1

To count the unread message, I created a count field and it works like this :

  1. set count_unread = 0
  2. If new_message : count_unread++
    • If user join chat room : set count_unread = 0
    • If user already in chat room, I have a listener to listen database change, then show the message in screen and set count_unread = 0

But, as you know, Firebase charge on every read/write command. So, if this user already in the chat room, every new message will create 2 writes ( set count_unread +1 and set count_unread=0 ). Is there any way to optimize it ?

boygiandi
  • 630
  • 10
  • 20
  • See https://stackoverflow.com/questions/52126744/how-to-structure-nosql-messages-to-get-unreads-by-1-query/52129025#52129025, and other from here: https://stackoverflow.com/search?q=%5Bfirebase%5D+unread+message+count for some ideas. But the trade-off you're describing is pretty much the same for all of them. – Frank van Puffelen Apr 05 '20 at 03:21
  • Not sure the logic is clear and without understanding the code, any answers would be a guess. For me, if I'm in a chat room and messages are being added, the count_unread would be incrementing for ME only and if there are multiple users, the number of uread messages would be different for each so they would each have their own counter. Then, as I read the message, MY count_unread would decrement. So I dont see where you would have two writes. – Jay Apr 06 '20 at 18:07

1 Answers1

0

Yes, you can optimize your Firebase Realtime Database usage by reducing the number of read and write operations that you perform. Here are a few suggestions:

Use Firebase Cloud Functions to update the unread message count for each user when new messages are added to a chat room. This way, you can avoid having to perform multiple read and write operations for each user in real-time. Instead, you can use Cloud Functions to calculate the unread message count based on the existing messages in the chat room, and update the count for each user only when new messages are added or deleted. This approach can help reduce the number of read and write operations you need to perform on the database, and improve the overall performance of your app.

Use Firebase's built-in transaction support to update the unread message count atomically. When you need to increment the unread message count, you can use a transaction to ensure that the count is incremented atomically, without any race conditions or conflicts with other clients. This approach can help reduce the number of write operations you need to perform on the database, and ensure that the unread message count is always accurate.

Consider using Firebase's Firestore database instead of the Realtime Database. Firestore has built-in support for server-side increment operations, which can help reduce the number of read and write operations you need to perform on the database. You can use Firestore's FieldValue.increment() method to atomically increment the unread message count without having to read the existing value first. This can be a more efficient approach if you have a large number of users and chat rooms, and need to update the unread message count frequently.

I hope these suggestions help you optimize your Firebase usage and reduce the number of read and write operations you need to perform on the database.

farhan meo
  • 37
  • 1
  • 1
  • 5