blob: e3ab04b6d2bfaea660b786fc1aa5187b4f1725a1 [file] [log] [blame]
import{_ as a,O as n,P as e,ah as t,aW as s}from"./framework-62ad666a.js";const o={},r=s(`<h2 id="influxdb-protocol" tabindex="-1"><a class="header-anchor" href="#influxdb-protocol" aria-hidden="true">#</a> InfluxDB-Protocol</h2><h3 id="_1-切换方案" tabindex="-1"><a class="header-anchor" href="#_1-切换方案" aria-hidden="true">#</a> 1.切换方案</h3><p>假如您原先接入 InfluxDB 的业务代码如下:</p><div class="language-java line-numbers-mode" data-ext="java"><pre class="language-java"><code><span class="token class-name">InfluxDB</span> influxDB <span class="token operator">=</span> <span class="token class-name">InfluxDBFactory</span><span class="token punctuation">.</span><span class="token function">connect</span><span class="token punctuation">(</span>openurl<span class="token punctuation">,</span> username<span class="token punctuation">,</span> password<span class="token punctuation">)</span><span class="token punctuation">;</span>
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div><p>您只需要将 InfluxDBFactory 替换为 <strong>IoTDBInfluxDBFactory</strong> 即可实现业务向 IoTDB 的切换:</p><div class="language-java line-numbers-mode" data-ext="java"><pre class="language-java"><code><span class="token class-name">InfluxDB</span> influxDB <span class="token operator">=</span> <span class="token class-name">IoTDBInfluxDBFactory</span><span class="token punctuation">.</span><span class="token function">connect</span><span class="token punctuation">(</span>openurl<span class="token punctuation">,</span> username<span class="token punctuation">,</span> password<span class="token punctuation">)</span><span class="token punctuation">;</span>
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div><h3 id="_2-方案设计" tabindex="-1"><a class="header-anchor" href="#_2-方案设计" aria-hidden="true">#</a> 2.方案设计</h3><h4 id="_2-1-influxdb-protocol适配器" tabindex="-1"><a class="header-anchor" href="#_2-1-influxdb-protocol适配器" aria-hidden="true">#</a> 2.1 InfluxDB-Protocol适配器</h4><p>该适配器以 IoTDB Java Session 接口为底层基础,实现了 InfluxDB 的 Java 接口 <code>interface InfluxDB</code>,对用户提供了所有 InfluxDB 的接口方法,最终用户可以无感知地使用 InfluxDB 协议向 IoTDB 发起写入和读取请求。</p><figure><img src="https://alioss.timecho.com/docs/img/UserGuide/API/IoTDB-InfluxDB/architecture-design.png?raw=true" alt="architecture-design" tabindex="0" loading="lazy"><figcaption>architecture-design</figcaption></figure><figure><img src="https://alioss.timecho.com/docs/img/UserGuide/API/IoTDB-InfluxDB/class-diagram.png?raw=true" alt="class-diagram" tabindex="0" loading="lazy"><figcaption>class-diagram</figcaption></figure><h4 id="_2-2-元数据格式转换" tabindex="-1"><a class="header-anchor" href="#_2-2-元数据格式转换" aria-hidden="true">#</a> 2.2 元数据格式转换</h4><p>InfluxDB 的元数据是 tag-field 模型,IoTDB 的元数据是树形模型。为了使适配器能够兼容 InfluxDB 协议,需要把 InfluxDB 的元数据模型转换成 IoTDB 的元数据模型。</p><h5 id="_2-2-1-influxdb-元数据" tabindex="-1"><a class="header-anchor" href="#_2-2-1-influxdb-元数据" aria-hidden="true">#</a> 2.2.1 InfluxDB 元数据</h5><ol><li>database: 数据库名。</li><li>measurement: 测量指标名。</li><li>tags : 各种有索引的属性。</li><li>fields : 各种记录值(没有索引的属性)。</li></ol><figure><img src="https://alioss.timecho.com/docs/img/UserGuide/API/IoTDB-InfluxDB/influxdb-data.png?raw=true" alt="influxdb-data" tabindex="0" loading="lazy"><figcaption>influxdb-data</figcaption></figure><h5 id="_2-2-2-iotdb-元数据" tabindex="-1"><a class="header-anchor" href="#_2-2-2-iotdb-元数据" aria-hidden="true">#</a> 2.2.2 IoTDB 元数据</h5><ol><li>storage group: 存储组。</li><li>path(time series ID):存储路径。</li><li>measurement: 物理量。</li></ol><figure><img src="https://alioss.timecho.com/docs/img/UserGuide/API/IoTDB-InfluxDB/iotdb-data.png?raw=true" alt="iotdb-data" tabindex="0" loading="lazy"><figcaption>iotdb-data</figcaption></figure><h5 id="_2-2-3-两者映射关系" tabindex="-1"><a class="header-anchor" href="#_2-2-3-两者映射关系" aria-hidden="true">#</a> 2.2.3 两者映射关系</h5><p>InfluxDB 元数据和 IoTDB 元数据有着如下的映射关系:</p><ol><li>InfluxDB 中的 database 和 measurement 组合起来作为 IoTDB 中的 storage group。