Amazon said Mountpoint for Amazon S3 is now generally available during the most recent AWS Storage Day event. The new open-source file client offers Amazon S3's elastic storage and throughput through a file interface, enabling data transmission between each EC2 instance and the object storage at up to 100 Gb/second.
Mountpoint for Amazon S3, which was first announced in preview earlier this year, is built for workloads like data lake apps that do sequential writes and sequential and random reads without requiring full POSIX semantics. Vice President at AWS and top evangelist Jeff Barr explains:
The S3 APIs and the AWS SDKs are used often by AWS customers to create software that can list, read, and manipulate the contents of an S3 bucket. To access files in UNIX manner, such as reading directories, opening & reading existing files, and creating & writing new ones, many clients already have apps, commands, tools, and processes in place. These customers have requested that we provide a formal, enterprise-ready client that supports fast, scalable access to S3.
The new open-source client is also frequently used for machine learning training, graphics rendering, simulating autonomous vehicles, and ETL. When compared to a managed network file system (NFS), Mountpoint offers data transfer rates of up to 100 Gb/second between each EC2 instance and the S3 bucket, converting local file system API calls to S3 object API calls and lowering expenses. Bart adds:
These operations—LIST, GET, PUT, and so forth—are really translated by the Linux Virtual Filesystem (VFS) into calls to Mountpoint, which then transform those calls into calls to S3. Utilizing network bandwidth effectively is a goal of Mountpoint, which boosts throughput and enables you to lower computation expenses by completing more work in a shorter amount of time.
Several Reddit users have questioned how the new offering stacks up against the open-source s3fs and for-profit ObjectiveFS alternatives that are already available. Senior cloud developer Fernando Schubert from Human Made says:
It's interesting that this is now officially supported, although s3fs/fuse had already been doing this for some time in order to benchmark features.
One of the objectives of the AWS client is highlighted by Jukka Forsgren, senior solutions architect at the company:
I was on call around-the-clock at my prior employer for a huge number of customer AWS setups. We frequently had to respond to notifications about stuck S3 mounts on EC2 Linux instances in the middle of the night. As a result, it was wonderful to see that we had recently released our own client, named Mountpoint for Amazon S3, with stability as one of its main design objectives.
Mountpoint has some limitations on file operations and is not a general-purpose networked file system. The rust-powered client, which is based on the AWS Common Runtime (CRT) library, is able to read files up to 5 TB in size but is unable to create new files or delete old ones. It also does not allow symbolic links or file locking.
The archive classes Glacier Flexible Retrieval, Glacier Deep Archive, Intelligent-Tiering Archive Access Tier, and Intelligent-Tiering Deep Archive Access Tier are not supported by Mountpoint at this time. As a result, not all S3 storage classes are currently supported.
Developers are responsible for the underlying S3 operations; there are no additional fees for using Mountpoint. The roadmap for Mountpoint is accessible on GitHub and is available in RPM format.
The tamilgun is a free public website that allows you to search for guides and it is sharing information about the business, education, finance and technology.