There is no need to connect different SDKs at each end, and there is no need to view data in different reports. At present, there is only one solution known in the market that can see the business panorama in one report.

2. Open source, free, free to customize

All codes are open source. What data to collect can be customized; batch running frequency can be customized (real-time statistics can be done during activities); display reports can be customized.

3. Private deployment, data automatic control

Using traditional saas statistical products, all data are reported to statistical service providers.

uni Statistics 2.0 is implemented based on uniCloud, cloud functions and statistical data are all hosted in the developer's own service space (Alibaba Cloud or Tencent Cloud optional), and developers have complete control over their own statistical data.

4. Effective Error Analysis

Traditional App statistics platforms do not have js error statistics. Developers cannot know on which devices their js code will report errors.

The error information of uni statistics is more comprehensive, including js front-end errors and app native layer crashes.

Because uni-app runs after compilation, the traditional web and MiniApp statistics platform, its js error cannot be traced back to the pre-compiled source code of uni-app, and the error is incomprehensible.

Uni statistics support sourcemap, which can intuitively understand which line of code is wrong. See details

5. Default feature rich

Device Statistics
Registered user statistics (based on uni-id)
页面统计
内容统计
New, active, retained, bounce rate analysis
Channel analysis: auxiliary channel promotion
Error statistics and reporting: auxiliary product quality improvement
Native App crash statistics, view crash rate by version
Custom management, open source and extensible reports

6. More suitable for uni-app and uniCloud

uni statistics goes deep into the bottom layer of uni-app and uniCloud framework, and provides many functions that other statistics platforms cannot provide:

uni-app full-end recognition, no need to connect to different sdk, no need to switch between different reports and accumulate data by yourself
Automatically identify uni-app routes and automatically capture page titles (navigationBar or uni-nav-bar components based on pages.json)
Automatically capture js errors and report native crash logs on the app side
Compatible with the uni-app channel package packaging system, automatically identifying the channel package
基于uni-id账户体系，自动出具注册用户（不是设备）的新增、活跃、留存报表
兼容uniCloud opendb规范，从服务器端统计各项数据

7. Open Ecosystem

uni统计的开源且基于uni-admin的插件规范提供了插件机制，会有更多插件作者提供各种丰富的统计插件（如电商统计、内容统计等）。见插件市场

**DCloud attaches great importance to the growth of data-driven developer business. Recently, it has cooperated with professional big data service providers Sensors and Aladdin will jointly improve the product modules and ecological construction of Uni Statistics 2.0. If you have any needs for data analysis, you are welcome to fill in the demand survey, we will Based on your valuable suggestions, we will complete the iterative update based on the original Uni statistical products. **

# uni statistics old version upgrade suggestion

For developers using uni stats 1.0, it is recommended to upgrade to uni stats 2.0 as soon as possible. uni stats 1.0 will no longer be updated and maintained.

Functions	uni stats 1.0	uni stats 2.0
Whether Open Source	No	Yes
Deployment	Central Deployment	Private Deployment
Customization	Cannot be customized	Free customization
uni-id user statistics	excludes	included by default
Error Analysis Effectiveness	Lower	Higher

# Environmental requirements

uni statistics 2.0 only supports uni-app on the client side, non-uni-app cannot be used.
uni Statistics 2.0 relies on uniCloud in the cloud, but does not require developers to develop all applications based on uniCloud. The main business of the application can be connected to the traditional server, and uniCloud can be used for statistical management and report presentation.

# Usage tutorial

uni stats 2.0 includes two modules:

Front-end collection module: built in the uni-app framework and configured in the manifest.json of the business app;
Cloud statistics module: built in the uni-admin project template, reuses the same uniCloud service space as the business app;

# Front-end collection configuration

In the manifest.json of the business app project, select the uni statistics configuration item, select to enable uni statistics as required, and check 2.0 to enable the new version of statistics.

开启统计

The above visualization operations actually correspond to the uniStatistics node of the manifest source view. The manifest specification is detailed in the following document. If you don't care about the specification details, you can skip this section and continue to read the next section "MiniApp Domain Name Whitelist".

The enable field under uniStatistics is set to true|false to enable and disable uni statistics

Set the version property to "2" to enable new version statistics

//...
"uniStatistics": {
	"enable": true,//全局开启
	"version": "2" // 开启新版uni统计，值为字符串
},
//...

uniStatistics description

Field	Type	Default Value	Optional Value	Description
enable	Boolean	false	true ， false	Enable or disable statistics globally, the sub-platform configuration will override the current configuration
version	String	"1"	"1" , "2"	Statistical version, if not filled in, the default version 1.0 is used, and the 2.0 version is recommended
debug	Boolean	false	true , false	Enable statistical debugging mode, a large number of logs will be generated, and data will be reported during the development stage. Please turn off this option when the application is released
reportInterval	Number	10	-	Front-end data reporting period HBuilderX 3.5.4+ support
collectItems	Object	-	-	collection item configuration HBuilderX 3.5.5+ support

collectItems collection item configuration description

Field	Type	Default Value	Optional Value	Description
uniPushClientID	Boolean	false	true/false	是否开启推送`PushClientID`的采集
uniStatPageLog	Boolean	true	true/false	是否开启页面数据采集，

Sub-platform settings

uniStatistics 支持分平台设置，比如仅开启微信小程序平台的 uni统计，则在 manifest.json mp-weixin节点下设置 uniStatistics ->enable 即可，如下：

{
	//...
	"mp-weixin":{
			"uniStatistics": {
					"enable": true //微信平台开启统计
			}
	}
}

Description of uniStatistics by platform

Field	Type	Default Value	Optional Value	Description
enable	Boolean	false	true ， false	Enable or disable statistics by platform, the configuration of each platform will override the global configuration, uniStatistics needs to be under the platform configuration

attention

If there is a uniStatistics -> enable field under a sub-platform, the configuration under the sub-platform will be used first, otherwise, the global statistics setting will be used
There is no need to set the version and debug properties on different platforms, and the two properties only take effect globally
There is no need to set the debug attribute on different platforms, this attribute only takes effect globally
The data will be reported normally only after the debug mode is turned on or the code is released

# MiniApp Domain Name Whitelist

Because each MiniApp needs to configure a whitelist for accessible domain names, otherwise it will not be able to connect to the Internet. Therefore, when the MiniApp needs to be released, it is necessary to configure the domain name whitelist in the MiniApp management background of each manufacturer.

When deploying uni statistics, you can choose to use the Alibaba Cloud version or Tencent Cloud version of uniCloud. The domain names corresponding to different cloud vendors are different (there will be service space related configurations in the later chapters of the article)

uniCloud service provider	request legal domain name
Alibaba Cloud Public Beta	api.bspapp.com
Alibaba Cloud official version	api.next.bspapp.com
Tencent Cloud	tcb-api.tencentcloudapi.com

According to the selected uniCloud cloud vendor, you can configure the corresponding domain name in the domain name whitelist configuration of the MiniApp.

Notice

Projects that have not been republished after Alibaba Cloud migrated to the official version will still access the public beta address api.bspapp.com

# debug mode

Set the debug field under manifest.json -> uniStatistics to true|false to enable and disable the uni statistics debug mode

In debug mode, the key information of the reported data will be printed to the console, which is convenient to observe whether the collected information is correct. It is mostly used when customizing extensions.

Log format

=== indicates statistics log related logs

// mark statistics on
=== uni统计开启,version:2.0

// Collect logs, see collection type: Description of reported data
=== 统计数据采集：{采集类型} ===
// here is the raw data collected
{
	fvts: 1647313662
	lang: "zh"
	lt: "1"
	lvts: 1650857441
	md: "PC"
	t: 1650857461
	ttpj: "view"
	tvc: 14
	url: "pages/component/view/view"
	usv: "0.0.1"
	ut: "h5"
	// ...
}
=== 采集结束 ===

// data reported successfully
=== 统计队列数据上报 ===
// here is the report data
{usv: '0.0.1', t: 1650857765, requests: '["lt=11&ut=h5&url=/pages/component/view/view&tt=&u…&ch=&usv=0.0.1&t=1650857765&ttn=&ttpj=view&ttc="]'}
=== 上报结束 ===

For the description of log fields, see Front-end Collection SDK

# Background report configuration

# Create admin project

The background statistics report of uni statistics 2.0 is a built-in plugin of uni-admin.

uni-admin is an open source management background. That is, to use uni statistics 2.0, you need to install this background system, find the menu of uni statistics and use it.

Currently, there are two ways to install uni-admin: one-click deployment and manual installation.

