IBM Cloud Docs
Using Java

Using Java

The IBM Cloud® Object Storage SDK for Java provides features to make the most of IBM Cloud Object Storage.

The IBM Cloud Object Storage SDK for Java is comprehensive, with many features and capabilities that exceed the scope and space of this guide. For detailed class and method documentation see the Javadoc. Source code can be found in the GitHub repository.

Getting the SDK

The easiest way to use the IBM Cloud Object Storage Java SDK is to use Maven to manage dependencies. If you aren't familiar with Maven, you can get up and running by using the Maven in 5-Minutes guide.

Maven uses a file that is called pom.xml to specify the libraries (and their versions) needed for a Java project. Here is an example pom.xml file for using the IBM Cloud Object Storage Java SDK to connect to Object Storage.

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>com.cos</groupId>
    <artifactId>docs</artifactId>
    <version>1.0.0-SNAPSHOT</version>
    <dependencies>
        <dependency>
            <groupId>com.ibm.cos</groupId>
            <artifactId>ibm-cos-java-sdk</artifactId>
            <version>2.8.0</version>
        </dependency>
    </dependencies>
</project>

Migrating from 1.x.x

The 2.0 release of the SDK introduces a namespacing change that allows an application to make use of the original AWS library to connect to AWS resources within the same application or environment. To migrate from 1.x to 2.0 some changes are necessary:

  1. Update by using Maven by changing all ibm-cos-java-sdk dependency version tags to 2.0.0 in the pom.xml. Verify that there are no SDK module dependencies in the pom.xml with a version earlier than 2.0.0.
  2. Update any import declarations from amazonaws to ibm.cloud.objectstorage.

Creating a client and sourcing credentials

In the following example, a client cos is created and configured by providing credential information (API key and service instance ID). These values can also be automatically sourced from a credentials file or from environment variables.

After generating a Service Credential, the resulting JSON document can be saved to ~/.bluemix/cos_credentials. The SDK will automatically source credentials from this file unless other credentials are explicitly set during client creation. If the cos_credentials file contains HMAC keys the client authenticates with a signature, otherwise the client uses the provided API key to authenticate by using a bearer token.

If migrating from AWS S3, you can also source credentials data from ~/.aws/credentials in the format:

[default]
aws_access_key_id = {API_KEY}
aws_secret_access_key = {SERVICE_INSTANCE_ID}

If both ~/.bluemix/cos_credentials and ~/.aws/credentials exist, cos_credentials takes preference.

For more details on client construction, see the Javadoc.

Code Examples

In your code, you must remove the angled brackets or any other excess characters that are provided here as illustration.

Let's start with an complete example class that will run through some basic functionality, then explore the classes individually. This CosExample class will list objects in an existing bucket, create a new bucket, and then list all buckets in the service instance.

Gather required information

  • bucketName and newBucketName are unique and DNS-safe strings. Because bucket names are unique across the entire system, these values need to be changed if this example is run multiple times. Note that names are reserved for 10 - 15 minutes after deletion.
  • apiKey is the value found in the Service Credential as apikey.
  • serviceInstanceId is the value found in the Service Credential as resource_instance_id.
  • endpointUrl is a service endpoint URL, inclusive of the https:// protocol. This is not the endpoints value found in the Service Credential. For more information about endpoints, see Endpoints and storage locations.
  • storageClass is a valid provisioning code that corresponds to the endpoint value. This is then used as the S3 API LocationConstraint variable.
  • location should be set to the location portion of the storageClass. For us-south-standard, this would be us-south. This variable is used only for the calculation of HMAC signatures, but is required for any client, including this example that uses an IAM API key.
    package com.cos;

    import java.time.LocalDateTime;
    import java.util.List;

    import com.ibm.cloud.objectstorage.ClientConfiguration;
    import com.ibm.cloud.objectstorage.auth.AWSCredentials;
    import com.ibm.cloud.objectstorage.auth.AWSStaticCredentialsProvider;
    import com.ibm.cloud.objectstorage.client.builder.AwsClientBuilder.EndpointConfiguration;
    import com.ibm.cloud.objectstorage.services.s3.AmazonS3;
    import com.ibm.cloud.objectstorage.services.s3.AmazonS3ClientBuilder;
    import com.ibm.cloud.objectstorage.services.s3.model.Bucket;
    import com.ibm.cloud.objectstorage.services.s3.model.ListObjectsRequest;
    import com.ibm.cloud.objectstorage.services.s3.model.ObjectListing;
    import com.ibm.cloud.objectstorage.services.s3.model.S3ObjectSummary;
    import com.ibm.cloud.objectstorage.oauth.BasicIBMOAuthCredentials;

    public class CosExample
    {
        public static void main(String[] args)
        {
            String bucketName = "<BUCKET_NAME>";  // eg my-unique-bucket-name
            String newBucketName = "<NEW_BUCKET_NAME>"; // eg my-other-unique-bucket-name
            String apiKey = "<API_KEY>"; // eg "W00YiRnLW4k3fTjMB-oiB-2ySfTrFBIQQWanc--P3byk"
            String serviceInstanceId = "<SERVICE_INSTANCE_ID"; // eg "crn:v1:bluemix:public:cloud-object-storage:global:a/3bf0d9003abfb5d29761c3e97696b71c:d6f04d83-6c4f-4a62-a165-696756d63903::"
            String endpointUrl = "https://s3.us-south.cloud-object-storage.appdomain.cloud"; // this could be any service endpoint

            String storageClass = "us-south-standard";
            String location = "us"; // not an endpoint, but used in a custom function below to obtain the correct URL

            System.out.println("Current time: " + LocalDateTime.now());
            AmazonS3 cosClient = createClient(apiKey, serviceInstanceId, endpointUrl, location);
            
            listObjects(cosClient, bucketName);
            createBucket(cosClient, newBucketName, storageClass);
            listBuckets(cosClient);
        }

        public static AmazonS3 createClient(String apiKey, String serviceInstanceId, String endpointUrl, String location)
        {
            AWSCredentials credentials = new BasicIBMOAuthCredentials(apiKey, serviceInstanceId);
            ClientConfiguration clientConfig = new ClientConfiguration()
                    .withRequestTimeout(5000)
                    .withTcpKeepAlive(true);

            return AmazonS3ClientBuilder
                    .standard()
                    .withCredentials(new AWSStaticCredentialsProvider(credentials))
                    .withEndpointConfiguration(new EndpointConfiguration(endpointUrl, location))
                    .withPathStyleAccessEnabled(true)
                    .withClientConfiguration(clientConfig)
                    .build();
        }

        public static void listObjects(AmazonS3 cosClient, String bucketName)
        {
            System.out.println("Listing objects in bucket " + bucketName);
            ObjectListing objectListing = cosClient.listObjects(new ListObjectsRequest().withBucketName(bucketName));
            for (S3ObjectSummary objectSummary : objectListing.getObjectSummaries()) {
                System.out.println(" - " + objectSummary.getKey() + "  " + "(size = " + objectSummary.getSize() + ")");
            }
            System.out.println();
        }

        public static void createBucket(AmazonS3 cosClient, String bucketName, String storageClass)
        {
            cosClient.createBucket(bucketName, storageClass);
        }

        public static void listBuckets(AmazonS3 cosClient)
        {
            System.out.println("Listing buckets");
            final List<Bucket> bucketList = cosClient.listBuckets();
            for (final Bucket bucket : bucketList) {
                System.out.println(bucket.getName());
            }
            System.out.println();
        }
    }

