alluxio enterprise Subscriptions

Alluxio Data Platform provides several editions designed for different workloads. Choose the edition based on the unique needs of your workload. 

Alluxio Enterprise AI is designed to accelerate model training and serving, with intelligent caching tailored to I/O patterns of AI.  Enterprise AI also provides efficient data loading to eliminate data replication, and maximizes the ROI of your AI platform by enabling up to 90% GPU utilization.

Alluxio Enterprise Data accelerates SQL queries, provides a unified data access point, and enables up to 70% in cloud savings. No need to copy data or rewrite your applications. You can spin up compute wherever it’s most cost-effective, enabling true multi-cloud freedom.

Alluxio Enterprise Edition is available as an annual subscription with license and support included. Alluxio Community Edition is available for free without support and only recommended for test/dev environments. 

Performance & Scalability

Community Edition

Enterprise AI

Enterprise DA

An advanced architecture for metadata and cache management that allows your existing platform to handle up to 100 billion objects.
Also provides RESTful API for Python-based applications to interact with storage easily at high speed.

Alluxio caches both data and metadata for underlying storage systems to provide lower latency (up to 10x) and higher throughput. Learn more

Reduce the number of operations to underlying cloud storage and eliminate associated charges. Learn more

Data access patterns for large scale analytics & AI/ML training workloads vary, with data either stored in some large structured files or many small semi-structured/unstructured files.
This feature intelligently identifies optimizations for a specific workload dynamically to tune the system for optimal performance based on the split between read & write operations, file sizes, file types, number of files, etc.

This feature allows multiple Alluxio clusters to mount the same (or intersecting) UFS path and keep files synchronized across the clusters. For example, assume there are two Alluxio clusters called C1 and C2. Both clusters mount the S3 bucket s3://my-bucket/ to the folder /mnt in Alluxio. Now when cluster C1 modifies a file on path /mnt the modification will also be visible on cluster C2. Learn More

Performance Comparison

numjobs

Community Edition

Enterprise Edition

FIO Hot Read  Sequential Read Performance (direct = 1, bs=256k filesize=100G)

1

0.6 GiB/s

2.0 GiB/s

2

1.1 GiB/s

4.3 GiB/s

4

2.3 GiB/s

7.2 GiB/s

8

4.5 GiB/s

7.4 GiB/s

16

7.9 GiB/s

8.1 GiB/s

32

8.8 GiB/s

8.0 GiB/s

64

NA

8.1GiB/s

FIO Hot Read Random Read Performance (direct = 1, bs=256k filesize=100G)

1

0.04 GiB/s

0.1 GiB/s

2

0.1 GiB/s

0.3 GiB/s

4

0.1 GiB/s

0.8 GiB/s

8

0.2 GiB/s

2.1 GiB/s

16

0.4 GiB/s

4.2 GiB/s

32

0.3 GiB/s

7.3 GiB/s

64

NA

7.4 GiB/s

* N/A: Due to some reasons, the FIO command reports an error

  • Alluxio AI 
  • Libfuse + Linux
  • FIO
  • 1 FUSE + 1 Worker (Access worker remotely via FUSE)
    •   FUSE：AWS c5n.metal
    •   Worker：AWS i3en.metal

* Latest Enterprise Edition: AI-3.1 version

* Latest community version: version 2.9.3

Product Support & Services

Community Edition

Enterprise AI

Enterprise DA

24×7 Support

Emergency Patching

Professional Services – Health Check

Services and Best Practices

DevOps Capabilities

Community Edition

Enterprise AI

Enterprise DA

Integrate Alluxio into the big data stack without any changes to the location metadata of the Hive Metastore. For example, when use S3, the code path stays as s3:// instead of alluxio:// so that client code or table definition doesn’t need to change. Learn more

Automatically mount folders that don’t exist in Alluxio space without any administrative intervention to simplify operations. With this, admin doesn’t have to manually identify and mount the buckets.  Learn more

Enterprise Security & Compliance

Community Edition

Enterprise AI

Enterprise DA

An audit log of data accesses including the time, user, operation, and success/failure. Learn more

Enable access to a Kerberos secured under store. Learn more

Authenticate users through Kerberos to validate identity, so that unauthenticated user cannot access Alluxio. Learn more

By adopting the open authentication protocol for S3 API, user identities will be verified before their requests are processed. User now can leverage SSO system, and provide token to Alluxio to be authenticated and passed onto authorization agent. Learn more

Enables administrators to centralize permission management for various resources by allowing plugins for Ranger (Apache, CDP, Privacera) and OPA (Styra). With this, one can achieve centralized table level or file level access control. Learn more about Ranger and OPA

Ensures the data is encrypted when on disk, so that attacker cannot access unencrypted data even if they get hold of the disk. Learn more

TLS 1.2 support for communication between all the Alluxio components (client, master, worker, job master, and job worker) as well as between Alluxio and the under store. When this enabled, attacker cannot intercept a communication and read unencrypted data. Learn more

Data Management

Community Edition

Enterprise AI

Enterprise DA

Abstraction that allows applications to access multiple independent storage systems through the same namespace and interface. Learn more

Manage storage resources with media-aware tiering policies across memory, SSD and HDD within Alluxio. Learn more

Replicate data to any supported data store to support HA or other operational needs. Learn more

Define policies based on date created, updated or accessed and move or migrate data across external storage systems online. One example could be: HDFS to S3 data migration. For example, one can set policy to archive data > 30 days old or haven’t been accessed to to S3, and leave data < 30 days old stays at the default location, while access URI is the same. Learn more

Deployment

Community Edition

Enterprise AI

Enterprise DA

This feature allows user to manage the Alluxio clusters through Alluxio operator. For example, user can deploy the Alluxio cluster with Alluxio operator and understand the resource consumption with consumption report. Learn More