</li><li>InfluxDB 中的 field key 作为 IoTDB 中 measurement 路径,InfluxDB 中的 field value 即是该路径下记录的测点值。</li><li>InfluxDB 中的 tag 在 IoTDB 中使用 storage group 和 measurement 之间的路径表达。InfluxDB 的 tag key 由 storage group 和 measurement 之间路径的顺序隐式表达,tag value 记录为对应顺序的路径的名称。</li></ol><p>InfluxDB 元数据向 IoTDB 元数据的转换关系可以由下面的公示表示:</p><p><code>root.{database}.{measurement}.{tag value 1}.{tag value 2}...{tag value N-1}.{tag value N}.{field key}</code></p><figure><img src="https://alioss.timecho.com/docs/img/UserGuide/API/IoTDB-InfluxDB/influxdb-vs-iotdb-data.png?raw=true" alt="influxdb-vs-iotdb-data" tabindex="0" loading="lazy"><figcaption>influxdb-vs-iotdb-data</figcaption></figure><p>如上图所示,可以看出:</p><p>我们在 IoTDB 中使用 storage group 和 measurement 之间的路径来表达 InfluxDB tag 的概念,也就是图中右侧绿色方框的部分。</p><p>storage group 和 measurement 之间的每一层都代表一个 tag。如果 tag key 的数量为 N,那么 storage group 和 measurement 之间的路径的层数就是 N。我们对 storage group 和 measurement 之间的每一层进行顺序编号,每一个序号都和一个 tag key 一一对应。同时,我们使用 storage group 和 measurement 之间每一层 <strong>路径的名字</strong> 来记 tag value,tag key 可以通过自身的序号找到对应路径层级下的 tag value.</p><h5 id="_2-2-4-关键问题" tabindex="-1"><a class="header-anchor" href="#_2-2-4-关键问题" aria-hidden="true">#</a> 2.2.4 关键问题</h5><p>在 InfluxDB 的 SQL 语句中,tag 出现的顺序的不同并不会影响实际的执行结果。</p><p>例如:<code>insert factory, workshop=A1, production=B1 temperature=16.9</code> 和 <code>insert factory, production=B1, workshop=A1 temperature=16.9</code> 两条 InfluxDB SQL 的含义(以及执行结果)相等。</p><p>但在 IoTDB 中,上述插入的数据点可以存储在 <code>root.monitor.factory.A1.B1.temperature</code> 下,也可以存储在 <code>root.monitor.factory.B1.A1.temperature</code> 下。因此,IoTDB 路径中储存的 InfluxDB 的 tag 的顺序是需要被特别考虑的,因为 <code>root.monitor.factory.A1.B1.temperature</code> 和 <code>root.monitor.factory.B1.A1.temperature</code> 是两条不同的序列。我们可以认为,IoTDB 元数据模型对 tag 顺序的处理是“敏感”的。</p><p>基于上述的考虑,我们还需要在 IoTDB 中记录 InfluxDB 每个 tag 对应在 IoTDB 路径中的层级顺序,以确保在执行 InfluxDB SQL 时,不论 InfluxDB SQL 中 tag 出现的顺序如何,只要该 SQL 表达的是对同一个时间序列上的操作,那么适配器都可以唯一对应到 IoTDB 中的一条时间序列上进行操作。</p><p>这里还需要考虑的另一个问题是:InfluxDB 的 tag key 及对应顺序关系应该如何持久化到 IoTDB 数据库中,以确保不会丢失相关信息。</p><p><strong>解决方案:</strong></p><p><strong>tag key 对应顺序关系在内存中的形式</strong></p><p>通过利用内存中的<code>Map&lt;Measurement, Map&lt;Tag Key, Order&gt;&gt;</code> 这样一个 Map 结构,来维护 tag 在 IoTDB 路径层级上的顺序。</p><div class="language-java line-numbers-mode" data-ext="java"><pre class="language-java"><code> <span class="token class-name">Map</span><span class="token generics"><span class="token punctuation">&lt;</span><span class="token class-name">String</span><span class="token punctuation">,</span> <span class="token class-name">Map</span><span class="token punctuation">&lt;</span><span class="token class-name">String</span><span class="token punctuation">,</span> <span class="token class-name">Integer</span><span class="token punctuation">&gt;</span><span class="token punctuation">&gt;</span></span> measurementTagOrder
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div><p>可以看出 Map 是一个两层的结构。</p><p>第一层的 Key String 类型的 InfluxDB measurement,第一层的 Value 是一个 &lt;String, Integer&gt; 结构的 Map。</p><p>第二层的 Key 是 String 类型的 InfluxDB tag key,第二层的 Value 是 Integer 类型的 tag order,也就是 tag 在 IoTDB 路径层级上的顺序。</p><p>使用时,就可以先通过 InfluxDB measurement 定位,再通过 InfluxDB tag key 定位,最后就可以获得 tag IoTDB 路径层级上的顺序了。</p><p><strong>tag key 对应顺序关系的持久化方案</strong></p><p>存储组为<code>root.TAG_INFO</code>,分别用存储组下的 <code>database_name</code>, <code>measurement_name</code>, <code>tag_name</code> 和 <code>tag_order</code> 测点来存储 tag key 极其对应的顺序关系。</p><div class="language-text line-numbers-mode" data-ext="text"><pre class="language-text"><code>+-----------------------------+---------------------------+------------------------------+----------------------+-----------------------+
| Time|root.TAG_INFO.database_name|root.TAG_INFO.measurement_name|root.TAG_INFO.tag_name|root.TAG_INFO.tag_order|
