由其他不同的具體 實施形態加以施行或應用。 請先參閱第1圖以瞭解本創作提供的應用於雲端計算 之控管裝置，而第1圖係為本創作之應用於雲端計算之控 管裝置100的應用架構圖。 如圖所示，控管裝置100係用於控管雲端計算系統中 M432075 __ , . . 第100220008號專利申請案 i〇l年3月13曰修正替換頁 的虛擬機集群30，而虛擬機集群30可包含虛擬機31、虛 擬機32、虛擬機33、及/或虛擬機34，且控管裝置1〇〇還 施針對虛擬機31、32、33、34進行如新建或佈署之控制與 管理’同時，控管裝置100至少包括收發單元11〇、觸發 單元130、以及驅動單元區15〇。 收發單元110係用於接收使用者由網路服務平台2〇 所輸入之控制指令訊息，並將接收到的控制指令訊息傳送 至位於指令發送緩衝區12〇中的接收佇列121，其中，接 鲁 收件列121係採用訊息仵列（Message Queue ;簡稱MQ)的 控制機制。此外，收發單元110係用於接收由指令回傳缓 衝區160所回傳之處理結果訊息。於本實施例中，網路服 務平台20可為提供雲端服務之入口網站，而收發單元110 可利用TCP/IP協定與網路服務平台20連結通訊，其中， 藉由基於OSI架構之傳輸層，能更加有效地整合不同的系 統環境。 • 觸發單元130係用於將接收佇列121中之控制指令訊 息傳送至驅動單元區150，而驅動單元區150係透過驅動 單疋（如151、152)將控制指令訊息傳送至虛擬機集群3〇 中之虛擬機31、32、33、34之至少一者，以驅動接收到控 制指令訊息的虛擬機31、32、 33、34執行特定的處理流程。 於一實施例中’觸發單元130係可先行將接收佇列121中 的控制指令訊息傳送至待處理佇列140中，進而再等待適 口的時機進行傳送，其中，待處理佇列140係採用訊息件 列的控制機制。 M432075 - 第100220008號專利申請案 101年3月13 &修正替換頁 於另一實施例中，控制指令訊息的類別係不相同，而 觸發單元130係可先針對控制指令訊息的内容進行識別並 分類，例如檢查CPU狀態、檢查磁碟使用量、或者檢查虛 擬機的快取記憶體(VM cache memory)等；並且在識別並分 類完成後，觸發單元130將不同類別之控制指令訊息傳送 至待處理佇列140中並紀錄該類別資訊。此種不同類別的 專屬性設計，能進一步提高處理效率。於又一實施例中， 傳送至待處理佇列140的控制指令訊息均可附加相關的時 間戳記(time stamp)，藉此，觸發單元130復可於判斷待處 理佇列140中的控制指令訊息的時間戳記已超過預定時間 時，自動取消或回收已超過預定時間的控制指令訊息之傳 送作業，避免後續發生錯誤或中斷之情形。 於一實施例中，驅動單元區150具有驅動單元151及 驅動單元152，而驅動單元區150所包含的驅動單元151、 152的數量係可調整，不限於圖式中的數量。驅動單元區 150係用以接收待處理佇列140的控制指令訊息，並且依 照控制指令訊息的類別，由對應的驅動單元151、152傳送 至對應的虛擬機31、32、33、34。例如檢查CPU狀態的 控制指令訊息由驅動單元151傳送至虛擬機31，以驅動虛 擬機31檢查CPU狀態；將檢查磁碟使用量的控制指令訊 息由驅動單元152傳送至虛擬機32，以驅動虛擬機32檢 查磁碟使用量。 當虛擬機3卜32、33、34在特定的處理流程完成後， 由驅動單元區150中與虛擬機31、32、33、34相對應的驅 M432075 __ • · · 第100220008號專利申請案 _ 101年3月13曰修正替換頁 動單元151、152接收對應的虛擬機31、32、33、34所回 傳之處理結果訊息。例如驅動單元151係與虛擬機3ι相對 應’則由驅動單元151接收虛擬機31所回傳之處理結果訊 息；驅動單元152係與虛擬機32相對應，則由驅動單元 152接收虛擬機32所回傳之處理結果訊息。 當驅動單元區150的驅動單元151、152接收到對應 的處理結果訊息後，驅動單元151、152將處理結果訊息傳 送至位於指令回傳緩衝區16〇中之回傳佇列161，進;將 •處理結果訊息依序經由收發單元110回傳至網路服務平台 2^0 ’以供網路服務平台2〇予以顯示，並進行相關的處理措 轭，其中，回傳佇列161係採用訊息佇列的控制機制。藉 此’使用者遂可透過網路服務平台2G從遠端對虛擬機集群 t的虛擬機31、32、33、34進行相關的處理與控管。 需說明的是，虛擬機31、32、33、34還分別可設有 处單元311、321、331、341，用於接收驅動單元區15〇 鲁 ^之控制才曰令訊息，並據以執行特定的處理流程，進 而傳送處理結果訊息。再者，於虛擬機3卜％、33、34 中一還刀別可預先裝載有相關的代理程式（mq吨咖〕（未 ，不）’而處理單元311、321、331、341復透過執行代理 f式的方式’接收驅動單元區15〇所傳送之控制指令訊 。並據以執行特定的處理流矛呈，以及產生並傳送處理結 果訊息。此種於虛擬機31、32、33、34安裝代理程式，並 透過相關的訊息仔列通訊介面作為控管裝置1〇〇與虛擬機 集群3〇之間的訊息傳遞模式，可令控管裝置⑽在不需擁 M432075 第100220008號專利申請案 101年3月13 6修正替換頁 有虛擬機31、32、33、34之登入權限的前提下，透過TCP/IP 協定直接與安裝於虛擬機31、32、33、34中的代理程式進 行通訊連結，從而提高虛擬機之安全性，並有效確保資料 的隱私性。 請再參閱第2圖，以瞭解本創作提供的控管系統10， 而第2圖係為本創作之應用於雲端計算之控管系統10的系 統架構圖。需先說明的是，圖中所示之第一控管裝置1〇〇’、 第二控管裝置200’及第三控管裝置300’之系統架構係相似 於前述的控管裝置100,僅額外增設了第一共享儲存單元 10卜第二共享儲存單元201及第三共享儲存單元301，故， 相同的系統架構不再於此贅述，且圖式中關於第二控管裝 置200’及第三控管裝置300’的細部架構亦予以省略。當 然，控管系統10可僅設置第一控管裝置100’及第二控管 裝置200’，或是設置更多的控管裝置。而如第2圖所示， 控管系統10還至少包括更新模組11、同步模組13及切換 模組15。 於本創作中，單備援的管理模式，即為將第一控管裝 置100’視為主機，而第二控管裝置200’則視為從機，以令 第二控管裝置200’提供第一控管裝置100’備援服務，而雙 備援之管理模式，即為將第二控管裝置200’及第三控管裝 置300’同時視為第一控管裝置100’之從機。 更新模組11係於偵測到第一收發單元110’接收到控 制指令訊息及/或處理結果訊息時，將第一收發單元110’ 接收之控制指令訊息及/或處理結果訊息更新至第一共享 第10〇22〇008號粵利申嘖案 儲存單元101。較佳地^ 卜3—月13 於财子第-收發單元110，t共享儲存單元⑻係具備用 處理住務儲存區（未圖 接收之控制指令訊息的第—待 所接收之處理結果訊自驗1存第—收發單元110, ⑻。同理，第二共享 已處理任務儲存區（未圖 二待處理任務健存區=子=、2二也可具備功能相同的第 (未圖示），而第三共享儲存:)第一已處理任務儲存區 第三待處理任務儲存301也可具備功能相同的 (未圖示）。 ΟΒ圖不）及第三已處理任務儲存區 同步模組13係於偵測到 結果訊息同步備份至第:共：：指广訊息及/或處理 2元加’以使第一 _ 儲 及卓三控管裝置300,中的資料妒伴牲^ 一好裝置2〇0, 置第三控管裝置300,的實施例二了同二二。當然，於未設 〜共享儲存同乂楔組13僅會將第 發生係於_第—控料置刚，的工作狀態 戈異常情形時，啟動第二控管裝置 切換=的第一控管裝置1〇0,來執行控管任務如 之、模、、且15該取代後之各控管褒置之狀態，其中，取代後 管置，將視為主機並執行與取代前之第一控 、置刚所執行之相同的控管任務，而若於雙備援的管 11 M432075 __ 第100220008號專利申請案 101年3月13扫修正替換頁 理模式中，取代發生異常情形的第一控管裝置1〇〇，的第二 控官裝置200則被視為主機，而第三控管裝置3〇()，係持續 被視為從機’並通知切換模組15該取代後之各控管裝置之 狀態，其中，取代後之第二控管裝置2〇〇，將執行與取代前 之第一控官裝置100’所執行之相同的控管任務。 當然，在雙備援之管理模式中，切換模組15可依據 預設之備援順序進行切換，而第一控管裝置1〇〇，、第二控 s裝置200及第二控官裝置3〇〇，能各自維護其具備的共享 儲存單元，並透過〜跳線機制（Heartbeat Protocol)來進行 通訊，當第一控管裝置1〇〇’發生故障或異常情形時可先 由鄰近的控管裝置接管其控管任務’或是由優先等級較高 之控管裝置接管其控管任務，並通知切換模組15各控管裝 置之狀態，且由切換模組15通知網路服務平台20置換後 之通訊接口。 此外’在另一實施態樣中，當切換模組15偵測第一 控管裝置100’的工作狀態恢復正常情形時，啟動第一控管 裝置100,以取代第二控管裝置200,以執行控管任務，並通 知切換模組15該取代後之各控管裝置之狀態。 綜上所述，本創作之應用於雲端計算之控管裝置，及 由多個應用於雲端計算之控管襞置所紕成之控管系統，係 可利用非同步的訊息傳遞方式來進行管理機制，以避免習 知同步化之通訊模式效能以及可靠度不佳的缺失。而透過 時間戮記的設計，可自動收回或停止超過預定時間的控制 指令訊息，以避免虛擬機回應中斷等異常情形發生。此外， 12 M432075 第100220008號專利申請案 101年3月13日修正替換頁 透過於虛擬機安裝的代理程式，在不需要擁有虛擬機之登 入權限的前提下，更可透過代理程式達到通訊連結的目 的，提高資料隱私性並降低主機登入資料於網路傳輸過程 中曝露的風險。再者，由於本創作提供的備援機制在視為 主機的控管裝置發生故障時，視為從機的控管裝置可即時 進行接管並完成無縫切換，故更能進一步提升整體的可靠 度。 然而，上述實施形態係用以例示性說明本創作之原理 鲁 及其功效，而非用於限制本創作。任何熟習此項技藝之人 士均可在不違背本創作之精神及範疇下，對上述實施形態 進行修改。因此本創作之權利保護範圍，應如後述之申請 專利範圍所列。 【圖式簡單說明】 第1圖係為本創作之應用於雲端計算之控管裝置的應 用架構圖；以及 I 第2圖係為本創作之應用於雲端計算之控管系統的系 統架構圖。 