性能分析

<p><strong>l&nbsp; 网络请求</strong></p> <p>对APP中的全量HTTP/HTTPS请求(包含自营和第三方的原生请求)性能进行汇总展示,列表区按照最差域名性能排序展示,图形区分别展示域名性能趋势图等展示,点选某个域名后则可展示出该域名下各个URL请求的性能及对应的性能趋势图</p> <p><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911205653-1165979994a5.png" style="height:390px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 网络请求步骤缓慢原因</strong></p> <p>点击某个域名后,在下方会展示出该域名下的各个URL请求的性能概况,右侧的趋势图会关联显示为该域名的响应时间趋势图,趋势图部分可展示各个网络请求环节耗时,如DNS耗时、TCP耗时、SSL耗时、服务器响应用时、接收用时等,除此外还有单域名请求量、性能缓慢与异常对比查看,快速定位耗时环节。</p> <p>分析思路:优先判断慢域名&rarr;掌握慢域名的解析主机&rarr;慢主机下的慢请求&rarr;快照分析。</p> <p>主机维度分析与请求分析逻辑均是三层结构分析,具体内容如下:</p> <p>&Oslash;&nbsp; <strong>一层所有慢域名统计分析</strong>,如下截图。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1" o:spid="_x0000_i1052" type="#_x0000_t75" style='width:415.5pt;height:147pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image003.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911205741-17ce7c099450.png" style="height:297px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_22" o:spid="_x0000_i1051" type="#_x0000_t75" style='width:414.75pt;height:143.25pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image005.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911205842-1d3343aa91d7.png" style="height:290px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_11" o:spid="_x0000_i1050" type="#_x0000_t75" style='width:387.75pt;height:145.5pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image007.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911205928-133a256b9df2.png" style="height:313px; width:835px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_23" o:spid="_x0000_i1049" type="#_x0000_t75" style='width:414.75pt; height:144.75pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image009.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911205943-148a308e9a8e.png" style="height:293px; width:840px" /></p> <p>&nbsp;</p> <p>&Oslash;&nbsp; <strong>二层指定域名分析</strong>:点击某域名后,展示该域名的性能堆叠图、HTTP错误率及吞吐率、健康请求分布与慢速比以及指定域名下的请求性能列表</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_21" o:spid="_x0000_i1048" type="#_x0000_t75" style='width:414.75pt; height:129.75pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image011.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210020-1fde6c5b9ae7.png" style="height:262px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_20" o:spid="_x0000_i1047" type="#_x0000_t75" style='width:414.75pt; height:146.25pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image013.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210033-121aa26995d6.png" style="height:296px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_19" o:spid="_x0000_i1046" type="#_x0000_t75" style='width:414.75pt; height:147pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image015.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210104-16ae92869148.png" style="height:299px; width:840px" /></p> <p>支持查看慢请求占比,定位对业务影响最大的慢请求。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_18" o:spid="_x0000_i1045" type="#_x0000_t75" style='width:414.75pt; height:56.25pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image017.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210117-13cf481d9f50.png" style="height:114px; width:840px" /></p> <p>&Oslash;&nbsp; 三层慢请求深度分析</p> <p>点选某次缓慢请求的URL地址后,即可展现慢请求网络传输过程中的拖累环节、城市运营商及主机的性能分布,支持切换指标与维度下钻。</p> <p>&nbsp;</p> <p><strong>判断问题思路:</strong></p> <p>1、 直观掌握影响请求各环节的主机、地域运营商分布情况,缩小问题范围。</p> <p>2、 通过慢请求下载大小,排查大文件对性能的影响。</p> <p>&Oslash;&nbsp; <strong>时序图说明:</strong></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_25" o:spid="_x0000_i1044" type="#_x0000_t75" style='width:414.75pt; height:55.5pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image019.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210219-15da4c0990f2.png" style="height:113px; width:840px" /></p> <p>按不同颜色显示所选请求响应时间中各个环节的用时。</p> <p>&nbsp;</p> <p><strong>l&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 主机维度分析</strong></p> <p>通过功能按钮选择&ldquo;显示请求&rdquo;或&ldquo;显示主机&rdquo;,&ldquo;显示请求&rdquo;则按照请求URL形式进行展示该于域名下的请求,&ldquo;显示主机&rdquo;则按照请求所到达的服务端IP进行展示,可定位影响性能的主机,并进一步分析慢请求的主机分析,步骤同上请求性能分析。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_10" o:spid="_x0000_i1043" type="#_x0000_t75" style='width:415.5pt;height:133.5pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image021.