Initializing configuration

private static String COS_ENDPOINT = "<endpoint>"; // eg "https://s3.us.cloud-object-storage.appdomain.cloud"
private static String COS_API_KEY_ID = "<api-key>"; // eg "0viPHOY7LbLNa9eLftrtHPpTjoGv6hbLD1QalRXikliJ"
private static String COS_AUTH_ENDPOINT = "https://iam.cloud.ibm.com/identity/token";
private static String COS_SERVICE_CRN = "<resource-instance-id>"; // "crn:v1:bluemix:public:cloud-object-storage:global:a/<CREDENTIAL_ID_AS_GENERATED>:<SERVICE_ID_AS_GENERATED>::"
private static String COS_BUCKET_LOCATION = "<location>"; // eg "us"

public static void main(String[] args)
{
    SDKGlobalConfiguration.IAM_ENDPOINT = COS_AUTH_ENDPOINT;

    try {
        _cos = createClient(COS_API_KEY_ID, COS_SERVICE_CRN, COS_ENDPOINT, COS_BUCKET_LOCATION);
    } catch (SdkClientException sdke) {
        System.out.printf("SDK Error: %s\n", sdke.getMessage());
    } catch (Exception e) {
        System.out.printf("Error: %s\n", e.getMessage());
    }
}

public static AmazonS3 createClient(String api_key, String service_instance_id, String endpoint_url, String location)
{
    AWSCredentials credentials = new BasicIBMOAuthCredentials(api_key, service_instance_id);
    ClientConfiguration clientConfig = new ClientConfiguration().withRequestTimeout(5000);
    clientConfig.setUseTcpKeepAlive(true);

    AmazonS3 cos = AmazonS3ClientBuilder.standard().withCredentials(new AWSStaticCredentialsProvider(credentials))
            .withEndpointConfiguration(new EndpointConfiguration(endpoint_url, location)).withPathStyleAccessEnabled(true)
            .withClientConfiguration(clientConfig).build();

    return cos;
}

Key Values

  • <endpoint> - public endpoint for your cloud Object Storage (available from the IBM Cloud Dashboard). For more information about endpoints, see Endpoints and storage locations.
  • <api-key> - api key generated when creating the service credentials (write access is required for creation and deletion examples)
  • <resource-instance-id> - resource ID for your cloud Object Storage (available through IBM Cloud CLI or IBM Cloud Dashboard)
  • <location> - default location for your cloud Object Storage (must match the region that is used for <endpoint>)

SDK References

Determining Endpoint

The methods below can be used to determine the service endpoint based on the bucket location, endpoint type (public or private), and specific region (optional). For more information about endpoints, see Endpoints and storage locations.

/**
* Returns a service endpoint based on the 
* storage class location (i.e. us-standard, us-south-standard), 
* endpoint type (public or private)
*/
public static String getEndpoint(String location, String endPointType) {
    return getEndpoint(location, "", endPointType);
}

/**
* Returns a service endpoint based on the 
* storage class location (i.e. us-standard, us-south-standard), 
* specific region if desired (i.e. sanjose, amsterdam) - only use if you want a specific regional endpoint,
* endpoint type (public or private)
*/
public static String getEndpoint(String location, String region, String endpointType) {
    HashMap locationMap = new HashMap<String, String>();
    locationMap.put("us", "s3-api.us-geo");
    locationMap.put("us-dallas", "s3-api.dal-us-geo");
    locationMap.put("us-sanjose", "s3-api.sjc-us-geo");
    locationMap.put("us-washington", "s3-api.wdc-us-geo");
    locationMap.put("us-south", "s3.us-south");
    locationMap.put("us-east", "s3.us-east");
    locationMap.put("eu", "s3.eu-geo");
    locationMap.put("eu-amsterdam", "s3.ams-eu-geo");
    locationMap.put("eu-frankfurt", "s3.fra-eu-geo");
    locationMap.put("eu-milan", "s3.mil-eu-geo");
    locationMap.put("eu-gb", "s3.eu-gb");
    locationMap.put("eu-germany", "s3.eu-de");
    locationMap.put("ap", "s3.ap-geo");
    locationMap.put("ap-tokyo", "s3.tok-ap-geo");
    locationMap.put("che01", "s3.che01");
    locationMap.put("mel01", "s3.mel01");
    locationMap.put("tor01", "s3.tor01");

    String key = location.substring(0, location.lastIndexOf("-")) + (region != null && !region.isEmpty() ? "-" + region : "");
    String endpoint = locationMap.getOrDefault(key, null).toString();

    if (endpoint != null) {
        if (endpointType.toLowerCase() == "private")
            endpoint += ".objectstorage.service.networklayer.com";
        else
            endpoint += ".objectstorage.s3.us-south.cloud-object-storage.appdomain.cloud.net";
    }

    return endpoint;
}

Creating a new bucket

public static void createBucket(String bucketName) {
    System.out.printf("Creating new bucket: %s\n", bucketName);
    _cos.createBucket(bucketName);
    System.out.printf("Bucket: %s created!\n", bucketName);
}

Create a bucket with a different storage class

A list of valid provisioning codes for LocationConstraint can be referenced in the Storage Classes guide.

cos.createBucket("sample", "us-vault"); // the name of the bucket, and the storage class (LocationConstraint)

SDK References

Creating a new text file

public static void createTextFile(String bucketName, String itemName, String fileText) {
    System.out.printf("Creating new item: %s\n", itemName);
    
    byte[] arr = fileText.getBytes(StandardCharsets.UTF_8);
    InputStream newStream = new ByteArrayInputStream(arr);
    
    ObjectMetadata metadata = new ObjectMetadata();
    metadata.setContentLength(arr.length);
    
    PutObjectRequest req = new PutObjectRequest(bucketName, itemName, newStream, metadata);
    _cos.putObject(req);
    
    System.out.printf("Item: %s created!\n", itemName);
}

Note that when adding custom metadata to an object, it is necessary to create an ObjectMetadata object by using the SDK, and not to manually send a custom header containing x-amz-meta-{key}. The latter can cause issues when authenticating by using HMAC credentials.

