The ScalarDB Java API is mainly composed of the Administrative API and Transactional API. This guide briefly explains what kinds of APIs exist, how to use them, and related topics like how to handle exceptions.

Administrative API

This section explains how to execute administrative operations programmatically by using the Administrative API in ScalarDB.

Note

Another method for executing administrative operations is to use Schema Loader.

Get a DistributedTransactionAdmin instance

You first need to get a DistributedTransactionAdmin instance to execute administrative operations.

To get a DistributedTransactionAdmin instance, you can use TransactionFactory as follows:

TransactionFactory transactionFactory = TransactionFactory.create("<CONFIGURATION_FILE_PATH>");
DistributedTransactionAdmin admin = transactionFactory.getTransactionAdmin();

For details about configurations, see ScalarDB Configurations.

After you have executed all administrative operations, you should close the DistributedTransactionAdmin instance as follows:

admin.close();

Create a namespace

Before creating tables, namespaces must be created since a table belongs to one namespace.

You can create a namespace as follows:

// Create the namespace "ns". If the namespace already exists, an exception will be thrown.
admin.createNamespace("ns");

// Create the namespace only if it does not already exist.
boolean ifNotExists = true;
admin.createNamespace("ns", ifNotExists);

// Create the namespace with options.
Map<String, String> options = ...;
admin.createNamespace("ns", options);

Creation options

In the creation operations, like creating a namespace and creating a table, you can specify options that are maps of option names and values (Map<String, String>). By using the options, you can set storage adapter–specific configurations.

Select your database to see the options available:

Name Description Default
replication-strategy Cassandra replication strategy. Must be SimpleStrategy or NetworkTopologyStrategy. SimpleStrategy
compaction-strategy Cassandra compaction strategy, Must be LCS, STCS or TWCS. STCS
replication-factor Cassandra replication factor. 1
Name Description Default
ru Base resource unit. 400
no-scaling Disable auto-scaling for Cosmos DB for NoSQL. false
Name Description Default
no-scaling Disable auto-scaling for DynamoDB. false
no-backup Disable continuous backup for DynamoDB. false
ru Base resource unit. 10

No options are available for JDBC databases.

Create a table

When creating a table, you should define the table metadata and then create the table.

To define the table metadata, you can use TableMetadata. The following shows how to define the columns, partition key, clustering key including clustering orders, and secondary indexes of a table:

// Define the table metadata.
TableMetadata tableMetadata =
    TableMetadata.newBuilder()
        .addColumn("c1", DataType.INT)
        .addColumn("c2", DataType.TEXT)
        .addColumn("c3", DataType.BIGINT)
        .addColumn("c4", DataType.FLOAT)
        .addColumn("c5", DataType.DOUBLE)
        .addPartitionKey("c1")
        .addClusteringKey("c2", Scan.Ordering.Order.DESC)
        .addClusteringKey("c3", Scan.Ordering.Order.ASC)
        .addSecondaryIndex("c4")
        .build();

For details about the data model of ScalarDB, see Data Model.

Then, create a table as follows:

// Create the table "ns.tbl". If the table already exists, an exception will be thrown.
admin.createTable("ns", "tbl", tableMetadata);

// Create the table only if it does not already exist.
boolean ifNotExists = true;
admin.createTable("ns", "tbl", tableMetadata, ifNotExists);

// Create the table with options.
Map<String, String> options = ...;
admin.createTable("ns", "tbl", tableMetadata, options);

Create a secondary index

You can create a secondary index as follows:

// Create a secondary index on column "c5" for table "ns.tbl". If a secondary index already exists, an exception will be thrown.
admin.createIndex("ns", "tbl", "c5");

// Create the secondary index only if it does not already exist.
boolean ifNotExists = true;
admin.createIndex("ns", "tbl", "c5", ifNotExists);

// Create the secondary index with options.
Map<String, String> options = ...;
admin.createIndex("ns", "tbl", "c5", options);

Add a new column to a table

You can add a new, non-partition key column to a table as follows:

// Add a new column "c6" with the INT data type to the table "ns.tbl".
admin.addNewColumnToTable("ns", "tbl", "c6", DataType.INT)