One-click deployment of uni-admin only needs to go to the Plugin Market to search and find the uni-admin plug-in, and choose to enter the uni-admin plug-in details page 😕/ext.dcloud.net.cn/plugin?id=3268) and click the one-click deployment plug-in to uniCloud button, then select the service space where the uni-admin project needs to be deployed, and wait patiently for the system to automatically deploy. Can.

Manually install uni-admin, please refer to [uni-admin](https://uniapp.dcloud.io/uniCloud/admin.html#uni-admin-%E6%A1%86%E6%9E%B6-% E5%8E%9F%E5%90%8D-unicloud-admin) document, complete the following operations:

Create a new uni-admin project (select the uni-admin template in the HBuilderX new project interface)
In the pop-up cloud service space initialization wizard, associate the service space (if your business app has used uniCloud, select the same service space; otherwise, create a new service space and associate the same service space in the business app)
Deploy cloud resources: upload and deploy cloud functions, public modules, and initialize database tables through db_init.json. If the previous table already has conflicting data, you need to manually merge it yourself
Complete other initialization configurations in the cloud configuration center, such as: fill in your own passwordSecret field in the uniCloud/cloudfunctions/common/uni-config-center/uni-id/config.json file (used for encrypted password storage key) and tokenSecret fields (to generate the secret key required for token generation, you can also skip this item during the test), then right-click on the uni-config-center public module to upload the update. Note: When the business-side App project and the report-side uni-admin project are associated with the same service space, the mutual coverage problem of uni-config-center may occur. In this case, single-point maintenance is recommended. See details.
Run the uni-admin project, usually to the browser
Set up an administrator account
Add a record of "Applied to Statistics" (appid, etc.) in the "App Management" on the left
After the configuration is complete, start the business app, and you will see the data of the business app in the uni statistics menu on the left side of the uni-admin page after running (if you can't see the data, see the FAQ section below)
After passing the test, issue the uni-admin project in the HBuilder issue menu, and choose to deploy it to the front-end web hosting of the uniCloud service space. You can customize the domain name in the uniCloud web console.

attention

To run "Connect Local Cloud Functions", right-click on the database folder of uni-admin and click "Download all DB schemas and extended validation functions". If the db_init initialization data was missed in the service space initialization wizard before, right-click on uniCloud/database/db_init.json to initialize

uni-admin old project upgrade

If the business you need to count has used uni-admin before, you can upgrade uni-admin. Right-click package.json under the project and select "Update from Plugin Marketplace".

The newly added files of uni statistics mainly include:

Cloud functions: uniCloud/cloudfunctions/uni-stat-cron (run and batch cloud functions regularly), uniCloud/cloudfunctions/uni-stat-receiver (receive cloud objects processed by the front end)
Cloud functions common module: uniCloud/cloudfunctions/common/uni-stat
Data table: several schema files at the beginning of uni-stat in the uniCloud/database directory
Statistics pages: several pages under the pages/uni-stat folder

When updating, please merge pages.json to ensure that all new pages are registered successfully. Otherwise, after running, clicking on the left menu will report that the xxx file cannot be found.

When db_init.json initializes the database, the menu table opendb-admin-menus of the old project already has data. At this time, the data with conflicting keys cannot be inserted and need to be merged manually.

If the old project has not changed the menus and permissions, you can delete the old table and reinitialize it
If the menu menu of the old project has been changed, several pages of uni statistics need to be merged into the menu data table.

# Set the trigger period of the scheduled task cloud function

In uni statistics 2.0, the default trigger period of the timed task cloud function (uni-stat-cron) is to trigger every hour (ie: every 1 hour, the system will automatically trigger the uni-stat-cron cloud function ).

You can modify the trigger period of the Scheduled Task Cloud Function (uni-stat-cron) to the minute level (that is, every n minutes) as needed, but pay attention to the modification of the trigger period, you need to ensure that Scheduled Task Configuration Item whether the number of minutes set in ) will be triggered, for example, your configuration item is set to trigger on the 10th minute of every hour (expression: * * * 10), while the timing trigger is set to trigger every 20 minutes 1 time (0 1/20 * * * * *), then this configuration item will never be triggered.

注意

阿里云正式版最低触发间隔为1分钟，支付宝云与腾讯云最低触发间隔为1秒
Since the cloud function can run for a maximum of 10 minutes at a time, if you want to reset the trigger time in the scheduled task after enabling the minute-level scheduled task, it is best to ensure that the trigger interval between the scheduled tasks is greater than or equal to 10 minutes. , to prevent the problem of running timeout. The default scheduled task type and trigger time can refer to the following [Scheduled Task Configuration Description](#%E5%AE%9A%E6%97%B6%E4%BB%BB%E5%8A%A1%E9%85%8D%E7 %BD%AE%E8%AF%B4%E6%98%8E).

Here we take the example of modifying the batch running cycle to trigger once every 10 minutes. The correct steps are:

Modify [uni statistics configuration item](#%E5%85%AC%E5%85%B1%E6%A8%A1%E5%9D%97%E9%85%8D%E7%BD%AE%E9% A1%B9%E8%AF%B4%E6%98%8E) Change the value of the cronMin parameter to true.
Modify the timing trigger configuration item in the package.json file under the Scheduled Task Cloud Function (uni-stat-cron). For the specific description of the timing trigger, please refer to the official document [Timing Trigger](https 😕/uniapp.dcloud.io/uniCloud/trigger.html).

"cloudfunction-config": {
	"concurrency": 1,
	"memorySize": 256,
	"timeout": 600,
	"triggers": [
		{
			"name": "uni-stat-cron",
			"type": "timer",
			"config": "0 */10 * * * * *"//每隔10分钟触发1次的cron表达式，如需复制此项请务必删除该注释
		}
	]
}

Check whether the options set in the cron parameter in the configuration file (config.json) will be triggered.
Re-upload and deploy the Scheduled Task Cloud Function (uni-stat-cron) and Configuration Center (uni-config-center).

# Enable redis cache

After enabling the redis cache, it can reduce the database query pressure and improve the uni statistical performance. You can decide whether to enable it on demand.

attention

Before enabling the redis cache, you need to confirm whether the redis service has been purchased in the service space where uni statistics are deployed. If not, you need to purchase the redis service first.

Open steps:

Modify [uni statistics configuration item](#%E5%85%AC%E5%85%B1%E6%A8%A1%E5%9D%97%E9%85%8D%E7%BD%AE%E9% A1%B9%E8%AF%B4%E6%98%8E) Change the value of the redis parameter to true.
Add the redis extension library to the package.json file under the data Report data receiver (uni-stat-receiver) and Scheduled task cloud function (uni-stat-cron) respectively.
Re-upload the deployment data Report data receiver (uni-stat-receiver), Scheduled task cloud function (uni-stat-cron) and Configuration center (uni-config-center).

Configure the redis extension library of uni-stat-receiver

Configure the redis extension library of uni-stat-cron

{
	"name": "uni-stat-receiver",
	"dependencies": {
		"uni-id": "file:../../../../uni-id/uniCloud/cloudfunctions/common/uni-id",
		"uni-stat": "file:../common/uni-stat"
	},
	"extensions": {
		"uni-cloud-jql": {},
		"uni-cloud-redis": {} // 配置为此云函数开启redis扩展库，值为空对象留作后续追加参数，暂无内容。如拷贝此配置项到package.json文件，切记去除注释。
	}
}

# Associated service space

In order for the data collected by the client app to be correctly received and counted by the cloud function in uni-admin, it is necessary to ensure that the client project and the admin project are associated with the same service space.

Select the client project (the project that needs to collect user data)
If uniCloud has not been enabled before, right click and select Create uniCloud cloud development environment -> Alibaba Cloud|Tencent Cloud; otherwise, go to step 3;

关联前后台数据

Right-click on the uniCloud directory and select Associate cloud service space or project, and select the service space associated with the uni-admin project in the opened window

关联前后台数据

# 错误解析 uni-admin 1.9.4+

This function only supports web platform, App platform, WeChat MiniApp platform

In order to facilitate developers to use sourceMap files to locate code problems, the error analysis function of using sourceMap in statistics has been added.

# Using the environment

HBuiderX 3.5.3+
不支持 IE，不支持hbx内置浏览器，推荐使用谷歌浏览器

# Generate sourceMap

The online running terminal may report errors in various user environments, which require statistics and analysis by developers. However, the three-party statistical system (such as Youmeng, Aladdin, and Baidu) collects the error reports on the running side, and it prompts the number of error lines in the code compiled by the uni-app compiler, which is equivalent to garbled characters, and cannot tell the developer that it is uni-app Which line of code in vue or js reports an error.

The sourceMap function of uni statistics can solve this problem. In the statistics background, you can clearly see the error reporting environment and accurate error code, which is the specific information of vue or js before uni-app compilation.

