About protecting data in Amazon for long-term retention
The following Amazon cloud storage options are available for long-term retention of data:
Consider the following table when deciding between GLACIER and GLACIER_VAULT storage classes:
GLACIER and GLACIER_DEEP_ARCHIVE storage class | GLACIER_VAULT storage class |
|---|---|
|
GLACIER and GLACIER_DEEP_ARCHIVE storage class corresponds to uploading data through S3 endpoint and transitioning the data to Glacier. |
GLACIER_VAULT storage class corresponds to uploading data using Amazon Glacier services to vault. |
|
For GLACIER and GLACIER_DEEP_ARCHIVE storage class, the metadata is stored in STANDARD storage class. |
For GLACIER_VAULT storage class, the metadata is stored in STANDARD and GLACIER_VAULT storage classes. |
|
Cost of operation for GLACIER is approximately 2% higher than GLACIER_VAULT. |
Cost of operation for GLACIER and GLACIER_VAULT storage class is approximately the same with GLACIER being approximately 2% higher than GLACIER_VAULT. |
|
Use GLACIER and GLACIER_DEEP_ARCHIVE storage class if you do not plan to use immutable vault lock capability. |
Use GLACIER_VAULT storage class if you plan to use the immutable vault lock policy for compliance or to protect your data from ransomware attack. |
|
GLACIER GLACIER_DEEP_ARCHIVE storage class has a configurable retrieval retention period. Thus, it is useful for restores that may take more time due to size and speed. |
The retrieval retention period for GLACIER_VAULT storage class is fixed, that is 24 hours. See Restoring from Amazon Glacier vault spans more than 24 hours for single fragment . |
|
As objects get uploaded, Amazon provides visibility for all objects and their storage class property through the Amazon S3 service console. Hence, NetBackup images that are created using GLACIER and GLACIER_DEEP_ARCHIVE storage class have better visibility through the Amazon S3 service console. |
Amazon takes 24 hours to refresh archive inventory. Hence, archives uploaded during backup done using GLACIER_VAULT storage class are reflected in the Amazon Glacier service console only after 24 hours. However, you can get some visibility of backups using the Amazon S3 service console through the metadata generated during the backup. Amazon Glacier service console does not provide any visibility for individual archives. |
|
There are architectural differences between GLACIER_VAULT storage class (using Amazon Glacier services) and GLACIER and GLACIER_DEEP_ARCHIVE storage class (using Amazon S3 services). This results in difference in speed that must be considered when selecting a storage class. |
There are architectural differences between GLACIER_VAULT storage class (using Amazon Glacier services) and GLACIER and GLACIER_DEEP_ARCHIVE storage class (using Amazon S3 services). This results in difference in speed that must be considered when selecting a storage class. |
|
Storage cleanup handling on failure is better for GLACIER and GLACIER_DEEP_ARCHIVE storage class. |
Storage cleanup handling on failure is better for GLACIER storage class as compared to GLACIER_VAULT storage class. |