Attention

You should carefully consider adding a new column to a table because the execution time may vary greatly depending on the underlying storage. Please plan accordingly and consider the following, especially if the database runs in production:

  • For Cosmos DB for NoSQL and DynamoDB: Adding a column is almost instantaneous as the table schema is not modified. Only the table metadata stored in a separate table is updated.
  • For Cassandra: Adding a column will only update the schema metadata and will not modify the existing schema records. The cluster topology is the main factor for the execution time. Changes to the schema metadata are shared to each cluster node via a gossip protocol. Because of this, the larger the cluster, the longer it will take for all nodes to be updated.
  • For relational databases (MySQL, Oracle, etc.): Adding a column shouldn’t take a long time to execute.

Truncate a table

You can truncate a table as follows:

// Truncate the table "ns.tbl".
admin.truncateTable("ns", "tbl");

Drop a secondary index

You can drop a secondary index as follows:

// Drop the secondary index on column "c5" from table "ns.tbl". If the secondary index does not exist, an exception will be thrown.
admin.dropIndex("ns", "tbl", "c5");

// Drop the secondary index only if it exists.
boolean ifExists = true;
admin.dropIndex("ns", "tbl", "c5", ifExists);

Drop a table

You can drop a table as follows:

// Drop the table "ns.tbl". If the table does not exist, an exception will be thrown.
admin.dropTable("ns", "tbl");

// Drop the table only if it exists.
boolean ifExists = true;
admin.dropTable("ns", "tbl", ifExists);

Drop a namespace

You can drop a namespace as follows:

// Drop the namespace "ns". If the namespace does not exist, an exception will be thrown.
admin.dropNamespace("ns");

// Drop the namespace only if it exists.
boolean ifExists = true;
admin.dropNamespace("ns", ifExists);

Get the tables of a namespace

You can get the tables of a namespace as follows:

// Get the tables of the namespace "ns".
Set<String> tables = admin.getNamespaceTableNames("ns");

Get table metadata

You can get table metadata as follows:

// Get the table metadata for "ns.tbl".
TableMetadata tableMetadata = admin.getTableMetadata("ns", "tbl");

Repair a table

You can repair the table metadata of an existing table as follows:

// Repair the table "ns.tbl" with options.
TableMetadata tableMetadata =
    TableMetadata.newBuilder()
        ...
        .build();
Map<String, String> options = ...;
admin.repairTable("ns", "tbl", tableMetadata, options);

Specify operations for the Coordinator table

The Coordinator table is used by the Transactional API to track the statuses of transactions.

When using a transaction manager, you must create the Coordinator table to execute transactions. In addition to creating the table, you can truncate and drop the Coordinator table.

Create the Coordinator table

You can create the Coordinator table as follows:

// Create the Coordinator table.
admin.createCoordinatorTables();

// Create the Coordinator table only if one does not already exist.
boolean ifNotExist = true;
admin.createCoordinatorTables(ifNotExist);

// Create the Coordinator table with options.
Map<String, String> options = ...;
admin.createCoordinatorTables(options);

Truncate the Coordinator table

You can truncate the Coordinator table as follows:

// Truncate the Coordinator table.
admin.truncateCoordinatorTables();

Drop the Coordinator table

You can drop the Coordinator table as follows:

// Drop the Coordinator table.
admin.dropCoordinatorTables();

// Drop the Coordinator table if one exist.
boolean ifExist = true;
admin.dropCoordinatorTables(ifExist);

Transactional API

This section explains how to execute transactional operations by using the Transactional API in ScalarDB.

Get a DistributedTransactionManager instance

You first need to get a DistributedTransactionManager instance to execute transactional operations.

To get a DistributedTransactionManager instance, you can use TransactionFactory as follows:

TransactionFactory transactionFactory = TransactionFactory.create("<CONFIGURATION_FILE_PATH>");
DistributedTransactionManager transactionManager = transactionFactory.getTransactionManager();

After you have executed all transactional operations, you should close the DistributedTransactionManager instance as follows:

transactionManager.close();

Begin or start a transaction

Before executing transactional CRUD operations, you need to begin or start a transaction.

