Q: How is Amazon EFS designed to provide high durability and availability?
By default, every Amazon EFS file system object (i.e. directory, file, and link) is redundantly stored across multiple AZs for all regional storage classes. If you select Amazon EFS One Zone storage classes, your data is redundantly stored within a single AZ. Amazon EFS is designed to sustain concurrent device failures by quickly detecting and repairing any lost redundancy. In addition, a file system can be accessed concurrently from all Availability Zones in the region where it is located, which means that you can architect your application to failover from one AZ to other AZs in the region in order to ensure the highest level of application availability. Mount targets are designed to be highly available within an AZ for all Amazon EFS storage classes.
Q: How durable is Amazon EFS?
Amazon EFS is designed to provide 99.999999999% (11 9’s) of durability over a given year. In addition, EFS Standard and EFS Standard-IA storage classes are designed to sustain data in the event of an entire Availability Zone loss. Because EFS One Zone storage classes store data in a single Amazon Web Services Availability Zone, data stored in these storage classes may be lost in the event of a disaster or other fault within the Availability Zone that affects all copies of the data, or in the event of Availability Zone destruction. As with any environment, best practice is to have a backup and to put in place safeguards against accidental deletion. For Amazon EFS data, that best practice includes replicating your file system across Regions using Amazon EFS Replication, and a functioning, regularly tested backup using Amazon Backup. File systems using EFS One Zone storage classes are configured to automatically back up files by default at file system creation, unless you choose to disable this functionality.
Q: What failure modes do I have to consider when using Amazon EFS One Zone compared to Standard storage classes?
File systems using Amazon EFS One Zone storage classes are not resilient to a complete AZ outage. In the event of an AZ outage, you will experience a loss of availability, because your file system data is not replicated to a different AZ. In the event of disaster or fault within an AZ affecting all copies of your data, or a permanent AZ loss, you may experience loss of data that has not been replicated using Amazon EFS Replication to keep an up-to-date copy of your file system in a second Amazon Web Services Region or an AZ. EFS Replication is designed to meet a recovery point objective (RPO) and recovery time objective (RTO) of minutes. You can also use Amazon Backup to store additional copies of your file system data and restore them to a new file system in an AZ or Region of your choice. Amazon EFS file system backup data created and managed by Amazon Backup is replicated to 3 AZs and is designed for 99.999999999% (11 9’s) durability.
Q. How can I guard my EFS One Zone file system against the loss of an AZ?
You can use Amazon EFS Replication or Amazon Backup to guard your EFS One Zone file system against the loss of an AZ. Amazon EFS Replication replicates your file system data to another Amazon Web Services Region or within the same Region in a few clicks, without requiring additional infrastructure or a custom process to monitor and synchronize data changes. EFS replication is continuous and designed to provide a recovery point objective (RPO) and a recovery time objective (RTO) of minutes for most file systems.
Backups are enabled by default for all file systems using Amazon EFS One Zone storage classes. You can disable this setting when creating file systems. You are able to restore your file data from a recent backup to a newly created file system in any operating AZ in the event of an AZ loss. If Amazon EFS is impacted by an AZ loss, and your data is stored in One Zone storage classes, you may experience data loss for files that have changed since the last automatic backup.
Q: What is Amazon EFS Replication?
EFS Replication allows you to replicate your file system data to another Amazon Web Services Region or within the same Region in a few clicks, without requiring additional infrastructure or a custom process to monitor and synchronize data changes. Amazon EFS Replication automatically and transparently replicates your data to a second file system in a Region or AZ of your choice. You can use the Amazon EFS console, Amazon Web Services CLI, and APIs to enable replication on an existing file system. EFS Replication is continuous and designed to provide a recovery point objective (RPO) and a recovery time objective (RTO) of minutes, enabling you to meet your compliance and business continuity goals.
Q: Why should I use EFS Replication?
If you have requirements to maintain a copy of your file system hundreds of miles apart for purposes of disaster recovery, compliance, or business continuity planning, EFS Replication can help you meet those requirements. For applications that require a low network latency cross-region access, Amazon EFS Replication provides a read-only copy in the region of your choice. With Amazon EFS Replication, you can cost-optimize and save up to 75% on your disaster recovery storage costs by using low-cost EFS One Zone storage classes and a 7-day age-off lifecycle management policy for your destination file system. There is no need to build and maintain a custom process for data replication. EFS Replication also makes it easy to monitor and alarm on your RPO status using Amazon CloudWatch.
Q: How do I get started with EFS Replication?
Using the Amazon EFS console, simply enable Replication on the file system you want to replicate (source file system) and choose the Region or AZ where you want to store the replica (destination). You can also use the CreateReplicationConfiguration API from the Amazon Web Services CLI or SDK to enable EFS Replication. As part of configuring EFS Replication, you’ll choose the Region in which to create your replica. If you choose to use EFS One Zone storage classes for your replica, you must also select your file system’s AZ. Once EFS Replication is enabled, Amazon EFS will automatically create a new destination file system in the destination Region or AZ you’ve selected. You can select the destination file system’s lifecycle management policy, backup policies, provisioned throughput, mount targets, and access points independent of the source file system. For example, you can optimize the destination file system storage costs by enabling EFS Lifecycle Management with a shorter age-off policy (such as 7 days) when compared to the source file system’s age-off policy (such as 7, 14, 30, 60, or 90 days). EFS Replication configurations such as the replication pair (source and destination), replication status, and last completed replication timestamp can be accessed using the DescribeReplicationConfigurations API.