【主要元件符號說明】 100 控管裝置 110 收發單元 120 指令發送緩衝區 121 接收佇列 130 觸發單元 140 待處理佇列 13 M432075 第100220008號專利申請案 101年3月13 &修正替換頁 150 151 、 152 160 161 20 30 31 、 32 、 33 、 34 311、321、331、341 10 100， 110’ 130’ 150， 101 200， 201 300， 301 11 13 驅動單元區 驅動單元 指令回傳緩衝區 回傳佇列 網路服務平台 虛擬機集群 虛擬機 處理單元 控管系統 第一控管裝置 第一收發單元 第一觸發單元 第一驅動單元區 第一共享儲存單元 第二控管裝置 第二共享儲存單元 第三控管裝置 第三共享儲存單元 更新模組 同步模組 切換模組 15 1 M432075 公告本 第100220008號專利申請案 101年3月13日修正替換頁 新型專利說羽"書 (本說明書格式、順序，請勿任意更動，※記號部分請勿填寫） ※申請案號：ροηοοΰ? ※申請曰：⑽、Ι〇Ά 分類：<^6F (2006.01) 一、新型名稱：（中文/英文）M432075-,. Patent Application No. 100220008 * * Amended and replaced page on March 13, 101 V. New type description: [Technical field to which new type belongs] This creation is about a virtual machine (Virtual Machine; VM) control technology is particularly related to a virtual machine cluster control technology that combines asynchronous communication mechanisms. [Previous technology] Cloud computing is a service model that uses related software through a network connection, allowing clients to obtain the required resources after entering the cloud computing system. Many cloud computing systems are deployed today in virtual machine clusters. Specifically, virtualization methods can virtualize computing resources to cut the hardware's computing and storage capabilities. In this cloud architecture, how to monitor and manage virtual machine clusters through effective and stable mechanisms has become a key factor in the pros and cons of cloud computing systems. The currently available virtual machine management platform (VMHypervis〇r) φ can control and control the creation, deployment, and management of virtual machines. However, this virtual machine management platform is not conducive to an open cloud environment. Disadvantages. Second, the management process of virtual machines requires a certain response time, and the highly differentiated hybrid virtual architecture further increases the uncertainty of the control mechanism. If a synchronized communication mode is used, it will affect the effectiveness of the overall mechanism. And reliability. In addition, when deploying related virtual machines, it is often necessary to log in to the virtual machine to obtain the corresponding use rights, but this move makes it easy for users to have a degree of doubt about the security and data privacy of the virtual machine. In addition, in the existing cloud computing system, for the virtual machine 3 M432075 No. 100220008 patent application March 13, 2011, the revision control process of the replacement page cluster did not provide an effective control management backup mechanism, so Once a failure occurs, important control information is lost and a virtual machine cluster is forced to break. In summary, how to provide a brand-new control solution to overcome the various defects of the existing control technology for virtual machine clusters is an issue that all sectors are anxious to solve. [New content] In view of the shortcomings of the above-mentioned conventional technologies, one of the main purposes of this creation is to provide a control device that can overcome the lack of synchronous communication mechanisms, provide a backup mechanism, and can be fully applied to cloud computing. And control systems. In order to achieve the above and other objectives, this creation provides a control device applied to cloud computing, which is used to control a virtual machine cluster composed of a plurality of virtual machines in a cloud computing system. The control device applied to cloud computing includes : Transceiving unit is used to receive the control instruction message input by the network service platform and transmit the control instruction message to the receiving queue of the instruction sending buffer; triggering unit is used to receive and transmit the receiving queue The control instruction message in the control unit; and the drive unit area, which is used to receive the control instruction message transmitted by the trigger unit, and transmit the control instruction message to at least one of the virtual machines in the virtual machine cluster to drive the virtual machine to execute a specific A processing flow, and receiving a processing result message returned by the virtual machine after execution is completed, and transmitting the processing result message to a return queue of an instruction return buffer, and then returning the processing result message through the transceiver unit To the web service platform. M432075 * * Patent application No. 100220008 amended the replacement page on March 13, 2010 Secondly, "This creation also provides a control system for cloud computing" includes: the first control device, which is used for receiving The control instruction message which is rotated by the network service platform and transmits the control instruction message to the first receiving and sending unit of the receiving queue of the instruction sending buffer, and the first receiving and transmitting unit for receiving and transmitting the control instruction message in the receiving queue. A trigger unit for receiving the control instruction message transmitted by the first trigger unit and transmitting the control instruction message to at least one of the virtual machine groups to drive the virtual machine to execute a specific processing flow and receive the virtual machine upon completion of execution And then return the processing result message to the return queue of the instruction return buffer, and then return the processing result message to the first of the network service platform via the first transceiver unit. The drive unit