You can begin a transaction as follows:

// Begin a transaction.
DistributedTransaction transaction = transactionManager.begin();

Or, you can start a transaction as follows:

// Start a transaction.
DistributedTransaction transaction = transactionManager.start();

Alternatively, you can use the begin method for a transaction by specifying a transaction ID as follows:

// Begin a transaction with specifying a transaction ID.
DistributedTransaction transaction = transactionManager.begin("<TRANSACTION_ID>");

Or, you can use the start method for a transaction by specifying a transaction ID as follows:

// Start a transaction with specifying a transaction ID.
DistributedTransaction transaction = transactionManager.start("<TRANSACTION_ID>");

Note

Specifying a transaction ID is useful when you want to link external systems to ScalarDB. Otherwise, you should use the begin() method or the start() method.

When you specify a transaction ID, make sure you specify a unique ID (for example, UUID v4) throughout the system since ScalarDB depends on the uniqueness of transaction IDs for correctness.

Join a transaction

Joining a transaction is particularly useful in a stateful application where a transaction spans multiple client requests. In such a scenario, the application can start a transaction during the first client request. Then, in subsequent client requests, the application can join the ongoing transaction by using the join() method.

You can join an ongoing transaction that has already begun by specifying the transaction ID as follows:

// Join a transaction.
DistributedTransaction transaction = transactionManager.join("<TRANSACTION_ID>");

Note

To get the transaction ID with getId(), you can specify the following:

tx.getId();

Resume a transaction

Resuming a transaction is particularly useful in a stateful application where a transaction spans multiple client requests. In such a scenario, the application can start a transaction during the first client request. Then, in subsequent client requests, the application can resume the ongoing transaction by using the resume() method.

You can resume an ongoing transaction that you have already begun by specifying a transaction ID as follows:

// Resume a transaction.
DistributedTransaction transaction = transactionManager.resume("<TRANSACTION_ID>");

Note

To get the transaction ID with getId(), you can specify the following:

tx.getId();

Implement CRUD operations

The following sections describe key construction and CRUD operations.

Note

Although all the builders of the CRUD operations can specify consistency by using the consistency() methods, those methods are ignored. Instead, the LINEARIZABLE consistency level is always used in transactions.

Key construction

Most CRUD operations need to specify Key objects (partition-key, clustering-key, etc.). So, before moving on to CRUD operations, the following explains how to construct a Key object.

For a single column key, you can use Key.of<TYPE_NAME>() methods to construct the key as follows:

// For a key that consists of a single column of INT.
Key key1 = Key.ofInt("col1", 1);

// For a key that consists of a single column of BIGINT.
Key key2 = Key.ofBigInt("col1", 100L);

// For a key that consists of a single column of DOUBLE.
Key key3 = Key.ofDouble("col1", 1.3d);

// For a key that consists of a single column of TEXT.
Key key4 = Key.ofText("col1", "value");

For a key that consists of two to five columns, you can use the Key.of() method to construct the key as follows. Similar to ImmutableMap.of() in Guava, you need to specify column names and values in turns:

// For a key that consists of two to five columns.
Key key1 = Key.of("col1", 1, "col2", 100L);
Key key2 = Key.of("col1", 1, "col2", 100L, "col3", 1.3d);
Key key3 = Key.of("col1", 1, "col2", 100L, "col3", 1.3d, "col4", "value");
Key key4 = Key.of("col1", 1, "col2", 100L, "col3", 1.3d, "col4", "value", "col5", false);

For a key that consists of more than five columns, we can use the builder to construct the key as follows:

// For a key that consists of more than five columns.
Key key = Key.newBuilder()
    .addInt("col1", 1)
    .addBigInt("col2", 100L)
    .addDouble("col3", 1.3d)
    .addText("col4", "value")
    .addBoolean("col5", false)
    .addInt("col6", 100)
    .build();

Get operation

Get is an operation to retrieve a single record specified by a primary key.