Q: How does EFS Replication work?
When you enable EFS Replication on a file system, Amazon EFS automatically creates a new file system in the destination region and begins copying your data into it. Once the initial copy is completed, EFS Replication copies changes incrementally to deliver an RPO of minutes for most file systems. EFS Replication preserves all metadata, such as owners and permissions, when copying changes to files and folders. While EFS Replication is enabled, your destination file system is in read-only mode and can be updated only by EFS Replication. In the event that your source file system is unavailable, you can failover to the destination file system by deleting replication. Deleting the Replication makes the destination file system writeable for your applications to use.
Q: Can I change my destination file system’s settings when EFS Replication is enabled?
Yes. When EFS Replication is enabled, you can modify your destination file system configuration settings, such as its lifecycle management policy including intelligent-tiering, backup policy, mount targets, access points, and provisioned throughput. All destination file systems are created with encryption of data at rest enabled irrespective of the source file system setting. You cannot change the performance mode of the destination file system. It always matches that of the source file system, except when you create a One Zone replica. In that case, General Purpose performance mode is used because Max I/O performance mode is not supported by EFS One Zone storage classes.
Q: Can I change which Region I’m replicating data to?
No. In order to change the Region of your destination, you first have to delete the replication configuration between your source and destination file system. You can then create a new replication configuration from the source by selecting the desired Region. Amazon EFS will create a new destination file system in the selected Region and begin to replicate the source file system's contents.
Q: Can I delete my source or destination file system if they’re part of a replication pair?
You cannot delete either your source or your destination file system if it’s part of a replication pair. In order to delete one of the file systems in the pair, you first need to delete the replication configuration.
Q: Is my replica file system point-in-time consistent?
No. EFS Replication doesn’t provide point-in-time consistent replication. EFS Replication publishes a timestamp metric on Amazon CloudWatch called TimeSinceLastSync. All changes made to your source file system at least as of the published time will be copied over to the destination. Changes to your source file system after the recorded time may not have been replicated over. You can monitor the health of your EFS Replication using Amazon CloudWatch. If you interrupt the replication process due to a disaster recovery event, some files from the source file system may have transferred over but are not yet copied to their final locations on your destination file system. These files and their contents can be found on your destination file system in a lost+found directory created by EFS Replication under the root directory.
Q: Can I select the VPC in which my mount targets are created?
Yes. When you enable EFS Replication for the first time, the replica file system will be automatically created for you. It’s created in the Region of your choosing without mount targets. You can then create mount targets for your replica file system in the VPC of your choosing. You can also change the VPC for your replica file system by deleting any existing mount targets and creating new ones in a VPC of your choosing.
Q: How can I utilize my destination file system while replication is enabled and when replication is deleted?
When your replication is in Enabled state, only EFS Replication is allowed to make changes to your destination file system. You can access your replica in the read-only mode during this time. In the event of a disaster you can fail over to your destination file system by deleting your replication configuration from the Amazon EFS console or by using the DeleteReplicationConfiguration API. When you delete the Replication, Amazon EFS will stop replicating additional changes and make the destination file system writeable. You can then point your application to your destination file system to continue your operations. You can use the Amazon EFS console or the DescribeReplicationConfigurations API call to check your destination file system status after you’ve failed over.
Q: Is the data for my file system replica encrypted in transit and at rest?
For all file systems, Amazon EFS automatically and transparently encrypts all Amazon EFS network traffic using Transport Layer Security (TLS) version 1.2. Your destination file system is created with encryption at rest enabled. You can select an encryption key from those available in the destination Region Amazon Key Management Service (KMS) or by using the default service “aws/elasticfilesystem” key in the Region where your destination file system is located.
Q: What permissions do I need to use EFS Replication?
To create and delete a replication, your Amazon IAM or resource-based policy must have permission for the Amazon EFS API calls CreateFileSystem, CreateReplicationConfiguration, and DescribeReplicationConfigurations.
Q: Does my replication traffic go over the public internet?
No. EFS Replication traffic always stays on the China Amazon Web Services backbone.
Q: Can I use EFS Replication to replicate my file system to more than one Amazon Web Services Region or to multiple file systems within a second Region?
No. EFS Replication supports replication between exactly two file systems.
Q: Can I replicate Amazon EFS file systems across Amazon Web Services accounts?
No. Amazon EFS does not support replicating file systems to a different Amazon Web Services account.
Q: Does EFS Replication consume my file system burst credits, IOPS limit, and throughput limits?
No. EFS Replication activity does not consume burst credits or count against the file system IOPS and throughput limits for either file system in a replication pair.
Q: Can I expect my destination file system to be available as soon as I enable EFS Replication?
Yes. When you first enable EFS Replication, your replica file system will be created in read-only mode and your entire source file system will be copied to the destination you selected. The time to complete this operation depends on the size of your source file system. Although you can failover to your destination file system at any time, it is recommended that you wait until the copy is complete to minimize data loss. You can monitor the progress of your replication from the Amazon EFS console, which displays a timestamp that indicates the last time your source file system and destination file system were synchronized.
 | Cloud File Storage | FAQs (2024))