Generate sourceMap file in HBuiderX
- For the web platform, check the Generate sourceMap option in Release -> Website.
- When the App platform is Release -> Native App-Cloud Packaging, check the Generate sourceMap option.
- When the WeChat MiniApp platform is Issue -> MiniApp- WeChat, check the Generate sourceMap option.

cli project generates sourceMap file

yarn build:h5 --sourcemap

yarn build:app-plus --sourcemap

yarn build:mp-weixin --sourcemap

After the project is compiled, you can view the generated sourceMap files for each platform in /unpackage/dist/build/.sourcemap.
After the WeChat MiniApp platform is uploaded and released, it will be compressed and packaged again, so an additional step is required:
1. Download the online version sourceMap file in Development Management/Operation and Maintenance Center/Error Log of WeChat Public Platform.
2. Unzip the downloaded file into the generated .sourcemap folder: /unpackage/dist/build/.sourcemap/mp-weixin/__WEIXIN__/ (WEIXIN is the newly created directory, and the parsing error will be based on this name lookup)
Precautions
1. If there is only __FULL__ folder in the downloaded sourceMap file, please rename it to __APP__
2. Only support release mode sourceMap file error parsing

# upload sourceMap

uni-admin 2.3.4+ 版本

注意：2.3.4+ 的版本同时支持阿里云和腾讯云空间。

将 uni-admin 项目运行到浏览器中，在 uni 统计 / 错误统计 / js报错 页面，错误信息列表表格的右上角有 上传 sourceMap 按钮。点击后展示如下：
上传请前请填写完整信息：应用、平台、版本
点击 选择文件并上传 按钮，弹出 选择文件夹 框后，请选中编译的对应该平台版本的 sourceMap。如：项目根目录/unpackage/dist/build/.sourcemap/h5，选中 h5 目录后点击上传。如果中途上传失败，在不刷新页面的情况下，重新选择文件夹上传可以跳过已上传文件。
上传完成

uni-admin 2.3.4 之前的版本

注意：2.3.4 之前的版本只支持腾讯云空间，不支持阿里云空间。

Configure uni-stat related information in uni-admin project/admin.config.js:
1. Search uploadSourceMapCloudSpaceId to supplement Tencent Cloud Service Space SpaceID
2. Search for cloudSourceMapUrl to supplement Tencent Cloud cloud storage access address (eg: https://xx-xx-xx.tcb.qcloud.la/__UNI__/uni-stat/sourcemap)
Run the uni-admin project into the browser, and on the uni statistics / error statistics / js error report page, there is an upload sourceMap button in the upper right corner of the error message list table. After clicking, it will display as follows:
Please fill in the complete information before uploading: Application, Platform, Version
Click the Choose file and upload button, after the Select Folder box pops up, please select the compiled sourceMap corresponding to the platform version. For example: Project root directory/unpackage/dist/build/.sourcemap/h5, select the h5 directory and click upload. If the upload fails in the middle, without refreshing the page, you can skip the uploaded file by re-selecting the folder to upload.
The contents of the folder will be uploaded to the cloud storage/__UNI__/uni-stat/sourcemap/application appId/platform (eg web, mp-weixin, ios)/version/ directory

Precautions

Cloud storage needs configuration permissions: If the Tencent Cloud service space used is not bound by the current project, it needs to be set to: All users can read
If there is cross-domain, you need to bind the security domain name in cross-domain configuration
The upload failure may be due to the following reasons:
1. If the file name, folder name contains ad or an advertisement field, please close the browser ad blocking extension and upload it again
2. If there is a file with the same name in this directory in the cloud storage, the upload may also fail. Therefore, before uploading the same platform and the same version of the sourceMap file again, please refer to the corresponding directory in the cloud storage (eg: cloud storage/__UNI__/uni-stat/sourcemap/appId/platform (eg: web, mp-weixin, ios)) /version/) delete

# Parse error

On the uni statistics / error statistics / js error report page, in the error information list table, click the details button on the right side of the table row to view the parsed error information in the pop-up window.

Example

Original error message:

原始错误：

Parsed error message:

原始错误：

Parsing errors are parsed line by line. If a line fails to parse, the original error message will be returned
runtime error Below the separation line is the runtime error stack information of the platform framework, you don't need to care about it

# Open source code interpretation

Front-end collection SDK source address, View
Cloud statistics module and source address of statistics background, View

# Front-end collection SDK

# Data reporting logic

The default data reporting interval is 10s. During the reporting interval, the data of each reporting node will be added to the statistics data queue. After 10s, the next reporting node will uniformly process the data queue and report it.

To save server-side resources, the data collected by the front-end is reported on a regular basis. So the escalation request does not happen in real time.

HBuilderX 3.5.4 or higher version can adjust the reporting period according to your own business needs, you can modify it in uni statistics configuration -> front-end data reporting period, in order to save server resources, it is recommended to fill in the range of 5- between 20.

When will the data be reported?

After the code is released, run the project report (HBuiderX -> Release -> Select the platform to open statistics)
Run the code with debug mode enabled and report (manifest.json -> uniStatistics -> debug:true)

In addition to the above two cases, data reporting will not occur in other cases, including App real machine runtime (including custom dock), crash and error data

# Collection Type

App launch

The program starts at the beginning of the visit, and the end of the visit is as follows: entering the background for more than 5 minutes, no operation in the foreground for more than 30 minutes, and a new source program

Reporting Fields	Description
lt	Statistical data type, the default value is 1, `type see below`
ut	Platform type, `type see below`
mpsdk	MiniApp SDK version
mpv	MiniApp platform version, such as WeChat, Alipay, etc.
mpn	Native platform package name, MiniApp appid
v	Application version. Native application
p	Mobile phone system, `type see below`
net	Network type, `type see below`
brand	Mobile phone brand
md	Mobile phone model
lang	language
lat	Latitude
lng	Longitude
pr	pixelRatio Device pixel ratio
ww	windowWidth can use window width
wh	windowHeight can use the window height
sw	screenWidth screen width
sh	screenHeight screen height
url	The full url of the current page, including parameters. Up to 255 characters
ch	Channel Information
fvts	First Visit Timestamp
lvts	Last Access Timestamp
en	Country
pn	province
ct	City
sc	Scene Value
tvc	The total number of visits by the user up to this visit
usv	stats sdk version
t	Timestamp when data was reported

App goes to background

When the application enters the background, it is triggered by the application's onHide life cycle in the SDK

Reporting Fields	Description
lt	Statistical data type, the default value is 3, `type see below`
ut	Platform type, `type see below`
p	Mobile system, `type see below`
urlref	Page where the app exits
urlref_ts	The duration of the last page when the app exits
ch	Channel Information
usv	stats sdk version
t	Timestamp when data was reported

Page switch

Report when the page jumps. In the SDK, it is triggered by the onHide life cycle of the page.

Reporting Fields	Description
lt	Statistical data type, the default value is 11, `type see below`
ut	Platform type, `type see below`
p	Mobile phone system, `type see below`
url	The full url of the current page, including parameters. Up to 255 characters
ttpj	title of the page defined in pages.json
ttn	title set via API uni.setnavigationbartitle
ttc	The title of the page reported via uni.report
ttct	title set in the title component
urlref	Page where the app exits
urlref_ts	The duration of the last page when the app exits
ch	Channel Information
usv	stats sdk version
t	Timestamp when data was reported

Event trigger

When the user triggers some business logic

Default event
- Login: User Information
- Payment: product name, amount
- share:
User defined events

Reporting Fields	Description
lt	Statistical data type, the default value is 21, `type see below`
ut	Platform type, `type see below`
p	Mobile phone system, `type see below`
url	The full url of the current page, including parameters. Up to 255 characters
e_n	Event Name
e_v	Event Parameters
ch	Channel Information
usv	stats sdk version
t	Timestamp when data was reported

Application error

Report an error in the application

Reporting Fields	Description
lt	Statistical data type, the default value is 21, `type see below`
ut	Platform type, `type see below`
p	Mobile phone system, `type see below`
ch	Channel Information
mpsdk	MiniApp SDK Version
mpv	MiniApp platform version, such as WeChat, Alipay, etc.
v	Application version. Native application
em	Error message
usv	stats sdk version
t	Timestamp when data was reported

lt: Statistics type

value	description
1	The application starts, corresponding to the `onLaunch` event
3	The application enters the background, corresponding to the application `onHide` event
11	Page jump, corresponding to page `onHide` event
21	Event Trigger
31	Application Error

ut: platform type

value	description
h5	h5
wx	WeChat
ali	Ali
bd	Baidu
qq	qq
qn	QuickApp
ks	Kuaishou
lark	Feishu
qw	QuickApp
dt	DingTalk

p: phone system

value	description
a	Android
i	iOS

net: network type

value	description
wifi	wifi network
2g	2g network
3g	3g network
4g	4g network
5g	5g network
none	No Internet
cable	cable

# Cloud Statistics

# Statistical report display page

In order to highlight the goal, only the folders and files related to uni statistics are commented out, and the rest are the same as ordinary uni-app projects. For new pages, please refer to similar pages in uni-stat.