You need to create a Get object first, and then you can execute the object by using the transaction.get() method as follows:

// Create a `Get` operation.
Key partitionKey = Key.ofInt("c1", 10);
Key clusteringKey = Key.of("c2", "aaa", "c3", 100L);

Get get =
    Get.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .projections("c1", "c2", "c3", "c4")
        .build();

// Execute the `Get` operation.
Optional<Result> result = transaction.get(get);

You can also specify projections to choose which columns are returned.

Handle Result objects

The Get operation and Scan operation return Result objects. The following shows how to handle Result objects.

You can get a column value of a result by using get<TYPE_NAME>("<COLUMN_NAME>") methods as follows:

// Get the BOOLEAN value of a column.
boolean booleanValue = result.getBoolean("<COLUMN_NAME>");

// Get the INT value of a column.
int intValue = result.getInt("<COLUMN_NAME>");

// Get the BIGINT value of a column.
long bigIntValue = result.getBigInt("<COLUMN_NAME>");

// Get the FLOAT value of a column.
float floatValue = result.getFloat("<COLUMN_NAME>");

// Get the DOUBLE value of a column.
double doubleValue = result.getDouble("<COLUMN_NAME>");

// Get the TEXT value of a column.
String textValue = result.getText("<COLUMN_NAME>");

// Get the BLOB value of a column as a `ByteBuffer`.
ByteBuffer blobValue = result.getBlob("<COLUMN_NAME>");

// Get the BLOB value of a column as a `byte` array.
byte[] blobValueAsBytes = result.getBlobAsBytes("<COLUMN_NAME>");

And if you need to check if a value of a column is null, you can use the isNull("<COLUMN_NAME>") method.

// Check if a value of a column is null.
boolean isNull = result.isNull("<COLUMN_NAME>");

For more details, see the Result page in the Javadoc of the version of ScalarDB that you’re using.

Execute Get by using a secondary index

You can execute a Get operation by using a secondary index.

Instead of specifying a partition key, you can specify an index key (indexed column) to use a secondary index as follows:

// Create a `Get` operation by using a secondary index.
Key indexKey = Key.ofFloat("c4", 1.23F);

Get get =
    Get.newBuilder()
        .namespace("ns")
        .table("tbl")
        .indexKey(indexKey)
        .projections("c1", "c2", "c3", "c4")
        .build();

// Execute the `Get` operation.
Optional<Result> result = transaction.get(get);

Note

If the result has more than one record, transaction.get() will throw an exception. If you want to handle multiple results, see Execute Scan by using a secondary index.

Scan operation

Scan is an operation to retrieve multiple records within a partition. You can specify clustering-key boundaries and orderings for clustering-key columns in Scan operations.

You need to create a Scan object first, and then you can execute the object by using the transaction.scan() method as follows:

// Create a `Scan` operation.
Key partitionKey = Key.ofInt("c1", 10);
Key startClusteringKey = Key.of("c2", "aaa", "c3", 100L);
Key endClusteringKey = Key.of("c2", "aaa", "c3", 300L);

Scan scan =
    Scan.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .start(startClusteringKey, true)    // Include startClusteringKey
        .end(endClusteringKey, false)       // Exclude endClusteringKey
        .projections("c1", "c2", "c3", "c4")
        .orderings(Scan.Ordering.desc("c2"), Scan.Ordering.asc("c3"))
        .limit(10)
        .build();

// Execute the `Scan` operation.
List<Result> results = transaction.scan(scan);

You can omit the clustering-key boundaries or specify either a start boundary or an end boundary. If you don’t specify orderings, you will get results ordered by the clustering order that you defined when creating the table.

In addition, you can specify projections to choose which columns are returned and use limit to specify the number of records to return in Scan operations.

Execute Scan by using a secondary index

You can execute a Scan operation by using a secondary index.

Instead of specifying a partition key, you can specify an index key (indexed column) to use a secondary index as follows:

// Create a `Scan` operation by using a secondary index.
Key indexKey = Key.ofFloat("c4", 1.23F);

Scan scan =
    Scan.newBuilder()
        .namespace("ns")
        .table("tbl")
        .indexKey(indexKey)
        .projections("c1", "c2", "c3", "c4")
        .limit(10)
        .build();

// Execute the `Scan` operation.
List<Result> results = transaction.scan(scan);

Note

You can’t specify clustering-key boundaries and orderings in Scan by using a secondary index.