area, wherein the first control device further has a first shared storage unit; the second control device has a function for receiving network services The control instruction message input by the station and transmits the control instruction message to the second receiving and sending unit of the receiving queue of the instruction sending buffer; the second triggering unit for receiving and transmitting the control instruction message in the receiving queue; For receiving the control instruction message transmitted by the second trigger unit and transmitting the control instruction message to at least one of the virtual machine group to drive the virtual machine to execute a specific processing flow and receive the virtual machine after the execution is completed. The returned processing result message and the processing result message are sent to the return queue of the instruction return buffer to further return the processing result message to the second drive unit area of the network service platform via the second transceiver unit. Wherein, the first control device further has a second shared storage unit; the update module is configured to detect the first transceiver unit when receiving the control instruction message or the processing result message, and then the first transceiver unit Received the control instruction message 5 M432075-Patent No. 100220008 Patent Application March 13, 2011 & amended replacement page or the result of this process update The first shared storage unit; a synchronization module, which stores the first shared storage unit when it is detected that the update module has updated the control instruction message or the processing result message to the first shared storage unit. The control command message or processing result message is backed up to the second shared storage unit synchronously; and the switching module is used to start the second control device to replace the first control device when an abnormal working state is detected. The first control device performs a control task. Compared with the conventional technology, the control device of the present invention combines a non-synchronous message delivery mechanism to overcome the lack and bottleneck of the conventional technology caused by the use of the synchronous message delivery mechanism. In addition, the control system provided by this creation can also seamlessly take over other control devices that are considered as slaves when the control devices that are considered as masters fail or are abnormal. The control device regarded as the master and other control devices regarded as the slave will synchronize the backup data, so not only can prevent data loss, but also effectively improve the overall reliability. [Embodiment] The following describes the technical content of this creation with specific specific implementation forms. Those familiar with this technology can easily understand other advantages and effects of this creation from the content disclosed in this manual, and can also use other different specific Implementation forms are implemented or applied. Please refer to FIG. 1 first to understand the cloud computing control device provided in this creation, and FIG. 1 is an application architecture diagram of the cloud computing control device 100 in this creation. As shown in the figure, the control device 100 is used to control and manage the M432075 __ in the cloud computing system. Patent application No. 100220008, March 13, 2001, revises the virtual machine cluster 30 of the replacement page, and the virtual machine cluster 30 may include a virtual machine 31, a virtual machine 32, a virtual machine 33, and / or a virtual machine 34, and the control device 100 also exercises control and control over the virtual machines 31, 32, 33, and 34 as new or deployed. At the same time, the control device 100 includes at least a transceiver unit 110, a trigger unit 130, and a drive unit area 150. The transceiver unit 110 is used for receiving a control instruction message input by the user from the network service platform 20, and transmitting the received control instruction message to the receiving queue 121 in the instruction sending buffer 120. Lu Receiving Queue 121 adopts the control mechanism of Message Queue (MQ). In addition, the transceiver unit 110 is configured to receive a processing result message returned by the instruction return buffer 160. In this embodiment, the network service platform 20 may be an portal website providing cloud services, and the transceiver unit 110 may communicate with the network service platform 20 by using the TCP / IP protocol. Among them, by using a transmission layer based on the OSI architecture, Can more effectively integrate different system environments. • The trigger unit 130 is used to transmit the control instruction message in the receiving queue 121 to the driving unit area 150, and the driving unit area 150 is used to transmit the control instruction message to the virtual machine