Kafka你真的瞭解嗎？

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"​ Apache kafka 是一個分佈式的基於push-subscribe的消息系統，它具備快速、可擴展和可持久化的特點。它的最大的特性就是可以實時的處理大量數據以滿足各種需求場景：比如基於hadoop的批處理系統、低延遲的實時系統、Storm/Spark流式處理引擎。主要有以下特性：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Consumergroup：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"各個consumer可以組成一個group，每個partition消息只能被group內的一個consumer消費，如果一個消息可以被多個consumer消費的話，那麼這些consumer必須在不同的group","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"消息狀態：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"消息的狀態被保存在consumer中，broker不會關心哪個消息被消費了、被誰消費了，只記錄一個offset值","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"持久化：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Kafka中會把消息持久化到本地磁盤，支持數據備份防止數據丟失並保持極高的效率","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"高吞吐量：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Kafka支持以批量發送，以提高push效率。每秒可以處理幾十萬條消息","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"推拉模式：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Kafka中的producer和consumer採用的是push-and-pull模式，即Producer只管向broker push消息，consumer只管從broker pull消息，兩者對消息的生產和消費是異步的","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"同步異步：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Producer採用異步push方式，極大提高Kafka系統的吞吐率","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"分區機制：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Kafka的broker端支持消息分區，producer可以決定把消息發到哪個partition，在partition中消息是有序的，一個topic可以有多個partition","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":14}},{"type":"strong","attrs":{}}],"text":"一、使用場景","attrs":{}}]},{"type":"bulletedlist","content":[{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"用Kafka可以收集各種服務的log，通過kafka以統一接口的方式開放給各種consumer","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"解耦和生產者和消費者、緩存消息等","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"收集用戶行爲信息發佈到kafka的topic中，然後consumer通過訂閱這些topic來做實時的監控分析，或者離線分析和挖掘","attrs":{}}]}],"attrs":{}}],"attrs":{}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":14}},{"type":"strong","attrs":{}}],"text":"二、架構組成","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"    Kafka中發佈訂閱的對象是topic,可以爲每類數據創建一個topic，把向topic發佈消息的稱作producer，從topic消費消息的稱作consumer。Producers和consumers可以同時從多個topic讀寫數據。一個kafka集羣由一個或多個broker服務器組成，它負責持久化和備份具體的kafka消息。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Topic：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Topic是一個邏輯的歸集基礎單元，實際上是由一些partition logs組成，一個Topic可以設定多個partition，一個message僅屬於一個topic，其結構如下圖：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/56/560e7c4b597117fae8c54617f419977e.png","alt":null,"title":null,"style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 每個partition中的消息都是有序的，生產的消息被不斷追加到partition log上，其中的每一個消息都被標識了唯一的offset值。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Broker：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Broker是當前服務器上運行的kafka的實例，在kafka集羣中每一個broker對應一個唯一的ID，即brokerId。Broker只關心數據存儲，不關心消息的狀態。Kafka的消息保存一定時間(通常7天)後會刪除。Consumer可以rewind back到任意位置重新進行消費，當consumer故障時可以選擇最小的offset進行重新讀消費","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Message：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Message是消息的基礎通信單元，由offset、messageSize和data組成","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Producer：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka集羣中的每個broker都可以響應producer的請求並返回topic的元信息，包括哪些機器是存活的，topic的leader partition都在哪，哪些leader partition是可以被訪問的。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Producer自行控制消息被推送到哪些partition，實現的方式有隨機分配、分區算法，Producer將消息在內存中累計到一定數量後作爲一個batch發送請求，可以設置爲累計的消息的數量(如500條)、累計的時間間隔(如100ms)或者累計的數據大小(默認16KB)，以Batch的方式推送數據可以極大的提高處理效率。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Producer可以異步並行的向kafka發送消息，但是通常producer在發送完消息之後會得到一個future響應，返回offset值或發送過程中遇到的錯誤，可設置acks決定producer要求leader partition 收到確認的副本個數:","attrs":{}}]},{"type":"bulletedlist","content":[{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Acks=0 表示producer不會等待broker的響應，會導致數據丟失，但會得到最大的系統吞吐量","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Acks=1 表示producer會等待leader partition的確認，這樣會有更好的可靠性","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Acks=-1 表示producer會收到所有副本partition的確認，可得到最高的可靠性保證","attrs":{}}]}],"attrs":{}}],"attrs":{}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Kafka沒有限定單個消息的大小，但推薦大小不要超過1MB，通常一般消息大小都在1~10kB。