<?xml version="1.0" encoding="UTF-8" standalone="no"?><xmap-content xmlns="urn:xmind:xmap:xmlns:content:2.0" xmlns:fo="http://www.w3.org/1999/XSL/Format" xmlns:svg="http://www.w3.org/2000/svg" xmlns:xhtml="http://www.w3.org/1999/xhtml" xmlns:xlink="http://www.w3.org/1999/xlink" modified-by="ganling" timestamp="1542074803840" version="2.0"><sheet id="5gimpj6r68ai2llsfdlm0l87t0" modified-by="ganling" theme="6rk3ges1df2vvfvu7ek61ff89o" timestamp="1542074803840"><topic id="2vkscnetvja10ilvnun19mqq1u" modified-by="Administrator" structure-class="org.xmind.ui.map.unbalanced" timestamp="1516073856960"><title>SparkCore</title><children><topics type="attached"><topic id="4f7u44fp19t09vmbj2nglrj90b" modified-by="ganling" timestamp="1542073656646"><title>前言</title><children><topics type="attached"><topic id="4diakcgg7q2njbbojf3o57ubap" modified-by="ganling" timestamp="1542073657625"><title>对比MapReduce</title><children><topics type="attached"><topic id="088rbc99md6ohscnfdmb5kuj7t" modified-by="ganling" timestamp="1542073658862"><title>分布式计算框架</title><children><topics type="attached"><topic id="5f84c0b89dee80m3iap3hq0g1p" modified-by="ganling" timestamp="1542073669179"><title>就MapReduce单个而言它只是一个计算框架,没有什么环境的</title><children><topics type="attached"><topic id="0rklreqiieehv4sjfs4u8jcc81" modified-by="ganling" timestamp="1542073677953"><title>只是实际工作的时候把它运行在yarn上面</title><children><topics type="attached"><topic branch="folded" id="7bmsfg95d51oegui49ada7r3e6" modified-by="admin" timestamp="1525836616641"><title>MapReduce需要读取HDFS文件所以会结合HDFS</title></topic></topics></children></topic></topics><topics type="callout"><topic id="1g70kq5ph35pj8e32u8qeod9ik" modified-by="Administrator" timestamp="1509017594562"><title>就和Spring MVC一样</title><position svg:x="0" svg:y="0"/><children><topics type="attached"><topic id="0blemm3qn1ujiui1bsab39p3in" modified-by="Administrator" timestamp="1509017594562"><title>spring运行在tomcat上</title></topic></topics></children></topic></topics></children></topic></topics></children></topic><topic id="4lifdmd92nmr9ukpqril1a5i92" modified-by="ganling" timestamp="1542073664889"><title>缺点</title><children><topics type="attached"><topic id="7ume34lhjggqro6jkaisab0huv" modified-by="ganling" timestamp="1542073693874"><title>执行速度慢</title><children><topics type="callout"><topic id="7ongko6ggimk5vb5dngi2afhm1" modified-by="Administrator" timestamp="1509017594563"><title>3.0速度快了很多</title><position svg:x="0" svg:y="0"/></topic></topics><topics type="attached"><topic id="2l1t4pu7v50e70rbq4ss8amjkj" modified-by="ganling" timestamp="1542073695516"><title>IO瓶颈</title><children><topics type="attached"><topic id="7eec6l143qtk1ajsr9aagr4i18" modified-by="ganling" timestamp="1542073696761"><title>磁盘IO</title><children><topics type="attached"><topic branch="folded" id="67tnrhsg8t689nkftd1jvatrae" modified-by="admin" timestamp="1525836616641"><title svg:width="500">每个job的结果需要存储到HDFS集群机器的磁盘,而每个job又至少有三个备份因子,需要写3份,所以很慢</title></topic></topics></children></topic><topic branch="folded" id="3g1bv17t739t50lk92oa2v1fv6" modified-by="admin" timestamp="1525836616641"><title>网络IO</title></topic></topics></children></topic><topic id="5daho8oj6pq2403dvcuobsb9uf" modified-by="ganling" timestamp="1542073699807"><title>shuffle机制</title><children><topics type="attached"><topic id="2ppk1fbqacmc0a6jfd9pa82pvr" modified-by="ganling" timestamp="1542073705277"><title>数据需要输出到磁盘,而且每次都需要进行排序的操作</title><children><topics type="attached"><topic id="46hpkgpd51rnkpp475fc46jaj6" modified-by="ganling" timestamp="1542073706574"><title>当文件比较大的时候它执行的是外排序</title><children><topics