BigQuery Cross Join run like forever

I want to ask regarding Cross Join on BigQuery. I perform Cross Join where the 2 tables is actually huge (let's say 2Million and 1Million) the query is running like forever. Is there any way or alternative when it comes to cross join ?

If there are M rows from the first and N rows from the second, the result is M * N rows.
In your case, 1M *1M would be huge and the query would run forever.
Please go through the below link which defines cross-join to avoid anti-patterns:
https://cloud.google.com/bigquery/docs/best-practices-performance-patterns

Or, you can write specific problem which you are working on and people here might be able to help you.

Cross join means that server needs to map table A each row with table B each row to get all possible data combinations. In your case this results in 2 Million x 1 Million = 2×10¹² rows for server to generate ! If you really need something like cross join - sample your big tables with random() functions to get a reasonable small test sets and then perform cross join on them later, to reduce output set.

You must use cross join in cases where you need to associate each element from a table with each element from the other table. In this case, if you use a cross join you will get a 2 trillion records answer.
In this page you can find all the kinds of query in BigQuery and its usage.

Try to describe your problem better so I can help you to find a feasible solution since cross join is not a good possibility for this case.