├── cloudfunctions
├── common                             # 样式
│   │── uni.css                        # 公共样式
│   └── uni-icons.css                  # icon样式
├── components                         # 自定义组件
├── js_sdk                             # js sdk
│   └── uni-stat                       # uni统计相关工具方法
│       └── util.js
├── pages                              # 页面
│   └── uni-stat                       # uni统计页面
│       │── channel                    # 渠道（app）
│       │   │── channel.vue            # 页面（下同）
│       │   └── fieldsMap.js           # 字段配置（下同）
│       │── device                     # 设备统计
│       │   │── activity               # 渠道/场景分析
│       │   │   │── activity.vue
│       │   │   └── fieldsMap.js
│       │   │── comparison             # 平台对比
│       │   │   │── comparison.vue
│       │   │   └── fieldsMap.js
│       │   │── overview               # 今日概览
│       │   │   │── overview.vue
│       │   │   └── fieldsMap.js
│       │   │── retention              # 留存
│       │   │   │── retention.vue
│       │   │   └── fieldsMap.js
│       │   │── stickiness             # 粘性
│       │   │   │── stickiness.vue
│       │   │   └── fieldsMap.js
│       │   └── trend                  # 趋势分析
│       │       │── trend.vue
│       │       └── fieldsMap.js
│       │── error                      # 错误分析
│       │   │── error.vue
│       │   └── fieldsMap.js
│       │── event                       # 事件分析
│       │   │── event.vue
│       │   └── fieldsMap.js
│       │── index                       # 统计首页
│       │   │── index.vue
│       │   └── fieldsMap.js
│       │── page-ent                    # 入口页
│       │   │── page-ent.vue
│       │   └── fieldsMap.js
│       │── page-res                    # 受访页
│       │   │── page-res.vue
│       │   └── fieldsMap.js
│       │── page-content                # 内容统计
│       │   │── page-content.vue
│       │   └── fieldsMap.js
│       │── page-rule                   # 页面规则
│       │   │── page-rule.vue
│       │── pay-order                   # 支付统计
│           │── funnel                  # 支付/漏斗分析
│           │   │── funnel.vue
│           │   └── fieldsMap.js
│           │── list                    # 支付/订单明细
│           │   │── list.vue
│           │── overview                # 支付/订单概况
│           │   │── overview.vue
│           │   └── fieldsMap.js
│           └── ranking                 # 支付/用户价值排行
│               └── ranking.vue
│       │── scene                       # 场景值（小程序）
│       │   │── scene.vue
│       │   └── fieldsMap.js
│       └── user                        # 用户统计
│           │── activity                # 渠道/场景分析
│           │   │── activity.vue
│           │   └── fieldsMap.js
│           │── comparison              # 平台对比
│           │   │── comparison.vue
│           │   └── fieldsMap.js
│           │── overview                # 今日概览
│           │   │── overview.vue
│           │   └── fieldsMap.js
│           │── retention               # 留存
│           │   │── retention.vue
│           │   └── fieldsMap.js
│           │── stickiness              # 粘性
│           │   │── stickiness.vue
│           │   └── fieldsMap.js
│           └── trend                   # 趋势分析
│               │── trend.vue
│               └── fieldsMap.js
├── static
├── store
├── admin.config.js
├── App.vue
├── main.js
├── mainfest.json
├── pages.json
├── postcss.config.js
└── uni.scss

# Cloud function, general module description

1. Composition of uni statistics server

uni-config-center/uni-stat configuration module: Provides the necessary configuration parameters for uni stats to run.
uni-stat public module: data processing module, including the processing and storage of the collected and reported data and the data processing of timed tasks.
uni-stat-receiver report data receiver: Receive the data reported by the client and forward it to the public module for processing. Note: This cloud object depends on the uni-id public module.
uni-stat-cron timed task cloud function: triggers timed tasks and forwards them to public modules for processing.

2. Description of public modules

attention

Note: The uni statistics common module depends on the uniCloud configuration center (uni-config-center)

├── shared                              # 公共模块，提供公共函数库等支持。
│   │── create-api.js                   # 用来创建对外访问的实例
│   │── error.js                   		# 错误处理模块
│   │── index.js                   		# 入口文件，提供对外访问的基础模块
│   └── utils.js                     	# 工具函数库文件
├── stat                                # uni统计实际业务处理模块
│   │── lib                             # 工具类类库，提供日期计算、数据加密等额外功能支持。
│   │   │── date.js                     # 日期计算类文件
│   │   │── index.js                    # 入口文件，提供对外访问模块
│   │   └── uni-crypto.js               # 数据加密类文件，提供AES/MD5加密
│   │── mod                             # 数据模型，提供具体业务实现。
│   │   │── uni-pay                     # 支付统计目录
│   │   │   └── payResult.js            # 支付统计模型，跑批支付统计数据。
│   │   │── activeDevices.js            # 活跃设备模型，给周月维度的设备基础统计和留存统计提供数据，每日跑批合并，仅添加本周/本月首次访问的设备。
│   │   │── activeUsers.js              # 活跃用户模型，给周月维度的用户基础统计和留存统计提供数据，每日跑批合并，仅添加本周/本月首次访问的用户。
│   │   │── appCrashLogs.js             # 原生应用崩溃日志模型，记录原生应用的崩溃日志
│   │   │── base.js                     # 基类模型，提供基础服务支持
│   │   │── channel.js                  # 渠道模型，提供渠道和场景值数据
│   │   │── errorLog.js                 # 错误日志模型，记录上报的应用运行错误日志
│   │   │── errorResult.js              # 错误结果统计模型，统计汇总错误日志中的数据
│   │   │── event.js                    # 事件统计模型，提供应用的事件字典
│   │   │── eventLog.js                 # 事件日志模型，记录上报的事件日志
│   │   │── eventResult.js              # 事件结果统计，统计汇总事件日志中的数据
│   │   │── index.js                    # 入口文件，提供对外访问模块
│   │   │── loyalty.js                  # 设备/用户忠诚度（粘性）统计模型，统计设备/用户的粘性，粘性判断依据为：访问时长和访问页面数量
│   │   │── page.js                     # 页面模型，提供应用的页面字典
│   │   │── pageDetail.js               # 页面内容模型，提供页面规则匹配到的页面详情字典
│   │   │── pageDetailResult.js         # 页面内容统计模型，汇总统计页面内容的访问设备、时长、分享等数据
│   │   │── pageLog.js                  # 页面日志模型，记录上报的页面访问日志
│   │   │── pageResult.js               # 页面结果统计模型，统计汇总页面访问日志中的数据
│   │   │── platform.js                 # 应用平台模型，提供应用的平台字典
│   │   │── runErrors.js                # 运行错误日志，记录数据统计时运行出错的日志
│   │   │── scenes.js                   # 场景值模型，提供应用渠道和小程序场景值的数据字典
│   │   │── sessionLog.js               # 基础会话日志模型，记录设备访问时产生的会话日志
│   │   │── shareLog.js                 # 分享日志模型，记录触发分享事件的日志
│   │   │── statResult.js               # 基础数据结果统计模型，统计汇总会话数据包括不限于设备/用户的数量、访问量、活跃度（日活、周活、月活）、留存率（日留存、周留存、月留存）、跳出率、访问时长等数据
│   │   │── uniIDUsers.js               # uni-id 用户模型，提供uni-id用户数据查询
│   │   │── userSessionLog.js           # 用户会话日志模型，记录登录用户的会话日志
│   │   └── version.js                  # 应用版本模型，提供应用的版本号字典
│   │── receiver.js                     # 上报数据接收器，数据上报功能的分发入口文件
│   └── stat.js                         # 数据统计调度处理模块，数据统计及日志清理功能的分发入口文件
└── index.js                            # 代理入口文件，提供对外访问的uni-stat对象

# Common module configuration item description

The uni statistics configuration items are stored in the uni-stat/config.json file under the uniCloud configuration center (uni-config-center). Users can customize the value of each configuration item according to their own system needs.

attention

Note: After modifying the uni statistics configuration items, you need to re-upload the public module uni-config-center to take effect.