type="callout"><topic id="79vhsjcupdq3svuh40evi2kd8b" modified-by="Administrator" timestamp="1509017594565"><title>外排序:在磁盘上存取文件排序</title><position svg:x="0" svg:y="0"/></topic></topics><topics type="attached"><topic branch="folded" id="3l5fhknfvtri8oofqnitoe7tf1" modified-by="admin" timestamp="1525836616641"><title>外排序是比较慢的</title><children><topics type="callout"><topic id="25amb2tjclarhj2ioggp0dm775" modified-by="Administrator" timestamp="1509017594567"><title>实质也是内存利用率方面的问题</title><position svg:x="0" svg:y="0"/><children><topics type="attached"><topic id="5ui4h5i2vqkejd6ss3g9in8hhv" modified-by="Administrator" timestamp="1509017594566"><title>MapReduce的maptask默认是分配1个g的内存</title><children><topics type="attached"><topic id="7k4rahr4ss76p4redkngjfmkdo" modified-by="Administrator" timestamp="1509017594566"><title>实质上MaptaskJVM的大小才200M</title><children><topics type="attached"><topic id="2c3qh7m1fvbfbbgrk7pjo4q21i" modified-by="Administrator" timestamp="1509017594566"><title>当输出到shuffle的时候,shuffle的缓冲区才100MB</title><children><topics type="attached"><topic id="33vvn5h60th85uridh3u590s0h" modified-by="Administrator" timestamp="1509017594566"><title>它希望剩余的100MB程序中用</title></topic></topics></children></topic></topics></children></topic></topics></children></topic><topic id="1eqm9s2snnnas4epurlu65to54" modified-by="Administrator" timestamp="1509017594567"><title>如果一个task分配500MB</title><children><topics type="attached"><topic id="32s4208em8famqrh9p5s8j3e45" modified-by="Administrator" timestamp="1509017594567"><title>shuffle分配400MB</title><children><topics type="attached"><topic id="0mkbcmnqm6l53stvl97mlun4p6" modified-by="Administrator" timestamp="1509017594567"><title>其它程序分配100MB</title><children><topics type="attached"><topic id="551k6f9qbtcsbasfrnoabs9gne" modified-by="Administrator" timestamp="1509017594566"><title>如果数据不太大的话就可以在内存中处理完。不用到磁盘中读写</title><children><topics type="attached"><topic id="6p52atmm0gqdo65hfhjcovom51" modified-by="Administrator" timestamp="1509017594566"><title>因为一个Maptask才处理128M数据</title></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic></topics></children></topic><topic id="7cl4214p8vup1sati32cqcaslk" modified-by="ganling" timestamp="1542073709990"><title>框架的缺陷</title><children><topics type="attached"><topic id="204fv4i1r474j1qs0ubpklj04q" modified-by="ganling" timestamp="1542073711719"><title svg:width="500">只有map和reduce两个操作算子,结构只能是 map ->reduce 或者map ->[map ->] *->reduce[->map]*</title><children><topics type="attached"><topic branch="folded" id="4p8k765j14sn5ugg5gg77bct35" modified-by="admin" timestamp="1525836616641"><title svg:width="500">针对比较复杂的任务,需要构建多个job来执行,而多个job有依赖关系的时候,它会增加数据的读写、网络数据传输等方面的问题</title></topic></topics></children></topic></topics></children></topic><topic id="0vbjaf14k8f0fmvm7r385ejaho" modified-by="ganling" timestamp="1542073724215"><title>进程方式启动</title><children><topics type="attached"><topic id="7ar2b2k9anncbj0r3g2or6m28j" modified-by="ganling" timestamp="1542073725532"><title>Task是以进程的方式来启动的</title><children><topics type="attached"><topic id="545ndv66cf73v2fep7pjau78u0" modified-by="ganling" timestamp="1542073727597"><title>你会发现每个MR运行的时候,前面总会占了几十秒</title><children><topics type="attached"><topic id="3c188evm84djjl7gde4o0ioa55" modified-by="ganling" timestamp="1542073730982"><title>那就是起进程的过程</title><children><topics type="attached"><topic b
没有合适的资源?快使用搜索试试~ 我知道了~
超级详细的spark体系思维导图
共210个文件
xml:147个
png:29个
7ldt5fj7d1hf2gpnpptv29tf46:1个
需积分: 39 35 下载量 46 浏览量
2018-11-13
14:40:27
上传
评论 1
收藏 34.52MB ZIP 举报
温馨提示
超级详细的spark思维导图,其中包含了spark-core、spark-streaming 、spark-sql...
