Announcing UNLIMITED Public Preview of Metadata Caching for Azure Premium SMB/REST File Shares

Azure Files We are pleased to announce this. unlimited Public preview of metadata caching for premium SMB/REST file sharing tiers.

The unlimited public preview allows customers to automatically self-service the onboarding process through Feature Enrollment (AFEC) in supported regions.

Feature Overview

Metadata caching is an enhancement that aims to reduce metadata latency for file workloads running on Windows/Linux clients. In addition to reducing metadata latency, workloads A consistent latency experience that can make metadata-intensive workloads more predictable and deterministic. Additionally, reduced metadata latency increases data read/write (IOPS) and throughput. Once the metadata caching feature is registered within your subscription, there are no additional costs or operational management overhead to use this feature.

The following metadata APIs can benefit from metadata caching:

• make: Create new file; maximum 30% faster
• open: File open maximum 55% faster
• close: Close file up 45% faster
• delete: Maximum file deletion 25% faster

Workloads that perform large amounts of metadata operations (create/open/close/delete) on SMB/REST premium file shares benefit the most compared to workloads that are primarily data IO (e.g. databases).

Examples of metadata-heavy workloads include:

• Web\App Service: Frequently accessed files for CMS\LMS services such as Moodle\WordPress.
• Indexing\Batch Operations: Large-scale processing using Azure Kubernetes or Azure Batch.
• Virtual Desktop Infrastructure: Azure Virtual Desktop\Citrix users who need to manage their home directories or VDI applications.
• Business applications: Custom LOBs or legacy applications that require “lift and shift”.
• CI\CD – DevOps Pipeline: Build, test, and deploy workloads like Jenkins open source automation

GitHub solution with metadata caching

Performance improvements are expected through the metadata cache.

• 2-3 times Improved metadata latency consistency
• Up to 3 times Increased scale of metadata operations
• Improved metadata latency More than 30%
• Increased IOPS and bandwidth Up to 60%

How to get started

To get started: register Subscribe to the metadata cache feature using the Azure portal or PowerShell.

• central australia
• Geo India West
• south india
• Central Mexico
• eastern norway
• Central Poland
• Central Spain
• central sweden
• northern switzerland
• UAE Northern
• western usa 3

memo: As we expand regional support for the Metadata Caching feature, premium file storage accounts in that region will be automatically onboarded for all subscriptions registered for the Metadata Caching feature.

Who should participate?

Whether you have a new workload looking to take advantage of file sharing, or an existing workload you want to improve. You should encourage onboarding all workloads/usage patterns that contain metadata, especially metadata-heavy workloads that primarily consist of create/open/close or delete requests.

To ensure that your workload contains metadata, you can use Azure Monitor to split transactions by API dimension as described below. article

thank you

Azure Files Team