cluster 3 through the driving unit (such as 151, 152). At least one of the virtual machines 31, 32, 33, and 34 is configured to drive the virtual machines 31, 32, 33, and 34 that have received the control instruction message to execute a specific processing flow. In one embodiment, the 'trigger unit 130 may first transmit the control instruction message in the receiving queue 121 to the pending queue 140, and then wait for palatable timing for transmission, wherein the pending queue 140 uses a message Control mechanism M432075-Patent Application No. 100220008 March 13, 2011 & Amendment Replacement Page In another embodiment, the type of the control instruction message is different, and the trigger unit 130 can first identify the content of the control instruction message and Classification, such as checking the CPU status, checking disk usage, or checking the VM cache memory of the virtual machine; and after the identification and classification are completed, the trigger unit 130 sends control instruction messages of different types to the waiting The processing queue 140 records the category information. This kind of special-purpose design can further improve the processing efficiency. In yet another embodiment, the control instruction message sent to the pending queue 140 may be appended with a relevant time stamp, thereby triggering the unit 130 to determine the control instruction message in the pending queue 140. When the time stamp of the elapsed time exceeds the predetermined time, the transmission operation of the control instruction message that has exceeded the predetermined time is automatically cancelled or recovered to avoid subsequent errors or interruptions. In an embodiment, the driving unit area 150 has a driving unit 151 and a driving unit 152, and the number of the driving units 151 and 152 included in the driving unit area 150 is adjustable and is not limited to the number in the drawing. The drive unit area 150 is used to receive the control instruction message of the queue 140 to be processed, and is transmitted by the corresponding drive units 151 and 152 to the corresponding virtual machines 31, 32, 33, and 34 according to the type of the control instruction message. For example, the control instruction message for checking the CPU status is transmitted from the drive unit 151 to the virtual machine 31 to drive the virtual machine 31 to check the CPU status; the control instruction message for checking the disk usage is transmitted from the drive unit 152 to the virtual machine 32 to drive the virtual machine. The machine 32 checks the disk usage. When the virtual machines 3, 32, 33, and 34 are completed in a specific processing flow, the driver M432075 corresponding to the virtual machines 31, 32, 33, and 34 in the drive unit area 150 is __ • · · Patent Application No. 100220008_ On March 13, 101, it was amended that the replacement pager units 151 and 152 received the processing result information returned by the corresponding virtual machines 31, 32, 33, and 34. For example, if the drive unit 151 corresponds to the virtual machine 3 ′, the drive unit 151 receives the processing result message returned by the virtual machine 31; if the drive unit 152 corresponds to the virtual machine 32, the drive unit 152 receives the virtual machine 32. The processing result message returned. When the driving units 151 and 152 in the driving unit area 150 receive the corresponding processing result message, the driving units 151 and 152 transmit the processing result message to the return queue 161 in the instruction return buffer 160, and then • The processing result message is transmitted back to the network service platform 2 ^ 0 'through the transceiver unit 110 in order to be displayed by the network service platform 20, and relevant processing measures are performed. Among them, the return queue 161 uses messages. Queue control mechanism. With this, the user can remotely process and control the virtual machines 31, 32, 33, and 34 of the virtual machine cluster t through the network service platform 2G. It should be noted that the virtual machines 31, 32, 33, and 34 may also be provided with processing units 311, 321, 331, and 341, respectively, for receiving control command messages from the driving unit area 150 and ^, and execute accordingly. Specific processing flow, and then send the processing result message. In addition, in the virtual machine 3%, 33, and 34, the relevant agent program (mq ton coffee) (not, no) can be loaded in advance, and the processing units 311, 321, 331, and 341 can be executed again. The agent f-type method 'receives the control command message transmitted by the drive unit area 15. Based on this, it executes a specific processing flow, and generates and transmits a processing result message. This type of virtual machine 31, 32, 33, 34 Install the agent program and use the relevant message queue communication interface as the message transmission mode between the control device 100 and the