Execute Scan without specifying a partition key to retrieve all the records of a table

You can execute a Scan operation without specifying a partition key.

Instead of calling the partitionKey() method in the builder, you can call the all() method to scan a table without specifying a partition key as follows:

// Create a `Scan` operation without specifying a partition key.
Scan scan =
    Scan.newBuilder()
        .namespace("ns")
        .table("tbl")
        .all()
        .projections("c1", "c2", "c3", "c4")
        .limit(10)
        .build();

// Execute the `Scan` operation.
List<Result> results = transaction.scan(scan);

Note

You can’t specify clustering-key boundaries and orderings in Scan without specifying a partition key.

Put operation

Put is an operation to put a record specified by a primary key. The operation behaves as an upsert operation for a record, in which the operation updates the record if the record exists or inserts the record if the record does not exist.

Note

When you update an existing record, you need to read the record by using Get or Scan before using a Put operation.

You need to create a Put object first, and then you can execute the object by using the transaction.put() method as follows:

// Create a `Put` operation.
Key partitionKey = Key.ofInt("c1", 10);
Key clusteringKey = Key.of("c2", "aaa", "c3", 100L);

Put put =
    Put.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .floatValue("c4", 1.23F)
        .doubleValue("c5", 4.56)
        .build();

// Execute the `Put` operation.
transaction.put(put);

You can also put a record with null values as follows:

Put put =
    Put.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .floatValue("c4", null)
        .doubleValue("c5", null)
        .build();

Delete operation

Delete is an operation to delete a record specified by a primary key.

Note

When you delete a record, you need to read the record by using Get or Scan before using a Delete operation.

You need to create a Delete object first, and then you can execute the object by using the transaction.delete() method as follows:

// Create a `Delete` operation.
Key partitionKey = Key.ofInt("c1", 10);
Key clusteringKey = Key.of("c2", "aaa", "c3", 100L);

Delete delete =
    Delete.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .build();

// Execute the `Delete` operation.
transaction.delete(delete);

Put and Delete with a condition

You can write arbitrary conditions (for example, a bank account balance must be equal to or more than zero) that you require a transaction to meet before being committed by implementing logic that checks the conditions in the transaction. Alternatively, you can write simple conditions in a mutation operation, such as Put and Delete.

When a Put or Delete operation includes a condition, the operation is executed only if the specified condition is met. If the condition is not met when the operation is executed, an exception called UnsatisfiedConditionException will be thrown.

Conditions for Put

You can specify a condition in a Put operation as follows:

// Build a condition.
MutationCondition condition =
    ConditionBuilder.putIf(ConditionBuilder.column("c4").isEqualToFloat(0.0F))
        .and(ConditionBuilder.column("c5").isEqualToDouble(0.0))
        .build();

Put put =
    Put.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .floatValue("c4", 1.23F)
        .doubleValue("c5", 4.56)
        .condition(condition) // condition
        .build();

// Execute the `Put` operation.
transaction.put(put);

In addition to using the putIf condition, you can specify the putIfExists and putIfNotExists conditions as follows:

// Build a `putIfExists` condition.
MutationCondition putIfExistsCondition = ConditionBuilder.putIfExists();

// Build a `putIfNotExists` condition.
MutationCondition putIfNotExistsCondition = ConditionBuilder.putIfNotExists();
Conditions for Delete

You can specify a condition in a Delete operation as follows:

// Build a condition.
MutationCondition condition =
    ConditionBuilder.deleteIf(ConditionBuilder.column("c4").isEqualToFloat(0.0F))
        .and(ConditionBuilder.column("c5").isEqualToDouble(0.0))
        .build();

Delete delete =
    Delete.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKey)
        .condition(condition) // condition
        .build();

// Execute the `Delete` operation.
transaction.delete(delete);

In addition to using the deleteIf condition, you can specify the deleteIfExists condition as follows:

// Build a `deleteIfExists` condition.
MutationCondition deleteIfExistsCondition = ConditionBuilder.deleteIfExists();

Mutate operation

Mutate is an operation to execute multiple mutations (Put and Delete operations).

