Data Access Gateway Chrome Extension
This is the source code for the Data Access Gateway library. The goal of this Chrome Extension is to receive statistics about how the data is fetched and saved by the library.
The extension goal is to provide insight about how the library is manipulating the data. It will indicate how many question are still pending (on-going request), and will give information about which source of data is trying read and write (memory, IndexDb or Http). It also give statistic to which of the three source the data is used as well as a complete list of URL (id) that is being accessed by the system.
The extension is literally the extension of the Data Access Gateway library. It gives insight about what is being used, when, what was the size of the payload and how long it tooks to be distributed to your application. It shows if the memory cache, or the persisted cache (IndexDb) is used and how many requests and bytes are you saving.
- Quick glance information
- about how many request are running in the background
- Percentage of fetch coming from the cache compared to server
- Distribution of memory cache versus browser indexDb cache
- Number of read, write and use of memory cache, index db cache and HTTP request
- Distribution of the used bytes coming from the three sources
- Percentile of the usage from indexdb vs http
- Historical view of all fetch, save in cache, read, on-going request
- Performance for each operation, bytes transfered and write in cache
- Drill in for further analysis
- Compare a single operation against previous one
- Allows to compare against similar query (different query string or URL fragment)
- Possible to filter by data source (memory, indexDb, http) and data action (fetch, save, use, etc.)
- Give the number of time the payload changed and the average time between changes
- Possility to save session and load to have bigger historical analysis
- Possible to switch in "demo mode" which blur all url for easy sharing without leaking confidential endpoints.
Diving In Console
The header of the console can be clicked. It will open a panel. The panel contains option to filter the grid. Changing the value will trim the number of result in the grid allowing to see large payload in size or long request.
Every line of the console can be clicked. It opens the detailled view of a single request. The view allows to see in detail every ID (URL) that was executed to compare it against the one selected. This will filter by the line "action". For example, if the selected time is an action of "use" it will generate a graph of all the "action" for the same url. However, this is configurable. It is possible to use a slider to move a threshold value which will be used to compare ID (URL) with the Levenshtein algorithm. The configuration allows to bring in the analysis ID (URL) that change because of query string value that are different. For example, in the image below, the threshold is set to 1. It allows 1 character permutation. The line is for "http://url2" but the analysis includes "http://url1" as well. This can be adjusted to work for REST Api URL format and allow to see if specific parameter influence the performance or the payload size.
If you want to contribute to this Chrome Extension here are some details.
The technologies used it TypeScript and React. TypeScript is used for the agent and the Chrome Extension which is a panel that is inserted in the Chrome's developer tool. React is used inside the panel where the information about the Data Access Gateway.
At the moment, there is some fake data allowing to see the UI and play around when modifying or adding fearures. A caveat is how to execute the NPM scripts. Some copy use
cpx but some not. For some reasons it does not work properly when moving code between the two projects. I have not invested any time on it and focused that it work on a Mac OS. I'll fix that later.
Here are the steps for developing the panel
The panel is what is injected inside Chrome Developer tool. This is what should be done when changing features on the extension most of the time.
- At the root:
- Move to the
- To run in browser:
npm run start
- To build:
npm run build. This will build and move the files in the root (extension)
dist. You can then use Chrome to load the root
chrome://extensionsand use the extension on real website using the Data Access Gateway library. However, you must follow the next section step to generate the manifest.json file in the
distfolder as well.
Here are the steps for building the Chrome's extension
The Chrome Extension is what communicate between the library and the Chrome's extension. This should rarely be changed.
- At the root:
npm run build
- Go in Chrome to the extension,
chrome://extensions/, select "Load unpacked" and select the
When developing the extension, make sure to always
- Build the code after each modification you want to see in Chrome with
npm run build. Chrome will not read the code from
run startbut only from the
- You will have to load only once the extension from the Chrome's Extension
chrome://extensionsto point to the root
- Close and open the Chrome's Dev Tools every time after
npm run buildis executed.
Publish into Chrome Web Store
.env file and add the following environment variables:
CLIENT_ID=XXXXX.apps.googleusercontent.com CLIENT_SECRET=XXXXX REFRESH_TOKEN=XXXXX
The values are coming from the readme.md from chrome-webstore-upload-cli project. Once the configuration is setup:
- Go in
npm run build
- Go back at the root and run
npm run publish
- Unit tests
- UI could be improved. A balance between simplicity and density of information should be always respected.
- Find a way to have the creat-react-app to be unminimized for debugging purpose