virtual machine cluster 30, which can make the control device not need to hold the M432075 patent application No. 100220008 March 13, 2011 6 Corrected that the replacement page has the login rights of virtual machines 31, 32, 33, and 34, and communicates directly with the agents installed in virtual machines 31, 32, 33, and 34 through the TCP / IP protocol. Link to improve the security of the virtual machine and effectively ensure the privacy of the data. Please refer to Figure 2 again to understand the control system 10 provided by this creation, and Figure 2 is the application of this creation to cloud computing Department of Control System 10 Architecture diagram. It should be noted that the system architecture of the first control device 100 ′, the second control device 200 ′ and the third control device 300 ′ shown in the figure is similar to the aforementioned control device. 100. Only the first shared storage unit 10, the second shared storage unit 201, and the third shared storage unit 301 are additionally added. Therefore, the same system architecture is not repeated here, and the second control device 200 is shown in the drawing. The detailed structure of the 'and third control device 300' is also omitted. Of course, the control system 10 may be provided with only the first control device 100 'and the second control device 200', or more control devices may be provided. As shown in Figure 2, the control system 10 also includes at least an update module 11, a synchronization module 13, and a switch module 15. In this creation, the single backup management mode is to control the first control The device 100 ′ is regarded as a master, and the second control device 200 ′ is regarded as a slave, so that the second control device 200 ′ provides a backup service of the first control device 100 ′, and a management mode of dual backup, That is, the second control device 200 'and the third control device 300' are regarded as the first The slave of the control device 100 '. The update module 11 detects the control instruction message and / or the processing result message received by the first transceiver unit 110', and controls the control instruction message received by the first transceiver unit 110 'and / Or the processing result message is updated to the first shared storage unit 101 of the Guangdong Proposal No. 10022002. Preferably ^ March 3-March 13 in the Caizi No.-transceiver unit 110, t shared storage unit is not Equipped with a processing house storage area (not shown in the control instruction message received in the first-to-be-received processing result self-inspection 1 storage-receiving unit 110, ⑻. Similarly, the second shared processed task storage area (not Figure 2 To-be-processed task storage area = sub =, 22 can also have the same function (not shown), and the third shared storage :) The first processed task storage area The third pending task storage 301 is also available Equipped with the same function (not shown). 〇Β 图 不) and the third processed task storage area synchronization module 13 is synchronously backed up to the detected result message to the total ::: means the message and / or process 2 yuan plus' to make the first _ Chu and Zhuo San The information in the control device 300, is jealous of the companion ^ A good device 2000, the third control device 300, the second embodiment is the same as the second two. Of course, when the shared storage and wedge group 13 is not set, the first control device will be switched to the first control device when the working condition of the first control device is rigid, and the second control device is switched. 100, to perform the control tasks such as the model, the model, and the state of each control device after the replacement. Among them, the replacement control device will be regarded as the host and the first control before the replacement, Zhigang performs the same control tasks, but if the dual backup management is 11 M432075 __ No. 100220008 Patent Application March 31, 2011, the correction and replacement page mode replaces the first control that occurs when an abnormal situation occurs. The second controller device 200 of device 100, is regarded as the master, and the third controller device 30 (), is continuously regarded as the slave ', and informs the switching module 15 of the replaced controllers. The state of the device. The second control device 200 after the replacement will perform the same control tasks as the first control device 100 'before the replacement. Of course, in the dual backup management mode, the switching module 15 can switch according to a preset backup sequence, and the first control device 100, the second control device 200, and the second control device 3 〇〇, can maintain its own shared storage unit, and communicate through the ~ Jumper mechanism (Heartbeat Protocol), when the first control device 100 ′ failure or abnormal situation can be first controlled by the adjacent control The device takes over its control task 'or a higher