You need to create mutation objects first, and then you can execute the objects by using the transaction.mutate() method as follows:

// Create `Put` and `Delete` operations.
Key partitionKey = Key.ofInt("c1", 10);

Key clusteringKeyForPut = Key.of("c2", "aaa", "c3", 100L);

Put put =
    Put.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKeyForPut)
        .floatValue("c4", 1.23F)
        .doubleValue("c5", 4.56)
        .build();

Key clusteringKeyForDelete = Key.of("c2", "bbb", "c3", 200L);

Delete delete =
    Delete.newBuilder()
        .namespace("ns")
        .table("tbl")
        .partitionKey(partitionKey)
        .clusteringKey(clusteringKeyForDelete)
        .build();

// Execute the operations.
transaction.mutate(Arrays.asList(put, delete));

Default namespace for CRUD operations

A default namespace for all CRUD operations can be set by using a property in the ScalarDB configuration.

scalar.db.default_namespace_name=<NAMESPACE_NAME>

Any operation that does not specify a namespace will use the default namespace set in the configuration.

// This operation will target the default namespace.
Scan scanUsingDefaultNamespace =
    Scan.newBuilder()
        .table("tbl")
        .all()
        .build();
// This operation will target the "ns" namespace.
Scan scanUsingSpecifiedNamespace =
    Scan.newBuilder()
        .namespace("ns")
        .table("tbl")
        .all()
        .build();

Commit a transaction

After executing CRUD operations, you need to commit a transaction to finish it.

You can commit a transaction as follows:

// Commit a transaction.
transaction.commit();

Roll back or abort a transaction

If an error occurs when executing a transaction, you can roll back or abort the transaction.

You can roll back a transaction as follows:

// Roll back a transaction.
transaction.rollback();

Or, you can abort a transaction as follows:

// Abort a transaction.
transaction.abort();

For details about how to handle exceptions in ScalarDB, see How to handle exceptions.

How to handle exceptions

When executing a transaction, you will also need to handle exceptions properly.

Attention

If you don’t handle exceptions properly, you may face anomalies or data inconsistency.

The following sample code shows how to handle exceptions:

public class Sample {
  public static void main(String[] args) throws Exception {
    TransactionFactory factory = TransactionFactory.create("<CONFIGURATION_FILE_PATH>");
    DistributedTransactionManager transactionManager = factory.getTransactionManager();

    int retryCount = 0;
    TransactionException lastException = null;

    while (true) {
      if (retryCount++ > 0) {
        // Retry the transaction three times maximum.
        if (retryCount >= 3) {
          // Throw the last exception if the number of retries exceeds the maximum.
          throw lastException;
        }

        // Sleep 100 milliseconds before retrying the transaction.
        TimeUnit.MILLISECONDS.sleep(100);
      }

      DistributedTransaction transaction = null;
      try {
        // Begin a transaction.
        transaction = transactionManager.begin();

        // Execute CRUD operations in the transaction.
        Optional<Result> result = transaction.get(...);
        List<Result> results = transaction.scan(...);
        transaction.put(...);
        transaction.delete(...);

        // Commit the transaction.
        transaction.commit();
      } catch (UnsatisfiedConditionException e) {
        // You need to handle `UnsatisfiedConditionException` only if a mutation operation specifies a condition.
        // This exception indicates the condition for the mutation operation is not met.

        try {
          transaction.rollback();
        } catch (RollbackException ex) {
          // Rolling back the transaction failed. Since the transaction should eventually recover, 
          // you don't need to do anything further. You can simply log the occurrence here.
        }

        // You can handle the exception here, according to your application requirements.

        return;
      } catch (UnknownTransactionStatusException e) {
        // If you catch `UnknownTransactionStatusException` when committing the transaction, 
        // it indicates that the status of the transaction, whether it was successful or not, is unknown.
        // In such a case, you need to check if the transaction is committed successfully or not and 
        // retry the transaction if it failed. How to identify a transaction status is delegated to users.
        return;
      } catch (TransactionException e) {
        // For other exceptions, you can try retrying the transaction.