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911210234-1141d253930c.png" style="height:270px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; WEBVIEW</strong></p> <p>WebView模块,对非原生的HTTP请求(自营和第三方)和H5页面性能进行监控,监控APP中H5页面加载性能,从首包时间、HTML下载用时、DOM建立用时、资源加载用时、渲染耗时等多方位分析页面加载性能</p> <p>&nbsp;</p> <p><strong>WEBVIEW中的HTTP请求分析</strong></p> <p>可通过请求/主机两大维度,对非原生请求进行分析,定位影响慢请求耗时环节。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_27" o:spid="_x0000_i1042" type="#_x0000_t75" style='width:414.75pt;height:133.5pt;visibility:visible; mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image023.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911211105-13feadaf9266.png" style="height:270px; width:840px" /></p> <p>注:WebView的HTTP请求分析与&ldquo;网络请求&rdquo;分析思路相同。</p> <p><strong>WEBVIEW中的H5页面分析</strong></p> <p>采集H5页面加载时间、白屏时间,渲染时间以及各元素的url、加载时间等信息,定位App中H5页面中的慢元素,有效优化H5页面访问效率。</p> <p>&Oslash;&nbsp; <strong>一层:页面分组域名的整体性能</strong></p> <p>H5页面性能以页面的域名进行分组,按选择指标(整体性能、吞吐率)进行排序,通过排序指标展示最差域名TOP5的趋势图、吞吐率最高域名TOP5、JS错误率最高域名TOP5</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1051" o:spid="_x0000_i1041" type="#_x0000_t75" style='width:414.75pt;height:141.75pt;visibility:visible; mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image025.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911211209-123791109e44.png" style="height:287px; width:840px" /></p> <p>l&nbsp; 整体性能:页面的加载用时,即页面从开始访问到全部打开的整体耗时。计算方式:loadEventEnd &ndash; navigationStart</p> <p>l&nbsp; 首包时间:从页面开始浏览到收到服务器第一包数据的时间。计算方式:responseStart-navigationStart</p> <p>l&nbsp; HTML下载用时:浏览器下载完整的HTML文档内容所需要的时间,计算方式:</p> <p>responseEnd-responseStart</p> <p>l&nbsp; DOM建立时间:浏览器建立文档对象模型(DOM)的用时,计算方式:</p> <p>domContentLoadedEventStart-domLoading</p> <p>l&nbsp; 资源加载时间:从浏览器开始下资源下载到页面渲染完成的时间间隔,计算方式:</p> <p>loadEventEnd-domContentLoadedEventStart</p> <p>l&nbsp; 正确访问次数:选定时间范围内容,排除了JS错误访问的所有访问次数。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1059" o:spid="_x0000_i1040" type="#_x0000_t75" style='width:414.75pt;height:129pt;visibility:visible; mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image027.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911211607-183a82d29f52.png" style="height:261px; width:840px" /></p> <p>l&nbsp; 吞吐率:平均每分钟的总访问次数。</p> <p>l&nbsp; 总请求次数:该页面总的请求次数。</p> <p><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911211859-128158d6935f.png" style="height:257px; width:840px" /></p> <p>l&nbsp; JS错误率:页面发生JS错误的比率,计算方式:发生JS错误的页面访问次数/该页面的总访问次数*100%</p> <p>l&nbsp; JS错误次数:JS发生的总次数。如一次页面访问中出现过10次JS错误,则错误次数为10。</p> <p>&Oslash;&nbsp; <strong>二层 指定域名性能</strong></p> <p>展示:整体性能堆叠图、平均页面错误次数及平均吞吐率、JS错误率最高页面TOP5</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1057" o:spid="_x0000_i1038" type="#_x0000_t75" style='width:415.5pt;height:132pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image031.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911211943-1b2bb6f39a01.png" style="height:268px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1056" o:spid="_x0000_i1037" type="#_x0000_t75" style='width:414.75pt;height:128.25pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image033.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911212023-128efce49efa.png" style="height:259px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1055" o:spid="_x0000_i1036" type="#_x0000_t75" style='width:415.5pt;height:131.25pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image035.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911212038-13e6eff19d6a.png" style="height:266px; width:840px" /></p> <p>&Oslash;&nbsp; <strong>三层 指定页面性能信息</strong></p> <p>点击某个域名下WebView页面,按平均加载时间排序信息,右侧展示选中域名下各WebView页面平均加载时间最差的TOP10页面趋势图。底部是所有慢加载页面列表,慢加载(加载时间&gt;1.2s)的发生时间、页面、加载时间和白屏时间。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1054" o:spid="_x0000_i1035" type="#_x0000_t75" style='width:414.75pt;height:146.25pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image037.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911212548-1315de149e44.png" style="height:295px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1053" o:spid="_x0000_i1034" type="#_x0000_t75" style='width:415.5pt;height:145.5pt;visibility:visible; mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image039.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213300-1d51d78a9560.png" style="height:294px; width:840px" /></p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1052" o:spid="_x0000_i1033" type="#_x0000_t75" style='width:414.75pt;height:132.75pt;visibility:visible; mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image041.