priority control device takes over its control task, and notifies the switching module 15 of the status of each control device, and the switching module 15 notifies the network service platform 20 to replace Later communication interface. In addition, in another embodiment, when the working state of the first control device 100 is detected by the switching module 15, the first control device 100 is activated to replace the second control device 200 to Perform control tasks, and notify the switching module 15 of the status of each control device after the replacement. To sum up, the control device applied to cloud computing and the control system formed by multiple control devices applied to cloud computing in this creation can be managed by asynchronous message transmission. Mechanism to avoid the lack of conventional synchronous communication mode performance and poor reliability. And through the design of time-keeping, it can automatically retract or stop the control instruction message that exceeds the predetermined time to avoid abnormal situations such as interruption of the virtual machine response. In addition, 12 M432075 Patent Application No. 100220008 on March 13, 101 amended the replacement page through the agent installed in the virtual machine. Without the need to have the login privileges of the virtual machine, the communication link can be achieved through the agent. The purpose is to improve the privacy of data and reduce the risk of exposure of host login data during network transmission. In addition, since the backup mechanism provided by this creation can be taken over as a slave when the control device of the master fails, the control device regarded as a slave can take over and complete seamless switching, which can further improve the overall reliability. . However, the above embodiments are used to exemplify the principle and effectiveness of this creation, not to limit this creation. Anyone who is familiar with this skill can modify the above implementation form without departing from the spirit and scope of this creation. Therefore, the scope of protection of the rights of this creation shall be as listed in the scope of patent application mentioned later. [Schematic description] Figure 1 is the application architecture diagram of the cloud computing control device created by the author; and Figure 2 is the system architecture diagram of the cloud computing control system created by the author. [Description of main component symbols] 100 Control device 110 Transceiving unit 120 Instruction sending buffer 121 Receiving queue 130 Trigger unit 140 Pending queue 13 M432075 Patent application No. 100220008 March 13, 2011 & Correction replacement page 150 151 , 152 160 161 20 30 31, 32, 33, 34 311, 321, 331, 341 10 100, 110 '130' 150, 101 200, 201 300, 301 11 13 Drive unit area Drive unit command return buffer return Queued network service platform virtual machine cluster virtual machine processing unit control system first control device first transceiver unit first trigger unit first drive unit area first shared storage unit second control device second shared storage unit Three control devices, third shared storage unit update module, synchronization module switch module 15 1 M432075 Notice No. 100220008 Patent Application March 13, 101 Revised Replacement Page New Patent Saying Feather " book (this specification format, Please do not change the order arbitrarily, ※ Please do not fill in the marked part) ※ Application number: ροηοοΰ? ※ Application: ⑽ 、 Ι〇Ά Classification: & l t; ^ 6F (2006.01) I. New name: (Chinese / English)
應用於雲端計算之控管裝置及控管系統 MONITORING DEVICE AND MONITORING SYSTEM APPLICABLE TO CLOUD ALGORITHM 二、中文新型摘要： 一種應用於雲端計算之控管裝置及控管系統，係採用 訊息佇列控制機制，進而以非同步的方式傳遞控制指令訊 息至虛擬機’並接收虛擬機回傳的處理結果訊息，藉此提 升系統整體效能與可靠性。再者，本創作的控管系統包括 多個能同步備份的控管裝置，所以還能於視為主機的控管 裝置發生故障或異常時，啟動視為從機的控管裝置接管原 先針對虛擬機之控管任務，藉以提升整體之穩定性。 • 三、英文新型摘要：MONITORING DEVICE AND MONITORING SYSTEM APPLICABLE TO CLOUD ALGORITHM applied to cloud computing II. Chinese new abstract: A control device and control system applied to cloud computing, which uses a message queue control mechanism, and further Pass the control instruction message to the virtual machine in an asynchronous manner and receive the processing result message returned by the virtual machine, thereby improving the overall system performance and reliability. In addition, the control system of this creation includes multiple control devices that can be backed up synchronously, so when the control device that is regarded as the master fails or is abnormal, the control device that is regarded as the slave can take over the original control for the virtual Machine control tasks to improve overall stability. • 3. New English Abstracts:
