Good Approach for Air Hotel Car booking without involving DB

sanman chavan

Hi Team,

In our application, we are rearchitecting a structure in the existing application flow

Current Flow

When the user tries to book for Air Hotel Car from the Search page, we do concurrent API requests in a particular area and store the data in SQL.

Once the user goes to respective list pages for booking,we show respective content for that request

For an example

1. API hit Search

Air =100 flights
Hotel=90 Hotels
Car =80 Cars

2. We store all the data in SQL.

3. The user reaches Air page shows 100 flights, hotel page show 90 hotels & car page show 80 cars

4. Once the user select air, hotel, and car we show the payment option

New Flow

When the user searches for Air Hotel Car from the Search page, we do concurrent API requests in a particular area but don’t want to store data in DB (need to store in memory, session local storage)

The main motive is to remove dependency to store data in DB (SQL /NoSQL). Bcz data is required only for a particular search. We are dumping the payload for logging purposes but we don’t need an extra hit to store in DB

Any idea, any suggestions?

5 Replies

Know the answer? Post it — somebody with the same question will find it here.

Varun Setia5y ago
If you want to scale the solution and persist the session data you must go for Redis Cache. You can set the eviction policy. You may cache the data based on location rather than based on users.
+2
Mahesh Chand5y ago
Here is a video on Microservices architecture how large companies build their software that service 100s of millions of customers. https://www.youtube.com/watch?v=eJJcbRu7bio
+3
Mahesh Chand5y ago
You can use an in-memory database that will sit between the app and the database. That way, you will have faster access of data. In-memory DB should have data mapped so your app doesnt have to do much querying from different tables. Less calculations on the fly, less complex queries, better results will be. On users machine, you use a local browser database to store data that is related to that user. Having 1000 user concurrency is not a major concern for applications these days. I'm assuming you're using a scalable public cloud to host your website.
+3
sanman chavan5y ago
But if 100 users are using our site at a time, will this temp approach will work ?

Just for add on,
- 100 users can access the website at a time
- We are going to use AWS Lambda for the individual API call
- Payload size on list page is approx 2-3 MB (combining Air Hotel car List page)
- The response is serving to WebAPI, Mobile, and Web application
+1
Sachin Singh5y ago
use TempData if you are using ASP.NET MVC or ASP.NET Core
First, hit the API, store data in
TempData["alldata"]=retrieveddata;
and then do a Peek Read on the next page as,
var data= TempData.Peek("allData") ;
loop through and
display all data.
+1