Basic parameters

配置项	默认值	说明
debug	false	开启调试模式 true: 开启，false:关闭，开启后会产生大量日志，生产环境请关闭。
redis	false	开启redis缓存，开启后可以降低数据库查询压力，提升uni统计性能，可按需决定是否开启。开启方法
cachetime	604800	redis缓存有效期，单位秒。
sessionExpireTime	1800	会话过期时间，该配置用来判断当前会话是否已过期，一般情况下无需修改此项。
realtimeStat	true	开启实时统计，true: 开启，false:关闭，开启后会每小时统计一次，数据库读写次数会增多，可按需决定是否开启。
cronMin	false	开启分钟级定时任务，true: 开启，false:关闭。开启后定时任务将细分到分钟级执行，分摊数据计算压力，适合应用日活较大或有特殊需求的用户群体。具体的开启方法见设置定时任务云函数的触发周期。
cron	-	用于配置定时任务触发时间，详情见下方定时任务配置说明。
pageDetailStat	false	开启页面内容统计 true: 开启，false:关闭，开启后会产生大量日志，请按需开启。uni-admin 2.4.0起支持配置此项。
batchInsertNum	5000	当有批量写入操作时，限制单次写入数据库的最大条数。为防止写入超时，最大值为5000条。
errorCheck	-	错误检测，此项用于在规定时间内限制相同的错误日志写入数据库，防止有高频错误产生时造成大量的数据库写入操作。详情
cleanLog	-	日志清理，此项用于配置定时清理过期的日志，减少数据库数据的存储量，提升uni统计性能。详情

# Scheduled task configuration instructions

The cron parameter is used to configure the trigger time of the scheduled task. Generally, you do not need to modify this item.

Parameters	Description
type	Scheduled task type: such as `stat`: basic data statistics
time	Trigger time expression: `* * * ` There are four digits in total, representing from left to right: week (1-7 represents Monday to Sunday)/day/hour/minute. Example: Triggered at 0:00 every night, it should be written as ` * 0 0`

The current types of scheduled tasks are (The content in brackets below indicates the trigger time of the scheduled task after the minute-level statistics are enabled):

stat: basic data statistics, statistical dimensions include:
- Real-time statistics, triggered by default every hour (0 minutes), statistics of basic data for the last hour
- Daily statistics, the default is triggered at 1:00 a.m. (10 minutes) every day, and the basic data of the previous day is counted
- Weekly statistics, triggered by every Monday at 1:00 am (20 minutes) by default, statistics the basic data of the previous week
- Monthly statistics, the default triggered at 3:00 am (30 minutes) on the 1st of each month, statistics the basic data of the previous month
retention-device: Device retention data statistics, statistical dimensions include:
- Daily statistics, the default is triggered by every day at 2:00 am (20 minutes), and the daily retained data of the device is counted (the initial statistics of the retention status from the day before yesterday to yesterday, so it takes at least three days to have data)
- Weekly statistics, triggered by every Monday at 2:00 am (30 minutes) by default, to count the weekly retained data of the device (the initial statistics of the retention from the last week to the last week, so it takes at least three weeks to have data)
- Monthly statistics, triggered by the default 4:00 am (30 minutes) on the 1st of each month, to count the monthly retained data of the device (the initial statistics are about the retention of the previous month to the previous month, so it will take at least three months to have data)
retention-user: User retention data statistics, statistical dimensions include:
- Daily statistics, triggered by every day at 3:00 a.m. (40 minutes) by default, to count the daily retained data of users (the initial statistics of retention from the day before yesterday to yesterday, so it takes at least three days to have data)
- Weekly statistics, triggered by every Monday at 5:00 am (30 minutes) by default, and statistics the user's weekly retention data (the initial statistics of the retention status from the last week to the last week, so there will be data after at least three weeks of operation)
- Monthly statistics, triggered by the default 6:00 am (40 minutes) on the 1st of each month, to count the monthly retention data of users (the retention status from the previous month to the previous month is initially counted, so it takes at least three months to have data)
active-device: Active device data archive, statistical dimensions include:
- Daily archiving, triggered by every day at 0:00 am (10 minutes) by default, and archives the active device data of the previous day. Note: this data should be kept in basic data statistics, device retention data statistics, user retention data Executed before the weekly statistics and monthly statistics in Statistics` are triggered.
active-user: Active user data archive, statistical dimensions include:
- Daily archiving, triggered by every day at 0:00 am (20 minutes) by default, and archives the active user data of the previous day. Note: this data must be kept in basic data statistics, device retention data statistics, user retention data Executed before the weekly statistics and monthly statistics in Statistics` are triggered.
page: page data statistics, statistical dimensions include:
- Daily statistics, the default is triggered at 3:00 a.m. (20 minutes) every day, and the page data of the previous day is counted
page-detail：页面内容统计，统计维度包括：
- 日统计，默认每天上午7点（10分钟）触发，统计前一天的页面内容数据
event: Event data statistics, statistical dimensions include:
- Daily statistics, the default is triggered at 4:00 a.m. (20 minutes) every day, and the event data of the previous day is counted
error: Error data statistics, statistical dimensions include:
- Daily statistics, the default triggered at 5:00 a.m. (20 minutes) every day, statistics the error data of the previous day
clean：过期日志清理定时任务：
- Log cleanup, triggered by default every day at 5:00 am (30 minutes), to clean up expired log data
pay-result: Payment data result statistics, statistical dimensions include:
- Real-time statistics, triggered by hourly (10 minutes) by default, statistics of the basic data of the last hour (during real-time statistics, the data of hour, day, week, month, quarter, and year will be automatically counted, no need to configure other dimensions statistics)

# Error detection configuration instructions

The errorCheck parameter is used to limit the same error log to be written to the database within a specified time to prevent a large number of database write operations caused by high-frequency errors. It can be turned on or off as needed.

Parameters	Description
needCheck	Need to check: true: yes; false: no
checkTime	Error check interval time, in `minutes`.

# Log cleaning configuration instructions

The cleanLog parameter is used to configure regular cleaning of expired logs, reduce the storage capacity of database data, and improve uni statistics performance.

Parameters	Description
open	Whether to enable log cleaning: true: yes; false: no
reserveDays	Retention days configuration for each log, parameter format: `log type: retention days`, for example: `sessionLog: 31` means to reserve session logs for 31 days, and if the retention days is set to 0, it means permanent retention (this will Accumulate a lot of useless data, not recommended)

The currently configurable log types are:

Basic session log: sessionLog, keep logs for 31 days by default. Note: Because the device retention statistics need to count the retained data after 30 days at the longest, the basic session logs must be stored for at least 31 days, otherwise the device retention statistics will be affected.
User session log: userSessionLog, keep logs for 31 days by default. Note: Because the user retention statistics need to count the retained data after 30 days at the longest, the user session log must be kept for at least 31 days, otherwise it will affect the user retention statistics.
Page log: pageLog, keep logs for 7 days by default.
Event log: eventLog, keep logs for 7 days by default.
Share log: shareLog, keep logs for 7 days by default.
Error log: errorLog, keep logs for 7 days by default.

Notes

The two projects of the client and the statistics background must be associated with the same service space, and all cloud functions, public templates, etc. in uni-admin have been uploaded and deployed to the service space
To use uni statistics, APPID must be configured for normal use. What is the use of DCloud's Appid, how to transfer the application
The application will not report statistical data during running and debugging. It will only report data after launching the new version of the app, h5, and MiniApp.
CLI projects are not supported

# Extensions and customizations

uni statistics provides basic data reports. If the expected data collection cannot be achieved, you can freely report data through uni.report(eventKey,param) on the client side, and add pages through uni-admin to collect statistics by yourself.

# 前端 uni.report 基础用法

The basic usage of uni.report(eventKey,param) is listed here, the complete API view: Details

uni.report(eventKey,param) takes two parameters.

eventKey custom event name
param custom event parameter

// parameter supports string
uni.report('购买','购买成功')

// parameter support object
uni.report('购买',{
	id:'1000',
	name:'上衣',
	price:'998',
	msg:'购买成功'
	// ...
})

# Data reported using custom events in uni-admin

Add new page in uni-admin, use uniCloud API to get all uni-report The reported raw data is stored in the uni-stat-event-logs table.

After getting the data, you can expand the required business logic by yourself

const db = uniCloud.database()
db.collection('uni-stat-event-logs')

不管是新上报的数据，还是opendb的数据、或者开发者自己的业务数据库里的数据，都可以自己编写报表进行统计分析。

Like ordinary uni-app pages, create a new page and write code.

When registering the left menu in uni-admin, you need to refer to the document: [uni-admin left window - menu bar](https://uniapp.dcloud.io/uniCloud/admin.html#%E5%B7% A6%E4%BE%A7%E7%AA%97%E5%8F%A3-%E8%8F%9C%E5%8D%95%E6%A0%8F)

# Version upgrade

# 升级到1.9.0+ 版本

uni-admin1.9.0 version starts (corresponding to HBuilderX 3.5.1), uni statistics and uni uni-upgrade-center center reuse the same application version table (ie opendb-app-versions table), discard the original The uni-stat-app-versions table.