The invention provides a monitoring device and a monitoring system applicable to a cloud algorithm, in which a message standing line control mechanism is employed to transmit control command messages to a virtual machine in a non-synchronic manner and to receive the feedback of processing results transmitted from the virtual machine, thereby enhancing the overall system efficacy and reliability. Further, the monitoring system comprises multiple synchronically performing backup monitoring devices, which enables slave devices to take over the original monitoring missions performed by the master device whenever any breakdown or irregularity occurs to thus increase reliability of performance. 1 M432075 第100220008號專利申請案 101年3月13日修正替換頁 六、申請專利範圍： 1. 一種應用於雲端計算之控管裝置，用於控管雲端計算系 統中由複數虛擬機組成的虛擬機集群，該應用於雲端計 算之控管裝置係包括： 收發單元，係用於接收由網路服務平台所輸入之控 制指令訊息，並將該控制指令訊息傳送至指令發送缓衝 區之接收彳宁列； 觸發單元，係用於接收與傳送該接收佇列中之控制 指令訊息；以及 驅動單元區，係用於接收該觸發單元所傳送之控制 指令訊息，而傳送該控制指令訊息至該虛擬機群中之至 少一者，以驅動該虛擬機執行特定的處理流程，以及接 收該虛擬機於執行完成後所回傳之處理結果訊息，並將 該處理結果訊息傳送至指令回傳緩衝區之回傳佇列，進 而將該處理結果訊息經由該收發單元回傳至該網路服 務平台。 2. 如申請專利範圍第1項所述之應用於雲端計算之控管 裝置，其中，該虛擬機係具備處理單元，用於接收該驅 動單元區所傳送之控制指令訊息，並據以執行該特定的 處理流程，進而傳送該處理結果訊息。 3. 如申請專利範圍第2項所述之應用於雲端計算之控管 裝置，其中，該虛擬機係裝載有代理程式，而該處理單 元係透過執行該代理程式的方式執行該特定的處理流 程，進而傳送該處理結果訊息。 15 第100220008號專利申培金 ωι年3月13 a修正替 4.如申請專利範圍m】s 裝置，JL 固第1項所述之應用於雲端計算之控管 指入=^中該觸發單元復用於將該接收仔列中的控制 _ α 5〜傳送至待處理佇列，進而等待傳送至該驅動單 區。 事申°月2利乾圍第4項所述之應用於雲端計算之控管 八中，該驅動單元區係包含至少一個驅動單元， 且該觸發單元復用於針對該控制指令訊息的内容進行 3別並分類’進而將不關別之該控制指令訊息經過該The invention provides a monitoring device and a monitoring system applicable to a cloud algorithm, in which a message standing line control mechanism is employed to transmit control command messages to a virtual machine in a non-synchronic manner and to receive the feedback of processing results transmitted from the virtual machine, thereby enhancing the overall system efficacy and reliability. Further, the monitoring system includes multiple synchronically performing backup monitoring devices, which enables slave devices to take over the original monitoring missions performed by the master device whenever any breakdown or irregularity occurs to Thus increased reliability of performance. 1 M432075 Patent Application No. 100220008 Amended and Replaced on March 13, 101 Page 6. Application Scope of Patent: 1. A control device applied to cloud computing, which is used to control cloud computing systems. A virtual machine cluster composed of a plurality of virtual machines. The control device applied to cloud computing includes: a transceiver unit, which is used for Receive the control instruction message input by the network service platform, and transmit the control instruction message to the receiving queue of the instruction sending buffer; the trigger unit is used to receive and transmit the control instruction message in the receiving queue And the drive unit area is used to receive the control instruction message transmitted by the triggering unit, and transmit the control instruction message to at least one of the virtual machine group to drive the virtual machine to execute a specific processing flow, and receive The processing result message returned by the virtual machine after execution is completed, and the processing result message is sent to a return queue of an instruction return buffer, and then the processing result message is returned to the network via the transceiver unit. Service Platform. 2. The control device for cloud computing as described in item 1 of the scope of patent application, wherein the virtual machine is provided with a processing unit for receiving a control instruction message transmitted by the driving unit area and executing the A specific processing flow, and then the processing result message is transmitted. 3. The control device for cloud computing described in item 2 of the scope of patent application, wherein the virtual machine is loaded with an agent program, and the processing unit executes the specific processing flow by executing the agent program , And then send the processing result message. 15 Patent No. 100220008 patent application March 13th year a Correction 4. According to the scope of the patent application m] s device, the control of the cloud computing application described in the first item of JL = = ^ This trigger unit is reused The control _ α 5 ~ in the receiving queue is transmitted to the queue to be processed, and then waiting to be transmitted to the driving single area. As mentioned above, the control unit No. 8 applied to cloud computing as described in item 4 of Liganwei, the drive unit area includes at least one drive unit, and the trigger unit is used to perform the content of the control instruction message. 3 Classification and classification '
y處理彳τ列傳送至該驅動單元區之驅動單元中。 申明專利範圍第4項所述之應用於雲端計算之控管 置/、中，傳送至該待處理彳宁列令的控制指令訊息係 設置有時間戳記。 〜” •如申請專利範圍第6項所述之應用於雲端計算之控管 裝置，其中，該觸發單元復用於在該待處理佇列中的該 控制指令訊息的該時間戳記超過預定時間時，取消該控 制指令訊息之傳送。The y processing sequence is transmitted to the driving unit in the driving unit area. The control instruction / applied to the cloud computing described in item 4 of the declaration patent scope, and the control instruction message transmitted to the pending Suining order is set with a time stamp. ~ "• The control device for cloud computing as described in item 6 of the scope of patent application, wherein the trigger unit is reused when the time stamp of the control instruction message in the pending queue exceeds a predetermined time To cancel the transmission of the control instruction message.