        // For `CrudConflictException`, `CommitConflictException`, and `TransactionNotFoundException`,
        // you can basically retry the transaction. However, for the other exceptions, the transaction
        // will still fail if the cause of the exception is non-transient. In such a case, you will 
        // exhaust the number of retries and throw the last exception.

        if (transaction != null) {
          try {
            transaction.rollback();
          } catch (RollbackException ex) {
            // Rolling back the transaction failed. The transaction should eventually recover,
            // so you don't need to do anything further. You can simply log the occurrence here.
          }
        }

        lastException = e;
      }
    }
  }
}

TransactionException and TransactionNotFoundException

The begin() API could throw TransactionException or TransactionNotFoundException:

  • If you catch TransactionException, this exception indicates that the transaction has failed to begin due to transient or non-transient faults. You can try retrying the transaction, but you may not be able to begin the transaction due to non-transient faults.
  • If you catch TransactionNotFoundException, this exception indicates that the transaction has failed to begin due to transient faults. In this case, you can retry the transaction.

The join() API could also throw TransactionNotFoundException. You can handle this exception in the same way that you handle the exceptions for the begin() API.

CrudException and CrudConflictException

The APIs for CRUD operations (get(), scan(), put(), delete(), and mutate()) could throw CrudException or CrudConflictException:

  • If you catch CrudException, this exception indicates that the transaction CRUD operation has failed due to transient or non-transient faults. You can try retrying the transaction from the beginning, but the transaction may still fail if the cause is non-transient.
  • If you catch CrudConflictException, this exception indicates that the transaction CRUD operation has failed due to transient faults (for example, a conflict error). In this case, you can retry the transaction from the beginning.

UnsatisfiedConditionException

The APIs for mutation operations (put(), delete(), and mutate()) could also throw UnsatisfiedConditionException.

If you catch UnsatisfiedConditionException, this exception indicates that the condition for the mutation operation is not met. You can handle this exception according to your application requirements.

CommitException, CommitConflictException, and UnknownTransactionStatusException

The commit() API could throw CommitException, CommitConflictException, or UnknownTransactionStatusException:

  • If you catch CommitException, this exception indicates that committing the transaction fails due to transient or non-transient faults. You can try retrying the transaction from the beginning, but the transaction may still fail if the cause is non-transient.
  • If you catch CommitConflictException, this exception indicates that committing the transaction has failed due to transient faults (for example, a conflict error). In this case, you can retry the transaction from the beginning.
  • If you catch UnknownTransactionStatusException, this exception indicates that the status of the transaction, whether it was successful or not, is unknown. In this case, you need to check if the transaction is committed successfully and retry the transaction if it has failed.

How to identify a transaction status is delegated to users. You may want to create a transaction status table and update it transactionally with other application data so that you can get the status of a transaction from the status table.

Notes about some exceptions

Although not illustrated in the sample code, the resume() API could also throw TransactionNotFoundException. This exception indicates that the transaction associated with the specified ID was not found and/or the transaction might have expired. In either case, you can retry the transaction from the beginning since the cause of this exception is basically transient.

In the sample code, for UnknownTransactionStatusException, the transaction is not retried because the application must check if the transaction was successful to avoid potential duplicate operations. For other exceptions, the transaction is retried because the cause of the exception is transient or non-transient. If the cause of the exception is transient, the transaction may succeed if you retry it. However, if the cause of the exception is non-transient, the transaction will still fail even if you retry it. In such a case, you will exhaust the number of retries.

Note

In the sample code, the transaction is retried three times maximum and sleeps for 100 milliseconds before it is retried. But you can choose a retry policy, such as exponential backoff, according to your application requirements.

Investigating Consensus Commit transaction manager errors

To investigate errors when using the Consensus Commit transaction manager, you can enable a configuration that will return table metadata augmented with transaction metadata columns, which can be helpful when investigating transaction-related issues. This configuration, which is only available when troubleshooting the Consensus Commit transaction manager, enables you to see transaction metadata column details for a given table by using the DistributedTransactionAdmin.getTableMetadata() method.

By adding the following configuration, Get and Scan operations results will contain transaction metadata:

# By default, this configuration is set to `false`.
scalar.db.consensus_commit.include_metadata.enabled=true