Eliminare caricamenti multiparte incompleti su Amazon S3 utilizzando un AWS SDK

Per arrestare i caricamenti in più parti in corso o incompleti per qualsiasi motivo, è possibile creare un elenco e poi eliminarli come mostrato nell’esempio seguente.

    /**
     * Aborts all incomplete multipart uploads from the specified S3 bucket.
     *
     * @param bucketName the name of the S3 bucket
     */
    public static void abortIncompleteMultipartUploadsFromList(String bucketName) {
        ListMultipartUploadsRequest listMultipartUploadsRequest = ListMultipartUploadsRequest.builder()
            .bucket(bucketName)
            .build();

        ListMultipartUploadsResponse response = s3Client.listMultipartUploads(listMultipartUploadsRequest);
        List<MultipartUpload> uploads = response.uploads();

        AbortMultipartUploadRequest abortMultipartUploadRequest;
        for (MultipartUpload upload : uploads) {
            abortMultipartUploadRequest = AbortMultipartUploadRequest.builder()
                .bucket(bucketName)
                .key(upload.key())
                .expectedBucketOwner(accountId)
                .uploadId(upload.uploadId())
                .build();

            AbortMultipartUploadResponse abortMultipartUploadResponse = s3Client.abortMultipartUpload(abortMultipartUploadRequest);
            if (abortMultipartUploadResponse.sdkHttpResponse().isSuccessful()) {
                logger.info("Upload ID [{}] to bucket [{}] successfully aborted.", upload.uploadId(), bucketName);
            }
        }
    }

Per eliminare i caricamenti in più parti incompleti avviati prima o dopo una data, è possibile eliminare selettivamente i caricamenti in più parti in base a un point-in-time specifico, come mostrato nell’esempio seguente.

    /**
     * Aborts incomplete multipart uploads older than the specified point in time.
     *
     * @param bucketName  the name of the S3 bucket
     * @param pointInTime the cutoff time; uploads initiated before this are aborted
     */
    static void abortIncompleteMultipartUploadsOlderThan(String bucketName, Instant pointInTime) {
        ListMultipartUploadsRequest listMultipartUploadsRequest = ListMultipartUploadsRequest.builder()
            .bucket(bucketName)
            .build();

        ListMultipartUploadsResponse response = s3Client.listMultipartUploads(listMultipartUploadsRequest);
        List<MultipartUpload> uploads = response.uploads();

        AbortMultipartUploadRequest abortMultipartUploadRequest;
        for (MultipartUpload upload : uploads) {
            logger.info("Found multipartUpload with upload ID [{}], initiated [{}]", upload.uploadId(), upload.initiated());
            if (upload.initiated().isBefore(pointInTime)) {
                abortMultipartUploadRequest = AbortMultipartUploadRequest.builder()
                    .bucket(bucketName)
                    .key(upload.key())
                    .expectedBucketOwner(accountId)
                    .uploadId(upload.uploadId())
                    .build();

                AbortMultipartUploadResponse abortMultipartUploadResponse = s3Client.abortMultipartUpload(abortMultipartUploadRequest);
                if (abortMultipartUploadResponse.sdkHttpResponse().isSuccessful()) {
                    logger.info("Upload ID [{}] to bucket [{}] successfully aborted.", upload.uploadId(), bucketName);
                }
            }
        }
    }

Se hai accesso all’ID di caricamento dopo aver avviato un caricamento in più parti, puoi eliminare il caricamento in corso utilizzando l’ID.

    /**
     * Aborts a multipart upload using the upload ID.
     *
     * @param bucketName the name of the S3 bucket
     * @param key        the object key
     */
    static void abortMultipartUploadUsingUploadId(String bucketName, String key) {
        String uploadId = startUploadReturningUploadId(bucketName, key);
        AbortMultipartUploadResponse response = s3Client.abortMultipartUpload(b -> b
            .uploadId(uploadId)
            .bucket(bucketName)
            .key(key));

        if (response.sdkHttpResponse().isSuccessful()) {
            logger.info("Upload ID [{}] to bucket [{}] successfully aborted.", uploadId, bucketName);
        }
    }

Per eliminare sistematicamente i caricamenti in più parti incompleti più vecchi di un determinato numero di giorni, imposta una configurazione del ciclo di vita per il bucket. L’esempio seguente mostra come creare una regola per eliminare i caricamenti incompleti con più di 7 giorni.

    /**
     * Configures a lifecycle rule to abort incomplete multipart uploads after 7 days.
     *
     * @param bucketName the name of the S3 bucket
     */
    static void abortMultipartUploadsUsingLifecycleConfig(String bucketName) {
        Collection<LifecycleRule> lifeCycleRules = List.of(LifecycleRule.builder()
            .abortIncompleteMultipartUpload(b -> b.
                daysAfterInitiation(7))
            .status("Enabled")
            .filter(SdkBuilder::build) // Filter element is required.
            .build());

        PutBucketLifecycleConfigurationResponse response = s3Client.putBucketLifecycleConfiguration(b -> b
            .bucket(bucketName)
            .lifecycleConfiguration(b1 -> b1.rules(lifeCycleRules)));

        if (response.sdkHttpResponse().isSuccessful()) {
            logger.info("Rule to abort incomplete multipart uploads added to bucket.");
        } else {
            logger.error("Unsuccessfully applied rule. HTTP status code is [{}]", response.sdkHttpResponse().statusCode());
        }
    }