8.如申請專利範圍第1項所述之應用於雲端計算之控管 裝置，其中，該收發單元係利用TCP/IP協定與該網路 服務平台連結通訊。 9. 一種應用於雲端計算之控管系統，係包括： 第一控管裝置，係具有用於接收由網路服務平台所 輸入之控制指令訊息並將該控制指令訊息.傳送至指令 發送緩衝區之接收佇列的第一收發單元、用於接收與傳 送該接收仔列中之控制指令訊息的第一觸發單元、用於 M432075 . 第100220008號專利申請案 - | 1〇丨年3月13日修正替換頁 收該第一觸發單元所傳送之控制指令訊息而傳送該 控制私令訊息至虛擬機群中之至少一者以驅動虛擬機 執行特定的處理流程及接收該虛擬機於執行完成後所 回傳之處理結果訊息並將該處理結果訊息傳送至指令 回，緩衝區之回傳佇列進而將該處理結果訊息經由該 第一收發單元回傳至該網路服務平台的第一驅動單元 區’其中，該第一控管裝置另具有第一共享儲存單元；8. The control device for cloud computing as described in item 1 of the scope of patent application, wherein the transceiver unit uses TCP / IP protocol to communicate with the network service platform. 9. A control and management system applied to cloud computing, comprising: a first control and control device having a control instruction message input by a network service platform and transmitting the control instruction message to an instruction sending buffer The first transceiver unit of the receiving queue, the first trigger unit for receiving and transmitting the control instruction message in the receiving queue, for M432075. Patent application No. 100220008-| March 13th, 2010 Modify the replacement page to receive the control instruction message sent by the first trigger unit and send the control private order message to at least one of the virtual machine groups to drive the virtual machine to execute a specific processing flow and receive the virtual machine after the execution is completed. The returned processing result message and the processing result message are transmitted to the instruction return, and the buffer return queue is further used to return the processing result message to the first driving unit area of the network service platform through the first transceiver unit. 'Wherein, the first control device further has a first shared storage unit;
第二控管裝置，係具有用於接收由網路服務平台所 輸入之控制才曰令訊息並將該控制指令訊息傳送至指令 發送緩衝區之接收㈣的第二收發單Tt、用於接收與傳 送該接收㈣中之控制指令訊息的第二觸發單元、用於 接收該第二觸發單元所傳送之控制指令訊息而傳送該 控制指令訊息至該虛擬機群中之至少一者以驅動該虛 擬機執行待定的處理流程及接收該虛擬機於執行完成 後所回傳之處理結果訊息並將該處理結果訊息傳送至 指令回傳緩衝區之回傳仵列進而將該處理結果訊息經 由該第二枚發單元回傳至該網路服務平台的第二驅動 單兀區，其中，該第二控管裝置另具有第二共享儲存單 元； 更新f组，係Μ·"第—收發單元接收到該控 制指，=測到該第一收發單元接收到該處理結 果H將該第-收發單元純之該 該第-收發單元接收之處理結果訊息更新至二ί 享儲存單元； 乐 ^ 17 M432075 同步模組，係於偵測 訊息或該處理結果吨阜更 斤棋組已將該控制指令 時’將該第-共享健存;元中：的第二儲t單元 理結果訊息同步備份至令- '^曰々訊心或處 切換模組，係於偵测到：:享:存單元;以及 發生異常時，啟動該第二控;=控管裝置之工作狀態 置執行控管任務。 取代該第-控管裝 ίο.如申請專利範圍第9項所 一 系統，其中，該第一共享 /，；农端计异之控管 一收發單元所接收^早元料制於儲存該第 訊自的第二 收發單元所接收之處理結果 I的弟-已處理任務儲存區 :::::存該第二收發單元所接收== 、、处理任務儲存區、及用於儲存該第二收發單 ’所接收之處理結果訊息的第二已處理任務儲存區。The second control device is a second transceiver Tt for receiving a control command message input from the network service platform and transmitting the control command message to the receiving unit of the command sending buffer, for receiving and Transmitting a second trigger unit for receiving the control instruction message in the frame, for receiving the control instruction message transmitted by the second trigger unit and transmitting the control instruction message to at least one of the virtual machine groups to drive the virtual machine Execute the pending processing flow and receive the processing result message returned by the virtual machine after the execution is completed and send the processing result message to the return queue of the instruction return buffer to further pass the processing result message through the second The sending unit sends back to the second driving unit area of the network service platform, wherein the second control device further has a second shared storage unit; group f is updated, and the M-&#; transmission unit receives the Control means, = it is detected that the first transceiver unit receives the processing result H updates the first transceiver unit from the processing result message received by the first transceiver unit to two Le ^ 17 M432075 Synchronous module, which is used to detect the message or the result of the processing. When the control group has sent the control instruction, the first-shared storage; the second storage t unit Synchronized backup of the processing result message to the order-'^ 々 々 々Xin Xun Xing Xing switching module, based on the detection of :: share: storage unit; and when an abnormality occurs, start the second control; = working state of the control device Set up control tasks. Instead of the first-control device, as in the system of the 9th patent application scope, wherein the first sharing / receiving unit received by the control-receiving unit of the farm-side meter control system is stored in the first storage device. Brother of the processing result I received by the second transceiver unit-processed task storage area ::::: Stores the received by the second transceiver unit ==, processing task storage area, and is used to store the second The second processed task storage area of the processing result message received by the sender and receiver.
1S1S