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213333-1f2cd86398f8.png" style="height:268px; width:840px" /></p> <p>WebView模块,帮助运维人员掌握业务部署情况,了解承载业务的机房、云主机及CDN等第三方服务质量。</p> <p>&nbsp;</p> <p><strong>l&nbsp; CDN性能</strong></p> <p>CDN性能模块通过四层级信息展示CDN性能问题:全部CDN厂商&rarr;选中CDN厂商下所有域名&rarr;选中域名的地区运营商覆盖&rarr;选中地域运营商,帮助运维人员透视跨地域/运营商访问的网络资源问题分布及其上支撑业务的服务质量,即时掌控运维基础服务能力。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_30" o:spid="_x0000_i1032" type="#_x0000_t75" style='width:414.75pt; height:129.75pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image043.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213420-1ce946ca96a3.png" style="height:263px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 解析建联</strong></p> <p>该模块主要对原生自营域名的DNS解析和自营主机的TCP建连性能进行监测,帮助运维人员快速掌握域名解析耗时和主机建连(TCP、SSL)耗时情况,掌控基础网络资源的服务能力,直观定位网络问题。</p> <p>1.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 解析分析</p> <p>解析分析主要是通过累计解析用时(默认显示)、解析用时、错误率等排序指标,分析自有主机解析服务质量。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1028" o:spid="_x0000_i1031" type="#_x0000_t75" style='width:415.5pt;height:136.5pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image045.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213435-1f759199957a.png" style="height:275px; width:840px" /></p> <p>2.&nbsp;&nbsp; 建连分析</p> <p>通过平均建连、累计建连用时、建连错误率等关键指标,分析建连最差主机,错误率最高主机以及建连次数最多TOP主机,帮助运维人员了解问题域名的主机分布和性能。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1029" o:spid="_x0000_i1030" type="#_x0000_t75" style='width:414.75pt; height:145.5pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image047.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213712-1c398ced9895.png" style="height:293px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 视图性能</strong></p> <p>该模块主要对应用视图启动性能进行监控,依据视图启动时间的健康度,划分视图启动分布区间, 从整体了解视图启动效果,优先定位启动慢视图,帮助研发人员直观掌握视图启动效果。</p> <p>可通过视图平均/累计启动时间、启动次数等指标进行排序,了解应用中视图启动耗时分布情况,掌握最差视图,便于研发人员深入分析。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1030" o:spid="_x0000_i1029" type="#_x0000_t75" style='width:414.75pt; height:121.5pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image049.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213750-13c471d795db.png" style="height:247px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 地区组合</strong></p> <p>从地区、运营商、接入方式等多维度组合分析,掌握请求各环节性能。帮助运维人员快速定位问题影响面。</p> <p>同时,利用筛选项,缩小问题范围,实现对网络、对资源的投入进行全局把控。</p> <p><strong><!--[if gte vml 1]><v:shape id="图片_x0020_1031" o:spid="_x0000_i1028" type="#_x0000_t75" style='width:414.75pt; height:133.5pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image051.png" o:title=""/> </v:shape><![endif]--></strong><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213817-154dbfba9ea5.png" style="height:270px; width:840px" /></p> <p><strong>l&nbsp; APP版本</strong></p> <p>该模块基于APP版本维度对请求进行分析,由APP版本号及SDK版本号组合分析。</p> <p><!--[if gte vml 1]><v:shape id="图片_x0020_1032" o:spid="_x0000_i1027" type="#_x0000_t75" style='width:415.5pt; height:122.25pt;visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image053.png" o:title=""/> </v:shape><![endif]--><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213837-1c44a0159294.png" style="height:247px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 设备组合</strong></p> <p>基于设备型号、操作系统版本组合信息对请求进行分析。</p> <p><strong><!--[if gte vml 1]><v:shape id="图片_x0020_1033" o:spid="_x0000_i1026" type="#_x0000_t75" style='width:414pt;height:115.5pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image055.png" o:title=""/> </v:shape><![endif]--></strong><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911213904-11ee35199d8a.png" style="height:236px; width:840px" /></p> <p>&nbsp;</p> <p><strong>l&nbsp; 活跃设备统计</strong></p> <p>该模块可从总览、设备型号、APP版本、操作系统版本和地区等五个维度进行活跃设备统计信息展示。</p> <p><strong><!--[if gte vml 1]><v:shape id="图片_x0020_1034" o:spid="_x0000_i1025" type="#_x0000_t75" style='width:372pt;height:178.5pt; visibility:visible;mso-wrap-style:square'> <v:imagedata src="file:///C:\Users\EX-ZHA~1\AppData\Local\Temp\47\msohtmlclip1\01\clip_image057.png" o:title="" cropleft="6529f"/> </v:shape><![endif]--></strong><img src="https://obs-cn-shanghai.yun.pingan.com/pcp-portal/20182911214028-1d9dcbac98b2.png" style="height:364px; width:840px" /></p>
以上内容是否解决了您的问题?
请补全提交信息!
咨询·建议

电话咨询 - 7x24 小时

400-151-8800

邮件咨询

cloud@pingan.com

在线客服

7x24 小时,急速解答

工单支持

解决云产品相关技术问题