Upload object from a file

This example assumes that the bucket sample exists.

cos.putObject(
    "sample", // the name of the destination bucket
    "myfile", // the object key
    new File("/home/user/test.txt") // the file name and path of the object to be uploaded
);

Upload object by using a stream

This example assumes that the bucket sample exists.

String obj = "An example"; // the object to be stored
ByteArrayOutputStream theBytes = new ByteArrayOutputStream(); // create a new output stream to store the object data
ObjectOutputStream serializer = new ObjectOutputStream(theBytes); // set the object data to be serialized
serializer.writeObject(obj); // serialize the object data
serializer.flush();
serializer.close();
InputStream stream = new ByteArrayInputStream(theBytes.toByteArray()); // convert the serialized data to a new input stream to store
ObjectMetadata metadata = new ObjectMetadata(); // define the metadata
metadata.setContentType("application/x-java-serialized-object"); // set the metadata
metadata.setContentLength(theBytes.size()); // set metadata for the length of the data stream
cos.putObject(
    "sample", // the name of the bucket to which the object is being written
    "serialized-object", // the name of the object being written
    stream, // the name of the data stream writing the object
    metadata // the metadata for the object being written
);

Alternatively, you can use a CipherInputStream to more easily encrypt the data stream without needing to overload the existing InputStream object.

public CipherInputStream encryptStream(InputStream inputStream) { 
       // Generate key
       KeyGenerator kgen = KeyGenerator.getInstance("AES");
       kgen.init(128);
       SecretKey aesKey = kgen.generateKey();
       // Encrypt cipher
       Cipher cipher = Cipher.getInstance("AES/CBC/PKCS5Padding");
       cipher.init(Cipher.ENCRYPT_MODE, aesKey);
       CipherInputStream cis = new CipherInputStream(inputStream, cipher);
       return cis;
} 

Download object to a file

This example assumes that the bucket sample exists.

GetObjectRequest request = new // create a new request to get an object
GetObjectRequest( // request the new object by identifying
    "sample", // the name of the bucket
    "myFile" // the name of the object
);

s3Client.getObject( // write the contents of the object
    request, // using the request that was just created
    new File("retrieved.txt") // to write to a new file
);

Download object by using a stream

This example assumes that the bucket sample exists.

S3Object returned = cos.getObject( // request the object by identifying
    "sample", // the name of the bucket
    "serialized-object" // the name of the serialized object
);
S3ObjectInputStream s3Input = returned.getObjectContent(); // set the object stream

Copy objects

// copy an object within the same Bucket
cos.copyObject( // copy the Object, passing…
    "sample",  // the name of the Bucket in which the Object to be copied is stored,
    "myFile.txt",  // the name of the Object being copied from the source Bucket,
    "sample",  // the name of the Bucket in which the Object to be copied is stored,
    "myFile.txt.backup"    // and the new name of the copy of the Object to be copied
);
// copy an object between two Buckets
cos.copyObject( // copy the Object, passing…
    "sample", // the name of the Bucket from which the Object will be copied,
    "myFile.txt", // the name of the Object being copied from the source Bucket,
    "backup", // the name of the Bucket to which the Object will be copied,
    "myFile.txt" // and the name of the copied Object in the destination Bucket
);

SDK References

putObject Exception

The putObject method might throw the following exception even if the new object upload was successful:

Exception in thread "main" java.lang.NoClassDefFoundError: javax/xml/bind/JAXBException
	at com.ibm.cloud.objectstorage.services.s3.AmazonS3Client.putObject(AmazonS3Client.java:1597)
	at ibmcos.CoSExample.createTextFile(CoSExample.java:174)
	at ibmcos.CoSExample.main(CoSExample.java:65)
Caused by: java.lang.ClassNotFoundException: javax.xml.bind.JAXBException
	at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:582)
	at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:190)
	at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:499)
	... 3 more

Root Cause: The JAXB APIs are considered to be Java EE APIs, and are no longer contained on the default class path in Java SE 9.

Fix: Add the following entry to the pom.xml file in your project folder and repackage your project

<dependency>
    <groupId>javax.xml.bind</groupId>
    <artifactId>jaxb-api</artifactId>
    <version>2.3.0</version>
</dependency>

List available buckets

public static void getBuckets() {
    System.out.println("Retrieving list of buckets");

    final List<Bucket> bucketList = _cos.listBuckets();
    for (final Bucket bucket : bucketList) {
        System.out.printf("Bucket Name: %s\n", bucket.getName());
    }
}

SDK References

List items in a bucket (v2)

The AmazonS3 object contains an updated method to list the contents (listObjectsV2). This method allows you to limit the number of records that are returned and retrieve the records in batches. This might be useful for paging your results within an application and improve performance.

public static void getBucketContentsV2(String bucketName, int maxKeys) {
    System.out.printf("Retrieving bucket contents (V2) from: %s\n", bucketName);

    boolean moreResults = true;
    String nextToken = "";

    while (moreResults) {
        ListObjectsV2Request request = new ListObjectsV2Request()
            .withBucketName(bucketName)
            .withMaxKeys(maxKeys)
            .withContinuationToken(nextToken);

        ListObjectsV2Result result = _cos.listObjectsV2(request);
        for(S3ObjectSummary objectSummary : result.getObjectSummaries()) {
            System.out.printf("Item: %s (%s bytes)\n", objectSummary.getKey(), objectSummary.getSize());
        }
        
        if (result.isTruncated()) {
            nextToken = result.getNextContinuationToken();
            System.out.println("...More results in next batch!\n");
        }
        else {
            nextToken = "";
            moreResults = false;
        }
    }
    System.out.println("...No more results!");
}

SDK References

Get file contents of particular item

public static void getItem(String bucketName, String itemName) {
    System.out.printf("Retrieving item from bucket: %s, key: %s\n", bucketName, itemName);

    S3Object item = _cos.getObject(new GetObjectRequest(bucketName, itemName));

    try {
        final int bufferSize = 1024;
        final char[] buffer = new char[bufferSize];
        final StringBuilder out = new StringBuilder();
        InputStreamReader in = new InputStreamReader(item.getObjectContent());

        for (; ; ) {
            int rsz = in.read(buffer, 0, buffer.length);
            if (rsz < 0)
                break;
            out.append(buffer, 0, rsz);
        }