资源推荐
资源详情
资源评论
收起资源包目录
超级详细的spark体系思维导图 (210个子文件)
0k6jk03i63hgdbs700drtf8cc0 116KB
165lgpgtq9uaprm6luh96nek1d 275KB
1aiomt98vi4inp3r9b3ri6cson 254KB
1ptnstusfq9v12c4cirshud6dg 339KB
1s5e6rnid79665blf6pgn6mtm2 75KB
1tkggmvf21vie6serpiaq3erdu 207KB
202tc83a0r9jlomehc9mrgvsj2 281KB
22ajf3seirbrajhttrgs379ml8 190KB
22btlke314jfgjrmeca5v97asp 227KB
2htpfevunsnfqcc3tllsahcoq9 313KB
2oh2bubdc76u6s9t2ettivunqo 244KB
2oqtkpll1ngmkcc9tcnt90uips 266KB
30olj1uqsu0civse6n9ma77313 138KB
3d53v5mj0i4qo1dd5j23m71isu 79KB
3e9815viubqd5cosseod7lvk80 338KB
3jjqqbs83d05ifb3ffqu531org 458KB
3mm35a4m2r2ejaeg2uks9kbtna 363KB
46h2urvjlk3fmbi57qtv5378m1 42KB
4kc89h4fdfjj5tvngka0kfsjs3 260KB
4qnisrfprcmmjptlqujgltndev 199KB
4ub38n83bumgmrga0gd1qbqa3k 220KB
56v2v1e8motl00ervt4cgkprdi 124KB
5lc7gnlslefude1g06d6tu55ie 328KB
632icg5l5ogd56h3kat4nfcf4h 216KB
67aas3mafg5gvdmnfb4jmkacsf 134KB
6gqubcvf1276kpm85d07mahqoj 572KB
6hm678ulqe4p6j6kkf1a0u18m7 218KB
6matvii59fjclk0cr5cda3h311 181KB
6phcai9rldjlms3vfv7ea611t2 81KB
70paco2plcfk2sb9e2j0d4h5o4 349KB
76uj5dl376du10e3ncq8nmempi 216KB
78pi548374i1mv1c2tcsiam6k6 127KB
79tijhbn8l9qk3phmlaf8p205v 174KB
7ldt5fj7d1hf2gpnpptv29tf46 405KB
7ullbevncem3rtlbe5scbj80k3.png 437KB
5bj6q7d0fts8md0n5j7fatdv76.png 377KB
3bsg04bb6ir09nk40lov3bqrfl.png 256KB
3vad2imleo0e3sdk1l4pt8gm8f.png 201KB
6rp500v34chlmb85gk00bmvqoh.png 179KB
0e6m6msgmhtku8i8c6k0inmj4l.png 165KB
thumbnail.png 148KB
4dm55gj7tjqr0r5vtob4ujrfr6.png 144KB
0gdl1mmse2us938bflfojdq4lc.png 131KB
71mvsa72hbv7e9dd2q8esbgkmg.png 128KB
0cls4e8cs2uok8vksadsp6fbbg.png 121KB
1ofjtpvsgih789dq6931q9dp08.png 101KB
7dfm3c3h1ec747idh5nkf6tibo.png 81KB
0qqcdu28mr6do3akbn9vek9crd.png 72KB
5j3bnaem65n5n8n2l1bvmisjpk.png 69KB
163bd8vvm2b7a28ecujur786ka.png 67KB
6v956hmgg531sovqp8bunsfv3q.png 67KB
1740kdlvank2juei5jhbqdbgdf.png 61KB
02b8df3fq30ah3qsi12e22t6mo.png 59KB
3fhepfl327i31bncu9huer7s7b.png 54KB
6804jeikk581jauhq5bqbgfvhl.png 52KB
2hd2k440f8mh15g01tth56d2ao.png 48KB
199eec9o0nbu04qrnfsbvtf4ca.png 44KB
5b427etn93b1fjapn1tih83o6q.png 37KB
7flk8bs0etptnojfp59o7i36ff.png 30KB
2b8d24m08oun27gga8uqoj0nnt.png 25KB
7o389lf2le59d9iooblra1ac0c.png 21KB
2f7vapcbuc2307241bbsnp0am2.png 17KB
7mfgraq6bldut9msaonoard8ee.png 17KB
content.xml 897KB
rev-76-1516678385244.xml 328KB
rev-74-1516073988816.xml 328KB
rev-75-1516269678722.xml 327KB
rev-77-1525836611892.xml 326KB
rev-78-1525836617782.xml 326KB
rev-79-1542073787749.xml 325KB
rev-80-1542074989689.xml 325KB
rev-73-1516063108754.xml 305KB
rev-69-1513237445596.xml 301KB
rev-72-1515137660839.xml 300KB
rev-66-1513220342736.xml 300KB
rev-60-1511748689933.xml 300KB
rev-68-1513237441002.xml 300KB
rev-53-1510538603228.xml 300KB
rev-56-1510543690202.xml 299KB
rev-58-1510562411476.xml 299KB
rev-57-1510543778820.xml 299KB
rev-59-1511252511150.xml 299KB
rev-71-1513240045265.xml 299KB
rev-51-1510311449667.xml 299KB
rev-65-1513220332529.xml 299KB
rev-64-1513220300250.xml 299KB
rev-49-1510224185382.xml 299KB
rev-55-1510543658821.xml 299KB
rev-50-1510304527279.xml 299KB
rev-67-1513236840252.xml 298KB
rev-62-1512461627022.xml 298KB
rev-70-1513240041494.xml 298KB
rev-46-1510218006490.xml 298KB
rev-63-1513162903230.xml 298KB
rev-61-1512461604676.xml 298KB
rev-52-1510538591397.xml 297KB
rev-47-1510220583768.xml 297KB
rev-48-1510221790179.xml 295KB
rev-45-1510216682306.xml 295KB
rev-54-1510543618044.xml 294KB
共 210 条
- 1
- 2
- 3
资源评论
梦梦王爷
- 粉丝: 4
- 资源: 3
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功