If you have enabled uni-statistics 2.0 but are using an older version of uni-admin, you need to pay attention to the migration and upgrade of the version table, otherwise the version-based statistics may be inaccurate.

Upgrade steps:

Update the uni-admin project from the plugin market
Upload overlay uni-stat public module
Re-upload overwriting all DB Schema
Import the data in the original uni-stat-app-versions table into the opendb-app-versions table; a code snippet is provided below, you can create a cloud function and copy the sample code to the cloud function , right-click to execute "Run - Local Cloud Function" to automatically complete the migration of data content;

'use strict';
const db = uniCloud.database()
exports.main = async (event, context) => {
	const oldVersionDataRes = await db.collection('uni-stat-app-versions')
		.aggregate()
		.lookup({
			from: 'uni-stat-app-platforms',
			localField: 'platform_id',
			foreignField: '_id',
			as: 'platform'
		})
		.limit(500)
		.end()

	if (oldVersionDataRes.data.length) {
		for (let oldKey in oldVersionDataRes.data) {
			// data of the old version table
			const oldVersionData = oldVersionDataRes.data[oldKey]

			//Assemble data
			const newVersionData = {
				_id: oldVersionData._id,//_id 两个表数据要保持一致
				appid: oldVersionData.appid,//appid
				platform: [],//默认为空数组即可
				uni_platform: oldVersionData.platform[0].code,//平台代码
				type: 'native_app',//类型 默认为native_app即可
				version: oldVersionData.version,
				create_env: 'uni-stat',//创建来源，设置为uni-stat
				create_date: oldVersionData.create_time//创建时间
			}
			await db.collection('opendb-app-versions').add(newVersionData)
		}
	}

	return true
};

# 升级到2.4.0+ 含内容统计的版本

uni-admin2.4.0 版本开始，新增了内容统计，可以通过设置页面规则进行内容统计。内容统计模块默认处于关闭状态，开启内容统计后会增加大量的数据库读写次数，请按需开启。开启内容统计的方式，可参考公共模块配置项说明。

演示图

内容统计

页面规则

设置页面规则

注意

如果升级uni-admin后，你的uni统计没有出现内容统计的菜单，则点击菜单管理里的【更新内置菜单】按钮即可（需uni-admin版本>=2.4.0）

# uni statistics 2.0 cost evaluation

# Preface

Recently, the uniCloud Alibaba Cloud version has been officially commercialized. Some developers have begun to struggle with cloud-integrated services such as `uni statistics' based on uniCloud. They don't know how much these services are expected to cost. The program is more cost-effective.

本文尝试算细账、算总账，以阿里云按量计费为例，详细预测uni统计在不同用户规模下的资源消耗及对应费用，帮助大家明智选择，无忧开发。

This article is mainly divided into three parts:

Estimation of resource costs consumed by uni statistics
Benefits brought to you by uni statistics
Comprehensive consideration, how should you choose

Uni Statistics Consumed Resource Cost Calculation

The parts of uni statistics involving expenses are mainly divided into:

Cloud function: uni statistics cloud function, there are 2 cloud functions
- uni-stat-receiver client data reporting function (add statistical data source)
- uni-stat-cron data run batch processing function (generate statistical data)
Cloud database: tables prefixed with uni-stat-
Front-end website hosting: deploy uni-admin, the administrator releases a new version

Next, we conduct cost assessments for different resources.

# cloud function

# uni-stat-receiver

After enabling uni statistics, each of your online users will request the uni-stat-receiver cloud function every 10 seconds by default (if your DAU is more than 10,000, you can change it to 60 seconds, which can reduce the cost. Time The interval can be set in manifest.json, if the user stays on one page, it will not be reported repeatedly)

我们按照uniCloud官网列出的按量计费规则，计算一下云函数的资源消耗。

Resource classification	Resource details	Price (yuan)
Cloud Function	Resource Usage (GBs)	0.000110592
	Number of calls (10,000 times)	0.0133
	Outbound traffic (GB)	0.8
Cloud Database	Capacity (GB/day)	0.07
	Number of read operations (10,000 times)	0.015
	Number of write operations (10,000 times)	0.05
Cloud storage	Capacity (GB/day)	0.0043
	Number of download operations (10,000 times)	0.01
	Number of upload operations (10,000 times)	0.01
	CDN Traffic (GB)	0.18
Front-end website hosting	Capacity (GB/day)	0.0043
	Traffic (GB)	0.18

We can simply derive the following formula:

Cloud function/cloud object cost = resource usage * 0.000110592 + number of calls * 0.0133 / 10000 + outbound traffic * 0.8

in:

Resource usage = cloud function memory (in G) * average execution time of a cloud function (in seconds) * number of calls
Number of calls = daily activity of the application * average daily number of reports by daily active users

We assume the following data model:

Cloud function running memory: 128M, which is 0.125G (cloud function memory is 512M by default, users can customize the settings, the minimum can be set to 128M)
Average number of reports per day by active users: 10 times
The average execution time of a cloud function: 100 milliseconds, or 0.1 seconds
Single request outbound traffic: 0.7 KB

According to the above formula, if there are 100 daily active users, the daily cost of the uni-stat-receiver cloud function of uni statistics is:

云函数费用（天） = 资源使用量 * 0.000110592  + 调用次数 * 0.0133 / 10000 + 出网流量 * 0.8
				= 云函数内存（单位为G） * 云函数平均单次执行时长（单位为秒） * 调用次数 * 0.000110592 + 调用次数 * 0.0133 / 10000 + 出网流量（单位GB） * 0.8
				= 0.125 * 0.1 * (100 * 10) * 0.000110592 + (100 * 10) * 0.0133/10000 + (100 * 10) * (0.7 / 1024 / 1024) * 0.8
				= 0.001382 + 0.00133 + 0.000534
				≈ 0.0032（元）

That is: your App DAU is 100, and after using the commercial version of uni statistic, the uni-stat-receiver cloud function consumes about 0.0032 yuan per day.

According to this, the monthly cost can be calculated as: 0.0032 * 30 ≈ 0.1, that is, when the daily activity is 100, the monthly uni-stat-receiver cloud function only needs 0.1 yuan.

In the same way, we can deduce that for apps with 1,000, 10,000, and 100,000 daily activities, the monthly cost of the uni-stat-receiver cloud function is as follows:

日活	资源使用量计费（元/日）	调用次数计费（元/日）	出网流量计费（元/日）	日合计（元/日）	月合计（元/月）
100	0.001382	0.00133	0.000534	0.0032	0.10
1000	0.01382	0.0133	0.00534	0.0325	0.98
10000	0.1382	0.133	0.0534	0.3246	9.74
100000	1.382	1.33	0.534	3.2460	97.38

# uni-stat-cron

uni statistic also has a cloud function uni-stat-cron, which is a scheduled data running batch task, which is used to count the reported data and generate statistical reports.

We assume the following data model:

Cloud function memory: 512M, which is 0.5G (it is recommended to set it to 512M for running batch cloud functions, because it only runs 24 times a day, and the larger the memory, the stronger the performance)
The average execution time of a cloud function: 1 second (the more data sources, the longer the running time)
Daily execution times: 24 times (fixed to run once per hour)
Outbound traffic: not returned to the client, fixed at 0

The daily cost of the uni-stat-cron cloud function of uni statistics is:

云函数费用（天） = 资源使用量 * 0.000110592  + 调用次数 * 0.0133 / 10000 + 出网流量 * 0.8
				= 云函数内存（单位为G） * 云函数平均单次执行时长（单位为秒） * 调用次数 + 调用次数 * 0.0133 / 10000 + 出网流量 * 0.8
				= 0.5 * 1 * 24 * 0.000110592 + 24 * 0.0133/10000 + 0
				= 0.001327104 + 0.00003192 + 0
				≈ 0.0014（元）

That is: your App DAU is 100, and after using the commercial version of uni statistic, the uni-stat-receiver cloud function consumes about 0.0014 yuan per day.

According to this, its monthly cost can be calculated as: 0.0014 * 30 ≈ 0.04, that is, when the daily activity is 100, the monthly uni-stat-receiver cloud function only needs 0.04 yuan.

In the same way, we can deduce that for apps with 1,000, 10,000, and 100,000 daily activities, the monthly cost of the uni-stat-receiver cloud function is as follows:

日活	平均耗时（秒）	资源使用量计费（元/日）	调用次数计费（元/日）	日合计（元/日）	月合计（元/月）
100	1	0.001327	0.00003192	0.0014	0.04
1000	3	0.003981	0.00003192	0.0040	0.12
10000	30	0.039813	0.00003192	0.0398	1.19
100000	90	0.119439	0.00003192	0.1195	3.59

Since the uni-stat-cron cloud function only runs 24 times a day regardless of the daily activity, the daily activity has little impact on its cost (only the duration of each operation is affected).