        System.out.println(out.toString());
    } catch (IOException ioe){
        System.out.printf("Error reading file %s: %s\n", name, ioe.getMessage());
    }
}

SDK References

Delete an item from a bucket

public static void deleteItem(String bucketName, String itemName) {
    System.out.printf("Deleting item: %s\n", itemName);
    _cos.deleteObject(bucketName, itemName);
    System.out.printf("Item: %s deleted!\n", itemName);
}

SDK References

Delete multiple items from a bucket

The delete request can contain a maximum of 1000 keys that you want to delete. While this is very useful in reducing the per-request performance hit, be mindful when deleting a large number of keys. Also take into account the sizes of the objects to ensure suitable performance.

public static void deleteItems(String bucketName) {
    DeleteObjectsRequest req = new DeleteObjectsRequest(bucketName);
    req.withKeys(
        "deletetest/testfile1.txt",
        "deletetest/testfile2.txt",
        "deletetest/testfile3.txt",
        "deletetest/testfile4.txt",
        "deletetest/testfile5.txt"
    );

    DeleteObjectsResult res = _cos.deleteObjects(req);
    
    System.out.printf("Deleted items for %s\n", bucketName);

    List<DeleteObjectsResult.DeletedObject> deletedItems = res.getDeletedObjects();
    for(DeleteObjectsResult.DeletedObject deletedItem : deletedItems) {
        System.out.printf("Deleted item: %s\n", deletedItem.getKey());
    }
}

SDK References

Delete a bucket

public static void deleteBucket(String bucketName) {
    System.out.printf("Deleting bucket: %s\n", bucketName);
    _cos.deleteBucket(bucketName);
    System.out.printf("Bucket: %s deleted!\n", bucketName);
}

SDK References

Check if an object is publicly readable

public static void getItemACL(String bucketName, String itemName) {
    System.out.printf("Retrieving ACL for %s from bucket: %s\n", itemName, bucketName);

    AccessControlList acl = _cos.getObjectAcl(bucketName, itemName);

    List<Grant> grants = acl.getGrantsAsList();

    for (Grant grant : grants) {
        System.out.printf("User: %s (%s)\n", grant.getGrantee().getIdentifier(), grant.getPermission().toString());
    }
}

SDK References

Execute a multi-part upload

public static void multiPartUpload(String bucketName, String itemName, String filePath) {
    File file = new File(filePath);
    if (!file.isFile()) {
        System.out.printf("The file '%s' does not exist or is not accessible.\n", filePath);
        return;
    }

    System.out.printf("Starting multi-part upload for %s to bucket: %s\n", itemName, bucketName);

    InitiateMultipartUploadResult mpResult = _cos.initiateMultipartUpload(new InitiateMultipartUploadRequest(bucketName, itemName));
    String uploadID = mpResult.getUploadId();

    //begin uploading the parts
    //min 5MB part size
    long partSize = 1024 * 1024 * 5;
    long fileSize = file.length();
    long partCount = ((long)Math.ceil(fileSize / partSize)) + 1;
    List<PartETag> dataPacks = new ArrayList<PartETag>();

    try {
        long position = 0;
        for (int partNum = 1; position < fileSize; partNum++) {
            partSize = Math.min(partSize, (fileSize - position));

            System.out.printf("Uploading to %s (part %s of %s)\n", name, partNum, partCount);  

            UploadPartRequest upRequest = new UploadPartRequest()
                    .withBucketName(bucketName)
                    .withKey(itemName)
                    .withUploadId(uploadID)
                    .withPartNumber(partNum)
                    .withFileOffset(position)
                    .withFile(file)
                    .withPartSize(partSize);

            UploadPartResult upResult = _cos.uploadPart(upRequest);
            dataPacks.add(upResult.getPartETag());

            position += partSize;
        } 

        //complete upload
        _cos.completeMultipartUpload(new CompleteMultipartUploadRequest(bucketName, itemName, uploadID, dataPacks));
        System.out.printf("Upload for %s Complete!\n", itemName);
    } catch (SdkClientException sdke) {
        System.out.printf("Multi-part upload aborted for %s\n", itemName);
        System.out.printf("Upload Error: %s\n", sdke.getMessage());
        _cos.abortMultipartUpload(new AbortMultipartUploadRequest(bucketName, itemName, uploadID));
    }
}

SDK References

Upload larger objects using a Transfer Manager

The TransferManager simplifies large file transfers by automatically incorporating multi-part uploads whenever necessary setting configuration parameters.

public static void largeObjectUpload(String bucketName, String itemName, String filePath) throws IOException, InterruptedException {
    File uploadFile = new File(filePath);

    if (!uploadFile.isFile()) {
        System.out.printf("The file '%s' does not exist or is not accessible.\n", filePath);
        return;
    }

    System.out.println("Starting large file upload with TransferManager");

    //set the part size to 5 MB    
    long partSize = 1024 * 1024 * 5;

    //set the threshold size to 5 MB
    long thresholdSize = 1024 * 1024 * 5;

    String endPoint = getEndpoint(COS_BUCKET_LOCATION, "public");
    AmazonS3 s3client = createClient(COS_API_KEY_ID, COS_SERVICE_CRN, endPoint, COS_BUCKET_LOCATION);

    TransferManager transferManager = TransferManagerBuilder.standard()
        .withS3Client(s3client)
        .withMinimumUploadPartSize(partSize)
        .withMultipartCopyThreshold(thresholdSize)
        .build();

    try {
        Upload lrgUpload = transferManager.upload(bucketName, itemName, uploadFile);

        lrgUpload.waitForCompletion();

        System.out.println("Large file upload complete!");
    }
    catch (SdkClientException e) {
        System.out.printf("Upload error: %s\n", e.getMessage());            
    }
    finally {
        transferManager.shutdownNow();
    }

SDK References

Using Key Protect

Key Protect can be added to a storage bucket to encrypt sensitive data at rest in the cloud.

Before You Begin

The following items are necessary in order to create a bucket with Key-Protect enabled:

Retrieving the Root Key CRN

  1. Retrieve the instance ID for your Key Protect service
  2. Use the Key Protect API to retrieve all your available keys
  3. Retrieve the CRN of the root key you will use to enabled Key Protect on the your bucket. The CRN will look similar to below:

crn:v1:bluemix:public:kms:us-south:a/3d624cd74a0dea86ed8efe3101341742:90b6a1db-0fe1-4fe9-b91e-962c327df531:key:0bg3e33e-a866-50f2-b715-5cba2bc93234

Creating a bucket with key-protect enabled

private static String COS_KP_ALGORITHM = "<algorithm>";
private static String COS_KP_ROOTKEY_CRN = "<root-key-crn>";

public static void createBucketKP(String bucketName) {
    System.out.printf("Creating new encrypted bucket: %s\n", bucketName);

    EncryptionType encType = new EncryptionType();
    encType.setKmsEncryptionAlgorithm(COS_KP_ALGORITHM);
    encType.setIBMSSEKMSCustomerRootKeyCrn(COS_KP_ROOTKEY_CRN);

    CreateBucketRequest req = new CreateBucketRequest(bucketName).withEncryptionType(encType);

    _cos.createBucket(req);

    System.out.printf("Bucket: %s created!", bucketName);
}

Key Values

  • <algorithm> - The encryption algorithm used for new objects added to the bucket (Default is AES256).
  • <root-key-crn> - CRN of the Root Key obtained from the Key Protect service.

