Spring Data cassandra throws: Cannot obtain where clauses for entity

Question

iam trying to use Spring data cassandra to communicate with cassandra/scyllaDB via REST api. I have entity

@Table
public class Transaction {

    @PrimaryKeyColumn(name = "id", ordinal = 0, type = PrimaryKeyType.PARTITIONED)
    private String id;

    @PrimaryKeyColumn(name = "timestamp", ordinal = 1, type = PrimaryKeyType.CLUSTERED)
    private Instant timestamp;

    private String currency;
}

with repository

@Repository
public interface TransactionRepository extends CassandraRepository<Transaction, String> {
}

and service which is called in REST controller

@Service
public class TransactionServiceImpl implements TransactionService{

    @Autowired
    private TransactionRepository transactionRepository;
    
    private Transaction getTransaction(String transactionId) {
        return transactionRepository.findById(transactionId)
                .orElseThrow(() -> new NotFoundException("Transaction with provided " + transactionId + " does not exist."));
    }
}

When i call required REST endpoint, where i provide correct transactionId, Exception is thrown.

Caused by: org.springframework.dao.InvalidDataAccessApiUsageException: Cannot obtain where clauses for 
entity [com.example.dao.entity.Transaction] using [id123456]

I have made some investigation, but composite key provided in entity class should be valid. What ma i doing wrong?

Model of the table:

CREATE TABLE etl.transaction
(
    id text,
    timestamp timestamp,
    currency ascii,
    PRIMARY KEY (id, timestamp)
)
WITH CLUSTERING ORDER BY (timestamp ASC) AND
default_time_to_live = 157680000; // 5 years in seconds

Answer 1

I never get this error message. What will you get error message if you have setter/getter methods or use Lombok project and add @Data annotation next to @Table? (If you use IntelliJ, you have to install Lombok plugin!)

In my answer, write comment if there is any result.

@Data
@Table
public class Transaction {

    @PrimaryKeyColumn(name = "id", ordinal = 0, type = PrimaryKeyType.PARTITIONED)
    private String id;

    @PrimaryKeyColumn(name = "timestamp", ordinal = 1, type = PrimaryKeyType.CLUSTERED)
    private Instant timestamp;

    private String currency;
}

---

UPDATE

TL;DR: Just add @Id above the attribute if you have one only @PrimaryKeyColumn. If you use more @PrimaryKeyColumn, you have to use:

• @PrimaryKey and @PrimaryKeyClass (and you can use findById)

public interface TestRepository extends CassandraRepository<Test, KeyClass> {
}

@Data
@Table
public class Test {
    
    @PrimaryKey
    private KeyClass id;    

    @Data
    @PrimaryKeyClass
    public static class KeyClass { // You either give it a different name or move it in another file. Nevermind.
        @PrimaryKeyColumn(name = "id", type = PrimaryKeyType.PARTITIONED)
        private Integer id;
    
        @PrimaryKeyColumn(name = "data")
        private Integer data;
    }
}

• or next interface:

public interface TestRepository extends CassandraRepository<Test, Integer> {
    Optional<Test> findByIdAndData(Integer id, Integer data);
}

@Data
@Table
public class Test {
    
    @PrimaryKeyColumn(name = "id", type = PrimaryKeyType.PARTITIONED)
    private Integer id;
    
    @PrimaryKeyColumn(name = "data")
    private Integer data;
}

---

With solo @PrimaryKeyColumn

Test data:

-- auto-generated definition
CREATE TABLE test
(
    id   int PRIMARY KEY,
    data int
)
    WITH CACHING = {'keys': 'ALL', 'rows_per_partition': 'NONE'}
     AND COMPACTION = {'max_threshold': '32', 'min_threshold': '4', 'class': 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy'}
     AND COMPRESSION = {'class': 'org.apache.cassandra.io.compress.LZ4Compressor', 'chunk_length_in_kb': '64'}
     AND DCLOCAL_READ_REPAIR_CHANCE = 0.1;

public interface TestRepository extends CassandraRepository<Test, Integer> {
}

@Data
@Table
public class Test {
    
    /*
     * You can only use "findById" or "findallById" with @Id annotation.
     * @PrimaryKey contains the @Id itself, but ofc you have to configuration @PrimaryKeyClass.
     */
    @PrimaryKeyColumn(name = "id", type = PrimaryKeyType.PARTITIONED)
    @Id // <-- add
    private Integer id;
    
    @Column("data")
    private Integer data;
}

Test:

@SpringBootTest
class ApplicationTests {
    
    @Autowired
    private TestRepository testRepository;
    
    @Test
    void contextLoads() {
        var x = testRepository.findById(1).orElse(null);
        
        if(x == null) {
            fail();
        }
    
        assertEquals((int) x.getData(), 5);
    }
}

Result:

---

With @PrimaryKey Table:

public interface TestRepository extends CassandraRepository<Test, KeyClass> {
}

@Data
@Table
public class Test {
    
    @PrimaryKey
    private KeyClass id;    

    @Data
    @PrimaryKeyClass
    public static class KeyClass { // You either give it a different name or move it in another file. Nevermind.
        @PrimaryKeyColumn(name = "id", type = PrimaryKeyType.PARTITIONED)
        private Integer id;
    
        @PrimaryKeyColumn(name = "data")
        private Integer data;
    }
}

---

With JPA API

The second genetic type (what defined type of id) doesn’t matter what, on this solution.

---

If you want just query by PARTITION. CLUSTERED is exist!

Solution 1:

The second genetic type (what defined type of id) doesn’t matter what, on this solution.

Solution 2:

Rename id to something else and configure JPA that way.

The second genetic type (what defined type of id) doesn’t matter what, on this solution.

Solution with @PrimaryKeyClass

public interface TestRepository extends CassandraRepository<Test, KeyClass> {
    // findby[ID-1]_[ID-2](...)
    // ID-1 is mean: Test's "id" attribute
    // "_" is mean chain/access instance
    // ID-2 is mean: KeyClass's "id" attribute
    Optional<Test> findById_Id(Integer id);

    // Look this both:
    Optional<Test> findById_IdAndId_Data(Integer id, Integer data);
}