# cloud database

按照uniCloud官网介绍，云数据库费用 = 容量费用 + 读操作次数费用 + 写操作次数费用，其中：

Capacity fee: database storage capacity (in G) * 0.07
Fee for the number of read operations: number of read operations (10,000 times) * 0.015
Number of write operations: number of write operations (thousands) * 0.05

uni statistics will generate a large amount of log data, but by default there will be a strategy for automatically clearing historical logs, for example: session log data 31 days ago will be deleted.

Since uni statistics involves a very complicated database, we have obtained the following results through the actual operation data of the official statistics sample project:

For an application with a DAU of 100, the resource usage of uni statistics database is as follows:

Average number of reports per day by active users: 10 times
On average, each report requires 2 database reads and 2 writes (an average of 1.1 new records and 0.9 modified records), so 1.1 new records are added and inserted, about 0.54KB (Note: In many cases, there may be no reports New data is added, just modify the session log data.)

The above figures are calculated from the Official Statistics Sample Project.

Therefore, the following data model can be obtained:

Average daily cloud database capacity: 0.00052G (= 100 * 10 * 0.54 / 1024 / 1024)
Daily average cloud database reading times: 2000 (= 100 * 10 * 2)
Average number of daily cloud database writes: 2000 (= 100 * 10 * 2)

Because the data source is only kept for 31 days, the data volume will remain at 31 days after stable operation. Therefore, the cloud database capacity in the formula below needs to be multiplied by 31

数据库费用（天） = 云数据库容量 * 31 * 0.07   + 读操作次数 * 0.015 / 10000 + 写操作次数 * 0.05/10000
				= 0.00052 * 31 * 0.07  + 2000 * 0.015/10000 + 2000 * 0.05/10000
				= 0.0011284 + 0.003 + 0.01
				≈ 0.0141（元）

That is: if your uni statistics business DAU is 100, after using the cloud service space of Alibaba Cloud Commercial Edition, the corresponding database will consume about 0.0141 yuan per day. Based on this, the monthly cost can be calculated as: 0.0141 * 30 ≈ 0.42, that is, when the DAU is 100, the monthly cloud database only needs 0.42 yuan.

Daily activity	Capacity fee (G)	Read operation fee (yuan/day)	Write operation fee (yuan/day)	Day total (yuan/day)	Month total (yuan/month)
100	0.0011284	0.003	0.01	0.0141	0.42
1000	0.011284	0.03	0.10	0.1413	4.24
10000	0.11284	0.30	1.00	1.4128	42.38
100000	1.1284	3.00	10.00	14.1284	423.85

# Cloud database with redis cost comparison

Redis only affects the read operation of the database. According to the comparison before and after using redis in the official statistical sample project, it can be seen that redis can reduce the number of database read operations by about 2/3 (equal to reducing the cost of database read times by 2/3).

daily activity	cost of read operations (yuan/month) (redis not enabled)	cost of read operations (yuan/month) (redis enabled)	cost reduction per month after redis is enabled (yuan)
100	0.09	0.03	0.06
1000	0.90	0.30	0.60
10000	9.00	3.00	6.00
100000	90.00	30.00	60.00

It can be seen that when the daily activity is lower than 100,000, the cost reduction of redis is not obvious. But when the daily activity is greater than 100,000, the role of redis becomes more and more obvious.

# Front-end web hosting

uni statistics needs to be used in conjunction with uni-admin, and uni-admin needs to be deployed in the front-end web hosting. uni-admin is mainly used by administrators, with less frequent use and lower traffic.

按照uniCloud官网介绍，前端网页托管费用 = 容量费 + 流量费。

# Capacity Fee

uni-admin is compiled to 4.7M. According to the rule of 0.0043 yuan per GB per day on the official website, the monthly capacity fee of uni-admin is: 4.7 / 1024 * 0.0043 * 30 = 0.00059, which is less than 1 cent. Ignorable.

# traffic fee

The administrator logs in to uni-admin and browses the statistical data on the uni statistics management page. The traffic required does not exceed 3M. Even if updates are released twice a month, the estimated traffic fee is: 3 / 1024 * 0.18 * 2 = 0.00105, which is less than 1 cent, can also be ignored.

# Summary of front-end web hosting

The monthly fee is less than 1 cent, which can be ignored.

# 费用合并

After comparing the detailed items, let's combine them to see how much they need to spend per month using uni statistics.

Daily Active	Cloud Function (RMB/month)	Cloud Database (RMB/month)	Front-end Web Hosting (RMB/month)	Monthly Total (RMB/month)
100	0.14	0.42	可忽略不计	0.56
1000	1.10	4.24	可忽略不计	5.34
10000	10.93	42.38	可忽略不计	53.31
100000	100.97	423.85	Negligible	524.82

# uni statistics bring you benefits

Using uni statistics, free access, one-click installation, you will get:

1. Full port

Full terminal traffic statistics, one report can view the operation data of all terminals (iOS, Android, Web and various MiniApp).

There is no need to connect different SDKs at each end, and there is no need to view data in different reports. At present, the only solution known in the market is to view the business panorama in one report.

2. Open source, free, free customization

All codes are open source. What data to collect can be customized; batch running frequency can be customized (real-time statistics can be done during activities); display reports can be customized.

3. Private deployment, data self-control

Using traditional saas statistical products, all data are reported to statistical service vendors.

uni Statistics 2.0 is implemented based on uniCloud. Cloud functions and statistical data are all hosted in the developer's own service space (Alibaba Cloud or Tencent Cloud optional), and developers have complete control over their own statistical data.

4. Effective error analysis

Traditional app statistics platforms do not have js error statistics. Developers cannot know on which devices their js code will report errors.

The error information collected by uni is more comprehensive, including js front-end errors and app native layer crashes.

Because uni-app runs after compilation, the traditional web and MiniApp statistics platform, its js error cannot be traced back to the source code of uni-app before compilation, and the error report is incomprehensible.

uni statistics support sourcemap, which can intuitively understand which line of code is wrong. See details

real case scenario

[A real example of acquiring 4 million customers in one and a half months](https://mp.weixin.qq.com/s?__biz=MzU3NTU5NDc0NA==&mid=2247491214&idx=1&sn=7e334d079146d9e31cea407f45bd8624&chksm=fd219719ca561e0f9a85b30017618eaf9551b46cdd6ecdf856bc4e47aee4ca93767fcf23147f&mpshare=1&scene=1&srcid=0713VwAOIuRllzMB6syoQssb&key= 15a2b72b2464b4fe73325967f733ac332583d5db37f1812c63613c083a8f5921bca2ada2140d45e07657b062dc451f27cc48fe4fd298f6456f300895a90bd471480afdc2c8dc5a45254fb1dc48d3b79a&ascene=1&uin=MTkzNjMxMzU%3D&devicetype=Windows+10&version=62060833&lang=zh_CN&pass_ticket=xW6dPp%2F565g5S8hl1lz%2F8FLQBEzW6KUHyyqyHPdT2nk%3D)

# Summary

Let’s go back to uni statistics again. From the cost calculation above, we can know that for a daily activity of 10,000, the annual cost of each daily activity is about 6 cents to 7 cents. This cost is already relatively It's cheap.

At present, statistics on the market, such as Youmeng statistics, are also charged. When the daily activity reaches 10,000, the annual fee needs to be 3109 yuan/year (and this is only the cost of web statistics, and the cost of APP statistics is calculated separately). At the same time, it is not a source code version, the code is not open source, and cannot be customized at will.

And uni statistics is full-platform statistics, open-source code, free customization, which cannot be achieved by three-party statistical platforms.

Looking back at the calculation table just now, assuming that the DAU is 10,000, the annual cost of uni statistics is only 53.31*12=639.72 yuan/year.

Daily Active	Cloud Function (RMB/month)	Cloud Database (RMB/month)	Front-end Web Hosting (RMB/month)	Monthly Total (RMB/month)
100	0.14	0.42	可忽略不计	0.56
1000	1.10	4.24	可忽略不计	5.34
10000	10.93	42.38	可忽略不计	53.31
100000	100.97	423.85	Negligible	524.82

It should be the pursuit of smart engineers not to repeat manufacturing wheels, focus on business, quickly verify models, and achieve business growth.

This review is for your reference.