SDK References

New Headers for Key Protect

The additional headers have been defined within Headers class:

public static final String IBM_SSE_KP_ENCRYPTION_ALGORITHM = "ibm-sse-kp-encryption-algorithm";
public static final String IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN = "ibm-sse-kp-customer-root-key-crn";

The same section of the create bucket implementation which already adds IAM service instance headers will add the 2 new encryption headers:

//Add IBM Service Instance Id & Encryption to headers
if ((null != this.awsCredentialsProvider ) && (this.awsCredentialsProvider.getCredentials() instanceof IBMOAuthCredentials)) {
    IBMOAuthCredentials oAuthCreds = (IBMOAuthCredentials)this.awsCredentialsProvider.getCredentials();
    if (oAuthCreds.getServiceInstanceId() != null) {
        request.addHeader(Headers.IBM_SERVICE_INSTANCE_ID, oAuthCreds.getServiceInstanceId());
        request.addHeader(Headers.IBM_SSE_KP_ENCRYPTION_ALGORITHM, createBucketRequest.getEncryptionType().getKpEncryptionAlgorithm());
        request.addHeader(Headers.IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN, createBucketRequest.getEncryptionType().getIBMSSEKPCustomerRootKeyCrn());
    }
}

The ObjectListing and HeadBucketResult objects have been updated to include boolean IBMSSEKPEnabled & String IBMSSEKPCustomerRootKeyCrn variables with getter & setter methods. These will store the values of the new headers.

GET bucket

public ObjectListing listObjects(String bucketName)
public ObjectListing listObjects(String bucketName, String prefix)
public ObjectListing listObjects(ListObjectsRequest listObjectsRequest)

The ObjectListing class will require 2 additional methods:

ObjectListing listing = s3client.listObjects(listObjectsRequest)
String KPEnabled = listing.getIBMSSEKPEnabled();
String crkId = listing.getIBMSSEKPCrkId();

The additional headers have been defined within the Headers class:

Headers.java
public static final string IBM_SSE_KP_ENABLED = "ibm-sse-kp-enabled";
public static final String IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN = "ibm-sse-kp-customer-root-key-crn";

The S3XmlResponseHandler which is responsible for unmarshalling all xml responses. A check has been added that the result is an instance of ObjectListing and the retrieved headers will be added to the ObjectListing object:

if (result instanceof ObjectListing) {
    if (!StringUtils.isNullOrEmpty(responseHeaders.get(Headers.IBM_SSE_KP_ENABLED)){
            ((ObjectListing) result).setIBMSSEKPEnabled(Boolean.parseBoolean(responseHeaders.get(Headers.IBM_SSE_KP_ENABLED)));
        }
    if (!StringUtils.isNullOrEmpty(responseHeaders.get(Headers.IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN))) {
            ((ObjectListing) result).setIBMSSEKPCrk(responseHeaders.get(Headers.IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN));
        }
}

HEAD bucket

The additional headers have been defined within Headers class:

Headers.java
public static final String IBM_SSE_KP_ENABLED = "ibm-sse-kp-enabled";
public static final String IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN = "ibm-sse-kp-customer-root-key-crn";

These variables are populated in the HeadBucketResponseHandler.

HeadBucketResultHandler
result.setIBMSSEKPEnabled(response.getHeaders().get(Headers.IBM_SSE_KP_ENABLED));
result.setIBMSSEKPCrk(response.getHeaders().get(Headers. IBM_SSE_KP_CUSTOMER_ROOT_KEY_CRN));

Head Bucket Example
HeadBucketResult result = s3client.headBucket(headBucketRequest)
boolean KPEnabled = result.getIBMSSEKPEnabled();
String crn = result.getIBMSSEKPCUSTOMERROOTKEYCRN();

Using Aspera High-Speed Transfer

By installing the Aspera high-speed transfer library you can utilize high-speed file transfers within your application. The Aspera library is closed-source, and thus an optional dependency for the COS SDK (which uses an Apache license).

Each Aspera high-speed transfer session spawns an individual ascp process that runs on the client machine to perform the transfer. Ensure that your computing environment can allow this process to run.

You will need instances of the S3 Client and IAM Token Manager classes to initialize the AsperaTransferManager. The s3Client is required to get FASP connection information for the COS target bucket. The tokenManager is required to allow the Aspera high-speed transfer SDK to authenticate with the COS target bucket.

Initializing the AsperaTransferManager

Before initializing the AsperaTransferManager, make sure you've got working s3Client and tokenManager objects.

It is advised to use TokenManager tokenManager = new DefaultTokenManager(new DelegateTokenProvider(apiKey)); and avoid .withTokenManager(tokenManager) when building AsperaTransferManager with AsperaTransferManagerBuilder.

There isn't a lot of benefit to using a single session of Aspera high-speed transfer unless you expect to see significant noise or packet loss in the network. So we need to tell the AsperaTransferManager to use multiple sessions using the AsperaConfig class. This will split the transfer into a number of parallel sessions that send chunks of data whose size is defined by the threshold value.

The typical configuration for using multi-session should be:

  • 2500 MBps target rate
  • 100 MB threshold (this is the recommended value for most applications)
AsperaTransferManagerConfig transferConfig = new AsperaTransferManagerConfig()
    .withMultiSession(true);

AsperaConfig asperaConfig = new AsperaConfig()
    .withTargetRateMbps(2500L)
    .withMultiSessionThresholdMb(100);
    
TokenManager tokenManager = new DefaultTokenManager(new DelegateTokenProvider(API_KEY));

AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(API_KEY, s3Client)
    .withAsperaTransferManagerConfig(transferConfig)
    .withAsperaConfig(asperaConfig)
    .build();

In the above example, the sdk will spawn enough sessions to attempt to reach the target rate of 2500 MBps.

Alternatively, session management can be explicitly configured in the sdk. This is useful in cases where more precise control over network utilization is desired.