+-----------------------------+---------------------------+------------------------------+----------------------+-----------------------+
|2021-10-12T01:21:26.907+08:00| monitor| factory| workshop| 1|
|2021-10-12T01:21:27.310+08:00| monitor| factory| production| 2|
|2021-10-12T01:21:27.313+08:00| monitor| factory| cell| 3|
|2021-10-12T01:21:47.314+08:00| building| cpu| tempture| 1|
+-----------------------------+---------------------------+------------------------------+----------------------+-----------------------+
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div></div></div><h4 id="_2-3-实例" tabindex="-1"><a class="header-anchor" href="#_2-3-实例" aria-hidden="true">#</a> 2.3 实例</h4><h5 id="_2-3-1-插入数据" tabindex="-1"><a class="header-anchor" href="#_2-3-1-插入数据" aria-hidden="true">#</a> 2.3.1 插入数据</h5><ol><li><p>假定按照以下的顺序插入三条数据到 InfluxDB (database=monitor):</p><p>(1)<code>insert student,name=A,phone=B,sex=C score=99</code></p><p>(2)<code>insert student,address=D score=98</code></p><p>(3)<code>insert student,name=A,phone=B,sex=C,address=D score=97</code></p></li><li><p>简单对上述 InfluxDB 的时序进行解释,database monitor; measurement studenttag 分别是 namephonesex addressfield score。</p></li></ol><p>对应的InfluxDB的实际存储为:</p><div class="language-text line-numbers-mode" data-ext="text"><pre class="language-text"><code>time address name phone sex socre
---- ------- ---- ----- --- -----
1633971920128182000 A B C 99
1633971947112684000 D 98
1633971963011262000 D A B C 97
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div></div></div><ol start="3"><li><p>IoTDB顺序插入三条数据的过程如下:</p><p>(1)插入第一条数据时,需要将新出现的三个 tag key 更新到 table 中,IoTDB 对应的记录 tag 顺序的 table 为:</p><table><thead><tr><th>database</th><th>measurement</th><th>tag_key</th><th>Order</th></tr></thead><tbody><tr><td>monitor</td><td>student</td><td>name</td><td>0</td></tr><tr><td>monitor</td><td>student</td><td>phone</td><td>1</td></tr><tr><td>monitor</td><td>student</td><td>sex</td><td>2</td></tr></tbody></table><p>(2)插入第二条数据时,由于此时记录 tag 顺序的 table 中已经有了三个 tag key,因此需要将出现的第四个 tag key=address 更新记录。IoTDB 对应的记录 tag 顺序的 table 为:</p><table><thead><tr><th>database</th><th>measurement</th><th>tag_key</th><th>order</th></tr></thead><tbody><tr><td>monitor</td><td>student</td><td>name</td><td>0</td></tr><tr><td>monitor</td><td>student</td><td>phone</td><td>1</td></tr><tr><td>monitor</td><td>student</td><td>sex</td><td>2</td></tr><tr><td>monitor</td><td>student</td><td>address</td><td>3</td></tr></tbody></table><p>(3)插入第三条数据时,此时四个 tag key 都已经记录过,所以不需要更新记录,IoTDB 对应的记录 tag 顺序的 table 为:</p><table><thead><tr><th>database</th><th>measurement</th><th>tag_key</th><th>order</th></tr></thead><tbody><tr><td>monitor</td><td>student</td><td>name</td><td>0</td></tr><tr><td>monitor</td><td>student</td><td>phone</td><td>1</td></tr><tr><td>monitor</td><td>student</td><td>sex</td><td>2</td></tr><tr><td>monitor</td><td>student</td><td>address</td><td>3</td></tr></tbody></table></li><li><p>(1)第一条插入数据对应 IoTDB 时序为 root.monitor.student.A.B.C</p><p>(2)第二条插入数据对应 IoTDB 时序为 root.monitor.student.PH.PH.PH.D (其中PH表示占位符)。</p><p>需要注意的是,由于该条数据的 tag key=address 是第四个出现的,但是自身却没有对应的前三个 tag 值,因此需要用 PH 占位符来代替。这样做的目的是保证每条数据中的 tag 顺序不会乱,是符合当前顺序表中的顺序,从而查询数据的时候可以进行指定 tag 过滤。</p><p>(3)第三条插入数据对应 IoTDB 时序为 root.monitor.student.A.B.C.D</p><p>对应的 IoTDB 的实际存储为:</p></li></ol><div class="language-text line-numbers-mode" data-ext="text"><pre class="language-text"><code>+-----------------------------+--------------------------------+-------------------------------------+----------------------------------+
| Time|root.monitor.student.A.B.C.score|root.monitor.student.PH.PH.PH.D.score|root.monitor.student.A.B.C.D.score|
+-----------------------------+--------------------------------+-------------------------------------+----------------------------------+