# How to lower your fees

1. Appropriately increase the front-end data reporting cycle Because the default is to report once every 10 seconds. Changing it to once every 60 seconds can reduce the number of visits by 6 times. For the specific adjustment method, please refer to the description of the front-end data reporting cycle in Data Reporting Logic above.
1. Enable redis cache This is also the most effective way to reduce the number of unistatistics 2.0 database queries. Because redis is billed according to capacity, no matter how many reads and writes are made, it will not cost more. For how to enable redis, please refer to Enable redis cache above.
1. Turn off real-time statistics Real-time statistics refer to cloud real-time computing statistical reports. But in reality, most people only care about yesterday's statistical report. Only when there is a promotion today will I pay attention to the data report in real time. It can be configured for daily statistics, and then adjusted for real-time statistics during activities. Modification method: Set the configuration item of real-time statistics to off, and then re-upload the configuration center (uni-config-center) to the associated service space. For the configuration item description, please refer to the above [common module configuration item description](#%E5%85%AC%E5%85%B1%E6%A8%A1%E5%9D%97%E9%85%8D%E7%BD %AE%E9%A1%B9%E8%AF%B4%E6%98%8E).

# common problem

# 1. 启动uni统计后，何时可以查看报表数据？

A: It is related to the configuration of timed tasks. By default, data such as Statistics Home and Today's Overview are visible after 1 hour, and the rest of the data are visible the next day. To learn more about the statistical time of various types of data, please refer to [Scheduled Task Configuration Instructions](#%E5%AE%9A%E6%97%B6%E4%BB%BB%E5%8A%A1%E9%85%8D% E7%BD%AE%E8%AF%B4%E6%98%8E).

# 2. 已经开启统计，定时任务配置也正常，但是后台还是看不到数据

answer:

Make sure to distinguish clearly, business app and admin are 2 projects. The business app is the collection end, and the admin is the report end
Make sure to use HBuilderX 3.4.14+. If it is a project created by cli, you need to upgrade cli to uni-app 3.4.14+
Make sure that Enable uni Statistics 2.0 is checked in the manifest of the business App project that needs statistics, and associate with the correct uniCloud service space
Make sure that the business app has been re-issued (you can issue it in HBuilder, you don't need to put it on the App Store or MiniApp Store). Data reporting only occurs after the project is released or the debug mode is enabled for running the project. In other cases, data will not be reported. Details
Make sure that the cloud functions in the uniCloud directory of the uni-admin project are uploaded to the same uniCloud service space as the App
In the cloud function log of uniCloud web console, you can see that the uni-stat-receiver cloud function has the correct request log
In the cloud function log of uniCloud web console, you can see that the uni-stat-cron cloud function has a regular execution log, and the log shows that the execution is successful ;If Not Found the cofnig file is displayed in the log, check the sixth question below;
If you want to see the statistics of the management side of uni-admin, you need to configure uni statistics 2.0 in the manifest of the uni-admin project and issue it again. Again, don't confuse business App and admin

# 3. 如何判断是否需要配置分钟级定时任务？

Answer: Generally, you do not need to configure it yourself, but if the time-out occurs in the Scheduled Task Cloud Function (uni-stat-cron), you should consider enabling the minute-level scheduled task.

# 4. 如何创建或授权`uni统计`运营管理员账号

答：参考uni-admin 给系统创建多个登录账户并设置不同的权限

# 5. 为什么总设备数比活跃设备数少？

Answer: The formula for calculating the total number of devices is: total number of devices = number of original devices + number of new devices, and the basis for judging whether a device is a new device is whether the time when the device last accessed an application has been stored in the client SDK If it is not stored, it is considered as a new device of the application (that is, when lvts=0, it is a new device, and when lvts>0, it is an old device). Therefore, if a device has accessed an application before, even if the data in the database is cleared at this time, since the time when the device accessed the application last time has been stored in the client SDK (that is, lvts > 0 at this time), the device It will no longer be considered as a new device of the application, so it will no longer be counted into the total number of devices of the application, but will only be counted into the number of active devices. At this time, the total number of devices may be less than the number of active devices. Condition.

# 6. uni-stat-cron运行日志显示 Not Found the cofnig file

Business App and admin are 2 projects. The business app is the collection end, and the admin is the report end; both projects contain uni-config-center; if these two projects are associated (multiplexed) with the same service space, it is easy to appear uni-config-center `The mutual coverage problem; at this time, it is recommended to maintain a single point, there are 2 solutions:

Focus on business apps: copy the uni-stat folder under uni-config-center in the uni-admin project to the uni-config-center directory under the business app project, and then Just re-upload the uni-config-center public module under the business App project.
Mainly based on uni-admin: Manually merge the configuration items under the uni-config-center under the business App project into the uni-config-center under the uni-admin project (Note: manual Merge the configuration items, do not overwrite the file as a whole), and then re-upload the uni-config-center public module under the uni-admin project.

# 7. 子账号需要赋予哪些表的权限才能正常查看uni统计

首先分三步骤

第一步：添加用户角色权限

去权限管理，添加一个权限id为：READ_UNI_STAT 的权限
去角色管理，添加一个角色id为：READ_UNI_STAT 的角色
去用户管理，赋予子账号角色 READ_UNI_STAT 角色

第二步：给相关的表设置read权限

涉及表如下：

所有 uni-stat- 开头的表
uni-pay-orders
opendb-app-list
opendb-app-versions

需要赋予上面的表的 read 权限

将上面的表的 read 权限全部设置为

"permission": {
	"read": "'READ_UNI_STAT' in auth.permission",
	"create": false,
	"update": false,
	"delete": false
},

第三步：前往菜单管理，对每一个uni统计的页面（包含子页面）设置下权限 READ_UNI_STAT（菜单只有拥有对应权限才会显示）

# 8. 某统计项突然没有数据怎么办？

答：首先登录uniCloud控制台，检查在出现问题的统计项配置的时间点(参考：定时任务配置说明)，uni-stat-cron云函数的运行日志，如果运行日志前面的状态标识是灰色的，代表云函数运行超时了，此时在云函数详情中将uni-stat-cron云函数的超时时间设置到最大值即可。如果运行日志的状态标识是绿色的，则需要检查日志内容是否有报错，然后根据报错内容做出调整。

# References

不掌握如下文档，很难对 uni统计2.0 吃透和做二次开发

uni-admin文档：详见
uni-id文档：详见
opendb文档：详见

products

HBuilderX uni-app uniCloud uniMPsdk 5+Runtime uni-id uni-search uniPay uniPush uni-verify sms uni-starter uni-admin uni-upgrade-center uni-im uni-ai uni-cms

Operating services

uni-AD uni-stat uni-portal Safety topic

Developer Services

forum dashboard

Docs

uni-app uniCloud native-support HBuilderX

Ecosystem

Marketplace OAuth

About： DCloud Jobs Terms of service Privacy policy acceptable use policy

# uni统计2.0

# demo experience

# feature of product

# uni statistics old version upgrade suggestion

# Environmental requirements

# Usage tutorial

# Front-end collection configuration

# MiniApp Domain Name Whitelist

# debug mode

# Background report configuration

# Create admin project

# Set the trigger period of the scheduled task cloud function

# Enable redis cache

# Associated service space

# 错误解析 uni-admin 1.9.4+

# Using the environment

# Generate sourceMap

# upload sourceMap

# Parse error

# Open source code interpretation

# Front-end collection SDK

# Data reporting logic

# Collection Type

# Cloud Statistics

# Statistical report display page

# Cloud function, general module description

# Common module configuration item description

# Scheduled task configuration instructions

# Error detection configuration instructions

# Log cleaning configuration instructions

# Extensions and customizations

# 前端 uni.report 基础用法

# Data reported using custom events in uni-admin

# Version upgrade

# 升级到1.9.0+ 版本

# 升级到2.4.0+ 含内容统计的版本

# uni statistics 2.0 cost evaluation

# Preface

# cloud function

# uni-stat-receiver

# uni-stat-cron

# cloud database

# Cloud database with redis cost comparison

# Front-end web hosting

# Capacity Fee

# traffic fee

# Summary of front-end web hosting

# 费用合并

# uni statistics bring you benefits

# Summary

# How to lower your fees

# common problem

# 1. 启动uni统计后，何时可以查看报表数据？

# 2. 已经开启统计，定时任务配置也正常，但是后台还是看不到数据

# 3. 如何判断是否需要配置分钟级定时任务？

# 4. 如何创建或授权uni统计运营管理员账号

# 5. 为什么总设备数比活跃设备数少？

# 6. uni-stat-cron运行日志显示 Not Found the cofnig file

# 7. 子账号需要赋予哪些表的权限才能正常查看uni统计

# 8. 某统计项突然没有数据怎么办？

# References

On This Page

# 4. 如何创建或授权`uni统计`运营管理员账号