0.1 million of SQL Server tables become bottlenecks

A netizen posted a message in one of my blog posts. I took it out and sorted it out.

The problem is as follows:

Our database has encountered a bottleneck in the process of increasing users. The database has a high frequency of concurrent operations. There are several tables with a large amount of data and tens of thousands of records, these tables are frequently operated, so deadlocks may occur when the number of online users reaches about two thousand.
The database has become a bottleneck! We also use ms SQL Server2000. I have read about database clusters, but I have said on the Internet that SQL Server Clusters do not support Server Load balancer and only support failover.
However, we have no practical experience.
I hope to give some suggestions or write articles on the knowledge of SQL Server Clusters and database load balancing. Thank you!
By the way, we are engaged in card and board games, so the database is used together by the game hall and website.

My personal opinion:

Is the number of online users about two thousand? Buddy, how do you count the number of online users? Concurrent users and online users are not necessarily linked. You won't ping it every 10 seconds. Make sure you are online.

If the data is only 0.1 million, you should check how much data is frequently used? The data should be buffered.
Do not access the database every time you read the data. That is very slow. I have seen several game programs that read all the basic data to the memory (personnel, equipment, etc.). These data are used too frequently and can be directly used when I read the memory, which means the efficiency is absolutely High. On the other hand, the data will be refreshed to the database on a regular basis. I remember it was only once an hour.

However, I still think
1. You have a problem with programming.
2. Your query statement is written incorrectly.
3. Your index settings are incorrect.
4. Your database structure is faulty and can be optimized.
5. Your machine is too bad.

Take a good look at which causes performance problems. Optimize statements, algorithms, and data structures.