|2021-10-12T01:21:26.907+08:00| 99| NULL| NULL|
|2021-10-12T01:21:27.310+08:00| NULL| 98| NULL|
|2021-10-12T01:21:27.313+08:00| NULL| NULL| 97|
+-----------------------------+--------------------------------+-------------------------------------+----------------------------------+
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div></div></div><ol start="5"><li><p>如果上面三条数据插入的顺序不一样,我们可以看到对应的实际path路径也就发生了改变,因为InfluxDB数据中的Tag出现顺序发生了变化,所对应的到IoTDB中的path节点顺序也就发生了变化。</p><p>但是这样实际并不会影响查询的正确性,因为一旦InfluxdbTag顺序确定之后,查询也会按照这个顺序表记录的顺序进行Tag值过滤。所以并不会影响查询的正确性。</p></li></ol><h5 id="_2-3-2-查询数据" tabindex="-1"><a class="header-anchor" href="#_2-3-2-查询数据" aria-hidden="true">#</a> 2.3.2 查询数据</h5><ol><li><p>查询student中phone=B的数据。在database=monitor,measurement=student中tag=phone的顺序为1,order最大值是3,对应到IoTDB的查询为:</p><div class="language-sql line-numbers-mode" data-ext="sql"><pre class="language-sql"><code><span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span>B
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div></li><li><p>查询studentphone=Bscore&gt;97的数据,对应到IoTDB的查询为:</p><div class="language-sql line-numbers-mode" data-ext="sql"><pre class="language-sql"><code><span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span>B <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span>
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div></li><li><p>查询studentphone=Bscore&gt;97且时间在最近七天内的的数据,对应到IoTDB的查询为:</p><div class="language-sql line-numbers-mode" data-ext="sql"><pre class="language-sql"><code><span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span>B <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span> <span class="token operator">and</span> <span class="token keyword">time</span> <span class="token operator">&gt;</span> <span class="token function">now</span><span class="token punctuation">(</span><span class="token punctuation">)</span><span class="token operator">-</span><span class="token number">7</span>d
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div></div></div></li><li><p>查询studentname=Ascore&gt;97,由于tag存储在路径中,因此没有办法用一次查询同时完成tagfield的<strong>或</strong>语义查询,因此需要多次查询进行或运算求并集,对应到IoTDB的查询为:</p><div class="language-sql line-numbers-mode" data-ext="sql"><pre class="language-sql"><code><span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span>A
<span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span>
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div><div class="line-number"></div></div></div><p>最后手动对上面两次查询结果求并集。</p></li><li><p>查询student中(name=Aphone=Bsex=C)且score&gt;97,由于tag存储在路径中,因此没有办法用一次查询完成tag的<strong>或</strong>语义, 因此需要多次查询进行或运算求并集,对应到IoTDB的查询为:</p><div class="language-sql line-numbers-mode" data-ext="sql"><pre class="language-sql"><code><span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span>A <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span>
<span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span>B <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span>
<span class="token keyword">select</span> <span class="token operator">*</span> <span class="token keyword">from</span> root<span class="token punctuation">.</span>monitor<span class="token punctuation">.</span>student<span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span><span class="token operator">*</span><span class="token punctuation">.</span>C <span class="token keyword">where</span> score<span class="token operator">&gt;</span><span class="token number">97</span>
</code></pre><div class="line-numbers" aria-hidden="true"><div class="line-number"></div><div class="line-number"></div><div class="line-number"></div></div></div><p>最后手动对上面三次查询结果求并集。</p></li></ol>`,55);function d(i,p){return n(),e("div",null,[t(`
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
`),r])}const c=a(o,[["render",d],["__file","InfluxDB-Protocol.html.vue"]]);export{c as default};