The typical configuration for using explicit multi-session should be:

  • 2 or 10 sessions
  • 100 MB threshold (this is the recommended value for most applications)
AsperaConfig asperaConfig = new AsperaConfig()
    .withMultiSession(2)
    .withMultiSessionThresholdMb(100);

TokenManager tokenManager = new DefaultTokenManager(new DelegateTokenProvider(API_KEY));

AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(API_KEY, s3Client)
    .withAsperaConfig(asperaConfig)
    .build();

For best performance in most scenarios, always make use of multiple sessions to minimize any processing associated with instantiating an Aspera high-speed transfer. If your network capacity is at least 1 Gbps you should use 10 sessions. Lower bandwidth networks should use two sessions.

Key Values

  • API_KEY - An API key for a user or service ID with Writer or Manager roles

You will need to provide an IAM API Key for constructing an AsperaTransferManager. HMAC Credentials are NOT currently supported. For more information on IAM, click here.

File Upload

String filePath = "<absolute-path-to-source-data>";
String bucketName = "<bucket-name>";
String itemName = "<item-name>";

// Load file
File inputFile = new File(filePath);

// Create AsperaTransferManager for FASP upload
AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(API_KEY, s3Client).build();

// Upload test file and report progress
Future<AsperaTransaction> asperaTransactionFuture = asperaTransferMgr.upload(bucketName, itemName, inputFile);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

Key Values

  • <bucket-name> - name of the bucket in your Object Storage service instance that has Aspera enabled.
  • <absolute-path-to-source-data> - directory and file name to upload to Object Storage.
  • <item-name> - name of the new object added to the bucket.

File Download

String bucketName = "<bucket-name>";
String outputPath = "<absolute-path-to-file>";
String itemName = "<item-name>";

// Create local file
File outputFile = new File(outputPath);
outputFile.createNewFile();

// Create AsperaTransferManager for FASP download
AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(COS_API_KEY_ID, s3Client)
    .withTokenManager(tokenManager)
    .withAsperaConfig(asperaConfig)
    .build();

// Download file
Future<AsperaTransaction> asperaTransactionFuture = asperaTransferMgr.download(bucketName, itemName, outputPath);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

Key Values

  • <bucket-name> - name of the bucket in your Object Storage service instance that has Aspera enabled.
  • <absolute-path-to-file> - directory and file name to save from Object Storage.
  • <item-name> - name of the object in the bucket.

Directory Upload

String bucketName = "<bucket-name>";
String directoryPath = "<absolute-path-to-directory-for-new-file>";
String directoryPrefix = "<virtual-directory-prefix>";
boolean includeSubDirectories = true;

// Load Directory
File inputDirectory = new File(directoryPath);

// Create AsperaTransferManager for FASP upload
AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(COS_API_KEY_ID, s3Client)
    .withTokenManager(tokenManager)
    .withAsperaConfig(asperaConfig)
    .build();

// Upload test directory
Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.uploadDirectory(bucketName, directoryPrefix, inputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

Key Values

  • <bucket-name> - name of the bucket in your Object Storage service instance that has Aspera enabled.
  • <absolute-path-to-directory> - directory of the files to be uploaded to Object Storage.
  • <virtual-directory-prefix> - name of the directory prefix to be added to each file upon upload. Use null or empty string to upload the files to the bucket root.

Directory Download

String bucketName = "<bucket-name>";
String directoryPath = "<absolute-path-to-directory>";
String directoryPrefix = "<virtual-directory-prefix>";
boolean includeSubDirectories = true;

// Load Directory
File outputDirectory = new File(directoryPath);

// Create AsperaTransferManager for FASP download
AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(COS_API_KEY_ID, s3Client)
    .withTokenManager(tokenManager)
    .withAsperaConfig(asperaConfig)
    .build();

// Download test directory
Future<AsperaTransaction> asperaTransactionFuture   = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

Key Values

  • <bucket-name> - name of the bucket in your Object Storage service instance that has Aspera enabled.
  • <absolute-path-to-directory> - directory to save downloaded files from Object Storage.
  • <virtual-directory-prefix> - name of the directory prefix of each file to download. Use null or empty string to download all files in the bucket.

Overriding Session Configuration on a Per Transfer Basis

You can override the multi-session configuration values on a per transfer basis by passing an instance of AsperaConfig to the upload and download overloaded methods. Using AsperaConfig you can specify the number of sessions and minimum file threshold size per session.

String bucketName = "<bucket-name>";
String filePath = "<absolute-path-to-file>";
String itemName = "<item-name>";

// Load file
File inputFile = new File(filePath);

// Create AsperaTransferManager for FASP upload
AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(API_KEY, s3Client)
.withTokenManager(TOKEN_MANAGER)
.withAsperaConfig(asperaConfig)
.build();

// Create AsperaConfig to set number of sessions
// and file threshold per session.
AsperaConfig asperaConfig = new AsperaConfig().
withMultiSession(10).
withMultiSessionThresholdMb(100);

// Upload test file and report progress
Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.upload(bucketName, itemName, inputFile, asperaConfig, null);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

Monitoring Transfer Progress

The simplest way to monitor the progress of your file/directory transfers is to use the isDone() property that returns true when your transfer is complete.

Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

while (!asperaTransaction.isDone()) {
    System.out.println("Directory download is in progress");

    //pause for 3 seconds
    Thread.sleep(1000 * 3);
}

You can also check if a transfer is queued for processing by calling the onQueue method on the AsperaTransaction. onQueue will return a Boolean with true indicating that the transfer is queued.

Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

while (!asperaTransaction.isDone()) {
    System.out.println("Directory download is in queueing: " + asperaTransaction.onQueue());

    //pause for 3 seconds
    Thread.sleep(1000 * 3);
}

To check if a transfer is in progress call the progress method in AsperaTransaction.

Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

while (!asperaTransaction.isDone()) {
    System.out.println("Directory download is in progress: " + asperaTransaction.progress());

    //pause for 3 seconds
    Thread.sleep(1000 * 3);
}

Every transfer by default will have a TransferProgress attached to it. The TransferProgress will report the number of bytes transferred and the percentage transferred of the total bytes to transfer. To access a transfer’s TransferProgress use the getProgress method in AsperaTransaction.

Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

while (!asperaTransaction.isDone()) {
    TransferProgress transferProgress = asperaTransaction.getProgress();

    //pause for 3 seconds
    Thread.sleep(1000 * 3);
}

To report the number of bytes transferred call the getBytesTransferred method on TransferProgress. To report the percentage transferred of the total bytes to transfer call the getPercentTransferred method on TransferProgress.

Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

while (!asperaTransaction.isDone()) {
    TransferProgress transferProgress = asperaTransaction.getProgress();

    System.out.println("Bytes transferred: " + transferProgress.getBytesTransferred());
    System.out.println("Percent transferred: " + transferProgress.getPercentTransferred());


    //pause for 3 seconds
    Thread.sleep(1000 * 3);
}

Pause/Resume/Cancel

The SDK provides the ability to manage the progress of file/directory transfers through the following methods of the AsperaTransfer object:

  • pause()
  • resume()
  • cancel()

There are no side-effects from calling either of the methods outlined above. Proper clean up and housekeeping is handled by the SDK.

The following example shows a possible use for these methods:

String bucketName = "<bucket-name>";
String directoryPath = "<absolute-path-to-directory>";
String directoryPrefix = "<virtual-directory-prefix>";
boolean includeSubDirectories = true;

AsperaTransferManager asperaTransferMgr = new AsperaTransferManagerBuilder(COS_API_KEY_ID, _cos)
    .withTokenManager(TOKEN_MANAGER)
    .build();

File outputDirectory = new File(directoryName);

System.out.println("Starting directory download...");

//download the directory from cloud storage
Future<AsperaTransaction> asperaTransactionFuture  = asperaTransferMgr.downloadDirectory(bucketName, directoryPrefix, outputDirectory, includeSubDirectories);
AsperaTransaction asperaTransaction = asperaTransactionFuture.get();

int pauseCount = 0;

while (!asperaTransaction.isDone()) {
    System.out.println("Directory download in progress...");

    //pause the transfer
    asperaTransfer.pause();

    //resume the transfer
    asperaTransfer.resume();

    //cancel the transfer
    asperaTransfer.cancel();
}

System.out.println("Directory download complete!");

Troubleshooting Aspera Issues

Issue: developers using the Oracle JDK on Linux or Mac OS X may experience unexpected and silent crashes during transfers

Cause: The native code requires its own signal handlers which could be overriding the JVM's signal handlers. It might be necessary to use the JVM's signal chaining facility.

IBM® JDK users or Microsoft® Windows users are not affected.

Solution: Link and load the JVM's signal chaining library.

  • On Linux locate the libjsig.so shared library and set the following environment variable:

    • LD_PRELOAD=<PATH_TO_SHARED_LIB>/libjsig.so
  • On Mac OS X locate the shared library libjsig.dylib and set the following environment variables:

    • DYLD_INSERT_LIBRARIES=<PATH_TO_SHARED_LIB>/libjsig.dylib
    • DYLD_FORCE_FLAT_NAMESPACE=0

Visit the Oracle® JDK documentation for more information about signal chaining.

Issue: UnsatisfiedLinkError on Linux

Cause: System unable to load dependent libraries. Errors such as the following may be seen in the application logs:

libfaspmanager2.so: libawt.so: cannot open shared object file: No such file or directory

Solution: Set the following environment variable:

LD_LIBRARY_PATH=<JAVA_HOME>/jre/lib/amd64/server:<JAVA_HOME>/jre/lib/amd64

Updating Metadata

There are two ways to update the metadata on an existing object:

  • A PUT request with the new metadata and the original object contents
  • Executing a COPY request with the new metadata specifying the original object as the copy source

Using PUT to update metadata

The PUT request overwrites the existing contents of the object so it must first be downloaded and re-uploaded with the new metadata.

public static void updateMetadataPut(String bucketName, String itemName, String key, String value) throws IOException {
    System.out.printf("Updating metadata for item: %s\n", itemName);

    //retrieve the existing item to reload the contents
    S3Object item = _cos.getObject(new GetObjectRequest(bucketName, itemName));
    S3ObjectInputStream itemContents = item.getObjectContent();

    //read the contents of the item in order to set the content length and create a copy
    ByteArrayOutputStream output = new ByteArrayOutputStream();
    int b;
    while ((b = itemContents.read()) != -1) {
        output.write(b);
    }

    int contentLength = output.size();
    InputStream itemCopy = new ByteArrayInputStream(output.toByteArray());

    //set the new metadata
    HashMap<String, String> userMetadata = new HashMap<String, String>();
    userMetadata.put(key, value);

    ObjectMetadata metadata = new ObjectMetadata();   
    metadata.setContentLength(contentLength);
    metadata.setUserMetadata(userMetadata);     

    PutObjectRequest req = new PutObjectRequest(bucketName, itemName, itemCopy, metadata);

    _cos.putObject(req);
    
    System.out.printf("Updated metadata for item %s from bucket %s\n", itemName, bucketName);
}

Using COPY to update metadata

public static void updateMetadataCopy(String bucketName, String itemName, String key, String value) {
    System.out.printf("Updating metadata for item: %s\n", itemName);

    //set the new metadata
    HashMap<String, String> userMetadata = new HashMap<String, String>();
    userMetadata.put(key, value);

    ObjectMetadata metadata = new ObjectMetadata();   
    metadata.setUserMetadata(userMetadata);     

    //set the copy source to itself
    CopyObjectRequest req = new CopyObjectRequest(bucketName, itemName, bucketName, itemName);
    req.setNewObjectMetadata(metadata);

    _cos.copyObject(req);

    System.out.printf("Updated metadata for item %s from bucket %s\n", itemName, bucketName);
}

Using Immutable Object Storage

Add a protection configuration to an existing bucket

This implementation of the PUT operation uses the protection query parameter to set the retention parameters for an existing bucket. This operation allows you to set or change the minimum, default, and maximum retention period. This operation also allows you to change the protection state of the bucket.

Objects written to a protected bucket cannot be deleted until the protection period has expired and all legal holds on the object are removed. The bucket's default retention value is given to an object unless an object specific value is provided when the object is created. Objects in protected buckets that are no longer under retention (retention period has expired and the object does not have any legal holds), when overwritten, will again come under retention. The new retention period can be provided as part of the object overwrite request or the default retention time of the bucket will be given to the object.