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Consumer：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 每個consumer屬於一個consumer group，一個分區的消息只能被同一個group中的一個consumer消費，每一個consumer實例可消費多個partition的數據，並且每個partition的消息都要被不同的group消費。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/3d/3d08a74fbc65dbb34452430f8453789a.png","alt":null,"title":null,"style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":14}},{"type":"strong","attrs":{}}],"text":"三、核心特性","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"Relications：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 副本機制大大提高了Kafka集羣的可靠性、穩定性，Kafka允許集羣中的節點掛掉後而不影響整個集羣工作，一個副本數量爲n的集羣允許n-1個節點失敗。在所有副本節點中，有一個節點作爲leader節點，這個節點保存了其它備份節點列表，並維持各個副本間的狀態同步。如下圖爲Kafka的副本機制：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/33/33a797026955a8c3cdcd8ca8704dae8a.png","alt":null,"title":null,"style":null,"href":null,"fromPaste":true,"pastePass":true}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka允許爲每個topic設置副本數量，副本數量決定了有幾個broker來存放寫入的數據。一般推薦副本數量至少爲2，這樣就可以保證增減、重啓機器時不會影響到數據消費。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"分區：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Partitions的設計目的根本原因是kafka基於文件存儲，通過分區可以將日誌內容分散到多個broker上避免文件大小達到單機磁盤的上限，每個partiton都會被當前broker保存,可以將一個topic切分任意多個partitions，越多的partitions意味着可以有更多的consumer並行消費,有效提升併發消費的能力。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"壓縮：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka支持批量發送消息，在此基礎上還可對消息集合進行壓縮，Producer端可以通過GZIP或Snappy格式對消息集合進行壓縮，壓縮的好處就是減少傳輸的數據量，減輕對網絡傳輸的壓力，在對大數據處理上瓶頸往往體現在網絡上而不是CPU。Kafka在消息頭部添加了一個壓縮屬性標識是否壓縮，這個屬性的後兩位表示消息壓縮採用的編碼，如果後兩位爲0則表示消息未被壓縮。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"消息可靠性：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 在消息系統中，保證消息在生產和消費過程中的可靠性是十分重要的，在實際消息傳遞過程中，可能會出現如下三種情況：","attrs":{}}]},{"type":"bulletedlist","content":[{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"一個消息發送失敗","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"一個消息被髮送多次","attrs":{}}]}],"attrs":{}},{"type":"listitem","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"最理想的情況：exactly-once ,一個消息發送成功僅發送了一次","attrs":{}}]}],"attrs":{}}],"attrs":{}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"  在實際中因爲網絡問題，producer或consumer在生產和消費過程中都有可能失敗。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 從producer端看：當一個消息被髮送後，producer會等待broker成功接收到消息的反饋，如果消息在途中丟失或是其中一個broker掛掉，Producer會重新發送","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 從consumer端看：Broker端記錄了partition中的一個offset值，這個值指向consumer下一個即將消費message。當consumer收到了消息，但在處理過程中掛掉，此時consumer可以通過offset值重新找到上一個消息重新消費。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"持久化：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka使用文件存儲消息，直接決定了kafka在性能上嚴重依賴文件系統的本身特性。因爲kafka是對日誌文件進行append操作，因此磁盤檢索的開支是較小的，另外利用OS Page Cache刷盤機制減少磁盤寫入的次數，減少磁盤IO調用的次數。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka高度依賴文件系統來存儲和緩存消息，一般認爲磁盤是緩慢的，其實磁盤遠比想象的要快或者慢，這決定於如何使用磁盤。一個和磁盤性能有關的關鍵事實是：磁盤驅動器的吞吐量跟尋道延遲是相背離的，也就是說線性寫的速度遠遠大於隨機寫。比如：在一個7200rpm SATA RAID-5 的磁盤陣列上線性寫的速度大概是600M/秒，但是隨機寫的速度只有100K/秒，兩者相差將近6000倍，並且對磁盤的線性讀在有些情況下可以比內存的隨機訪問要快一些。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" ","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong","attrs":{}}],"text":"集羣：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" Kafka集羣的大小需要根據硬件的配置、生產者和消費者的併發數量、數據的副本個數、數據的保存時長綜合確定。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 假設：W：每秒寫多少MB     R：副本數    C：consumer的數量","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 一般來說kafka集羣的瓶頸在於網絡和磁盤吞吐量，對於每條消息每個副本都要寫一遍，所以整體寫的速度是W*R。讀數據主要是集羣內部各個副本從leader同步消息和consumer讀，所以讀的速率是(R-1)*W，同時consumer讀的速度是C*W，因此：","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"  Write：W*R        Read：（R-1）*W+C*W","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":" 其次可以在讀的時候緩存部分數據來減少IO操作，如果一個集羣有M MB內存，寫的速度是W MB/sec，則允許M/(W*R) 秒的寫可以被緩存。如果集羣有32GB=32768MB內存，寫的速度是50MB/s的話，則可以至少緩存10分鐘的數據。","attrs":{}}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"horizontalrule","attrs":{}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":"center","origin":null},"content":[{"type":"text","marks":[{"type":"size","attrs":{"size":14}}],"text":"更多文章請加入公衆號","attrs":{}}]},{"type":"image","attrs":{"src":"https://static001.geekbang.org/infoq/81/81efbe4ffd49eb01ec73616667890bab.jpeg","alt":null,"title":"","style":[{"key":"width","value":"50%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}}]}