If a query is taking too long to execute, what checks do you perform and how do you troubleshoot it?

If a query is taking too long, I start by examining the execution plan to identify bottlenecks such as full table scans, hash joins, or data skew. I check whether the relevant columns are indexed or partitioned appropriately. I also look at the size of intermediate results and whether the query is pulling more data than necessary. If the issue is a join, I consider the join type and whether a broadcast join would be more efficient. I then test the query with sample data to isolate the problematic part. Once I identify the root cause, I apply fixes such as adding indexes, rewriting the query to use more selective predicates, or partitioning the data. Finally, I monitor the query performance after the changes to ensure the issue is resolved.