The minimum and maximum supported values for the retention period settings MinimumRetention, DefaultRetention, and MaximumRetention are a minimum of 0 days and a maximum of 365243 days (1000 years).

public static void addProtectionConfigurationToBucket(String bucketName) {
    System.out.printf("Adding protection to bucket: %s\n", bucketName);

    BucketProtectionConfiguration newConfig = new BucketProtectionConfiguration()
        .withStatus(BucketProtectionStatus.Retention)
        .withMinimumRetentionInDays(10)
        .withDefaultRetentionInDays(100)
        .withMaximumRetentionInDays(1000);

    cos.setBucketProtection(bucketName, newConfig);

    System.out.printf("Protection added to bucket %s\n", bucketName);
}

public static void addProtectionConfigurationToBucketWithRequest(String bucketName) {
    System.out.printf("Adding protection to bucket: %s\n", bucketName);

    BucketProtectionConfiguration newConfig = new BucketProtectionConfiguration()
        .withStatus(BucketProtectionStatus.Retention)
        .withMinimumRetentionInDays(10)
        .withDefaultRetentionInDays(100)
        .withMaximumRetentionInDays(1000);

    SetBucketProtectionConfigurationRequest newRequest = new SetBucketProtectionConfigurationRequest()
        .withBucketName(bucketName)
        .withProtectionConfiguration(newConfig);

    cos.setBucketProtectionConfiguration(newRequest);

    System.out.printf("Protection added to bucket %s\n", bucketName);
}

Check protection on a bucket

public static void getProtectionConfigurationOnBucket(String bucketName) {
    System.out.printf("Retrieving protection configuration from bucket: %s\n", bucketName;

    BucketProtectionConfiguration config = cos.getBucketProtection(bucketName);

    String status = config.getStatus();

    System.out.printf("Status: %s\n", status);

    if (!status.toUpperCase().equals("DISABLED")) {
        System.out.printf("Minimum Retention (Days): %s\n", config.getMinimumRetentionInDays());
        System.out.printf("Default Retention (Days): %s\n", config.getDefaultRetentionInDays());
        System.out.printf("Maximum Retention (Days): %s\n", config.getMaximumRetentionInDays());
    }
}

Upload a protected object

Objects in protected buckets that are no longer under retention (retention period has expired and the object does not have any legal holds), when overwritten, will again come under retention. The new retention period can be provided as part of the object overwrite request or the default retention time of the bucket will be given to the object.

Value Type Description
Retention-Period Non-negative integer (seconds) Retention period to store on the object in seconds. The object can be neither overwritten nor deleted until the amount of time specified in the retention period has elapsed. If this field and Retention-Expiration-Date are specified a 400 error is returned. If neither is specified the bucket's DefaultRetention period will be used. Zero (0) is a legal value assuming the bucket's minimum retention period is also 0.
Retention-expiration-date Date (ISO 8601 Format) Date on which it will be legal to delete or modify the object. You can only specify this or the Retention-Period header. If both are specified a 400 error will be returned. If neither is specified the bucket's DefaultRetention period will be used.
Retention-legal-hold-id string A single legal hold to apply to the object. A legal hold is a Y character long string. The object cannot be overwritten or deleted until all legal holds associated with the object are removed.
public static void putObjectAddLegalHold(String bucketName, String objectName, String fileText, String legalHoldId) {
    System.out.printf("Add legal hold %s to %s in bucket %s with a putObject operation.\n", legalHoldId, objectName, bucketName);

    InputStream newStream = new ByteArrayInputStream(fileText.getBytes(StandardCharsets.UTF_8));

    ObjectMetadata metadata = new ObjectMetadata();
    metadata.setContentLength(fileText.length());

    PutObjectRequest req = new PutObjectRequest(
        bucketName,
        objectName,
        newStream,
        metadata
    );
    req.setRetentionLegalHoldId(legalHoldId);

    cos.putObject(req);

    System.out.printf("Legal hold %s added to object %s in bucket %s\n", legalHoldId, objectName, bucketName);
}

public static void copyProtectedObject(String sourceBucketName, String sourceObjectName, String destinationBucketName, String newObjectName) {
    System.out.printf("Copy protected object %s from bucket %s to %s/%s.\n", sourceObjectName, sourceBucketName, destinationBucketName, newObjectName);

    CopyObjectRequest req = new CopyObjectRequest(
        sourceBucketName, 
        sourceObjectName, 
        destinationBucketName, 
        newObjectName
    );
    req.setRetentionDirective(RetentionDirective.COPY);
    

    cos.copyObject(req);

    System.out.printf("Protected object copied from %s/%s to %s/%s\n", sourceObjectName, sourceBucketName, destinationBucketName, newObjectName);
}

Extend the retention period of a protected object

The retention period of an object can only be extended. It cannot be decreased from the currently configured value.

The retention expansion value is set in one of three ways:

  • additional time from the current value (Additional-Retention-Period or similar method)
  • new extension period in seconds (Extend-Retention-From-Current-Time or similar method)
  • new retention expiry date of the object (New-Retention-Expiration-Date or similar method)

The current retention period stored in the object metadata is either increased by the given additional time or replaced with the new value, depending on the parameter that is set in the extendRetention request. In all cases, the extend retention parameter is checked against the current retention period and the extended parameter is only accepted if the updated retention period is greater than the current retention period.

Objects in protected buckets that are no longer under retention (retention period has expired and the object does not have any legal holds), when overwritten, will again come under retention. The new retention period can be provided as part of the object overwrite request or the default retention time of the bucket will be given to the object.

public static void extendRetentionPeriodOnObject(String bucketName, String objectName, Long additionalSeconds) {
    System.out.printf("Extend the retention period on %s in bucket %s by %s seconds.\n", objectName, bucketName, additionalSeconds);

    ExtendObjectRetentionRequest req = new ExtendObjectRetentionRequest(
        bucketName, 
        objectName)
        .withAdditionalRetentionPeriod(additionalSeconds);

    cos.extendObjectRetention(req);

    System.out.printf("New retention period on %s is %s\n", objectName, additionalSeconds);
}

List legal holds on a protected object

This operation returns:

  • Object creation date
  • Object retention period in seconds
  • Calculated retention expiration date based on the period and creation date
  • List of legal holds
  • Legal hold identifier
  • Timestamp when legal hold was applied

If there are no legal holds on the object, an empty LegalHoldSet is returned. If there is no retention period specified on the object, a 404 error is returned.

public static void listLegalHoldsOnObject(String bucketName, String objectName) {
    System.out.printf("List all legal holds on object %s in bucket %s\n", objectName, bucketName);

    ListLegalHoldsResult result = cos.listLegalHolds(
        bucketName, 
        objectName
    );

    System.out.printf("Legal holds on bucket %s: \n", bucketName);

    List<LegalHold> holds = result.getLegalHolds();
    for (LegalHold hold : holds) {
        System.out.printf("Legal Hold: %s", hold);
    }
}

Create a hosted static website

This operation requires an import statement to be added:

import com.ibm.cloud.objectstorage.services.s3.model.model.BucketWebsiteConfiguration;

This operation provides the following upon configuration and requires a correctly configured client:

  • Bucket configuration for suffix (index document)
  • Bucket configuration for key (error document)
cosClient.setBucketWebsiteConfiguration("<bucket_name>", new BucketWebsiteConfiguration("index.html", "error.html"));

Next Steps

For more information, see the Javadoc. The source code for the project can be found in the GitHub repository.