Absolutely, you can implement a timeout mechanism that stops long-running database operations both at your Spring backend level and terminates the corresponding MongoDB server-side process. Since you're using MongoDB Atlas (4+) with Spring Data's MongoRepository and custom extensions, here's a step-by-step guide tailored to your stack:

The most reliable way to ensure MongoDB terminates slow operations is to use the maxTimeMS() option, which tells MongoDB to abort any operation that exceeds the specified time limit (in milliseconds). For your 1-minute requirement, that's 60000 ms.

For Custom Repository Methods #

If you're using custom extension methods (e.g., extending MongoRepository with a custom interface and implementation), you can leverage MongoTemplate to set this timeout explicitly:

import org.springframework.data.mongodb.core.MongoTemplate;
import org.springframework.data.mongodb.core.query.Query;
import static org.springframework.data.mongodb.core.query.Criteria.where;

@Service
public class YourCustomRepositoryImpl implements YourCustomRepository {

    private final MongoTemplate mongoTemplate;

    public YourCustomRepositoryImpl(MongoTemplate mongoTemplate) {
        this.mongoTemplate = mongoTemplate;
    }

    @Override
    public List<YourEntity> findSlowQuery() {
        Query query = new Query(where("someField").exists(true))
                .maxTime(60000); // 1 minute timeout

        return mongoTemplate.find(query, YourEntity.class);
    }
}

For Annotated @Query Methods #

If you're using @Query annotations in your repository interface, you can add the maxTimeMS hint directly in the query string:

import org.springframework.data.mongodb.repository.MongoRepository;
import org.springframework.data.mongodb.repository.Query;

public interface YourEntityRepository extends MongoRepository<YourEntity, String> {

    @Query(value = "{ 'someField' : ?0 }", hints = "{ 'maxTimeMS' : 60000 }")
    List<YourEntity> findBySomeFieldWithTimeout(String someValue);
}

To apply a default timeout to all MongoDB operations across your application, you can configure it via Spring Boot properties. This ensures even standard MongoRepository methods (like findAll(), findById()) respect the timeout:

Add these properties to your application.properties:

spring.data.mongodb.socket-timeout=60000
spring.data.mongodb.connect-timeout=10000 # Optional: Connection timeout
spring.data.mongodb.max-wait-time=60000

Or in application.yml:

spring:
  data:
    mongodb:
      socket-timeout: 60000
      connect-timeout: 10000
      max-wait-time: 60000

⚠️ Note: While these properties handle application-level timeouts, pairing them with the server-side maxTimeMS ensures MongoDB itself terminates the operation if it runs too long, preventing lingering server processes.

MongoDB Atlas adds an extra layer of protection to complement your code:

• Configure Slow Query Logs: Set a 1-minute threshold in the Atlas UI to log slow queries, helping you identify problematic operations proactively.
• Use Atlas Performance Advisor: It will suggest index additions or query optimizations to reduce the chance of long-running operations in the first place.
• Kill Stuck Operations Manually: If an operation slips through, use the Atlas Shell to run db.currentOp() to find long-running processes, then db.killOp(opId) to terminate them.

When an operation hits the timeout, MongoDB throws a MongoCommandException with error code 50 (MaxTimeMSExpired). Catch this in your service layer to handle it gracefully:

try {
    yourEntityRepository.findBySomeFieldWithTimeout("value");
} catch (MongoCommandException e) {
    if (e.getErrorCode() == 50) {
        // Handle timeout: log, return user-friendly message, etc.
        log.error("Query exceeded 1-minute timeout", e);
        throw new YourCustomTimeoutException("Operation took too long to complete");
    }
    throw e; // Re-throw other MongoDB exceptions
}

内容的提问来源于stack exchange，提问作者Nitzan S