Welcome to the WEKA Documentation Portal, your guide to the latest WEKA version. Whether you're a newcomer or a seasoned user, explore topics from system fundamentals to advanced optimization strategies. Choose your WEKA version from the top menu for version-specific documentation.
Important: This documentation applies to the WEKA system's latest minor version (4.2.X). For information on new features and supported prerequisites released with each minor version, refer to the relevant release notes available at get.weka.io.
Check the release notes for details about any updates or changes accompanying the latest releases.
Sevii AI quickly delivers answers from WEKA documentation. Type your question and click . For the best results, ask clear, context-rich questions.
This portal encompasses all documentation essential for comprehending and operating the WEKA system. It covers a range of topics:
WEKA system overview: Delve into the fundamental components, principles, and entities constituting the WEKA system.
Planning and installation: Discover prerequisites, compatibility details, and installation procedures for WEKA clusters on bare metal, AWS, GCP, and Azure environments.
Getting started with WEKA: Initiate your WEKA journey by learning the basics of managing a WEKA filesystem through the GUI and CLI, executing initial IOs, and exploring the WEKA REST API.
Performance: Explore the results of FIO performance tests on the WEKA filesystem, ensuring optimal system performance.
WEKA filesystems & object stores: Understand the role and management of filesystems, object stores, filesystem groups, and key-management systems within WEKA configurations.
Additional protocols: Learn about the supported protocols—NFS, SMB, and S3—for accessing data stored in a WEKA filesystem.
Operation guide: Navigate through various WEKA system operations, including events, statistics, user management, upgrades, expansion, and more.
Billing & licensing: Gain insights into WEKA system licensing options and alternative billing approaches.
Monitor the WEKA cluster: Effectively monitor your WEKA cluster by deploying the WEKA Management Server (WMS) alongside tools like Local WEKA Home, WEKAmon, and SnapTool.
WEKA support: Find guidance on obtaining support for the WEKA system and effectively managing diagnostics.
Best practice guides: Explore our carefully selected guides, starting with WEKA and Slurm integration, to discover expert-recommended strategies and insights for optimizing your WEKA system and achieving peak performance in various scenarios.
The documentation marks the CLI mandatory parameters with an asterisk (*).
We welcome your feedback to improve our documentation. Include the document version and topic title with your suggestions and email them to . For technical inquiries, contact our . Thank you for helping us maintain high-quality resources.
4.2.18
Maintenance release.
Removed Azure deployment sections due to a known network issue in Azure deployments that causes performance degradation in versions 4.1 to 4.4.0. This issue is resolved in version 4.4.1 and later.
For updated guidance, see the sections in the Version 4.4 documentation.
4.2.17
DPDK and AWS Xen-on-Nitro incompatibility identified: Customers using legacy Amazon EC2 Xen-on-Nitro instances must contact AWS Support to request their account be added to the deny list.
Corrected client address representation in WEKA events: Resolved an issue where client addresses were displayed as loopback addresses; they now display the correct IP address.
Duplicate management address alert added: WEKA now generates an alert when duplicate management addresses are configured, indicating a binding failure.
Synchronous Snap: The Synchronous Snap feature, which allows incremental snapshots to be downloaded from an object store, was temporarily disabled in version 4.2.3. It has been re-enabled in version 4.3.0.
4.2.16
New event: DriverNotAccepting: Introduced a new event signaling that all I/O operations on a frontend are unresponsive or hanging.
New event: DriveImmediateShutdown: Added a new event to indicate an NVMe failure, triggering an immediate shutdown of the affected drive to prevent further system impact.
Network virtual function limitation: Implemented the ability to limit the quantity of network virtual functions (VFs). This feature is particularly relevant for clusters with high core counts using Intel E810 NICs.
Maximum number of processes increased: The maximum number of backend processes, drive processes, management processes, and total processes have all been increased.
Amazon Linux 2023: Added WEKA backend and client support for Amazon Linux 2023 with x86_64 kernel distribution.
New CLI reference guide: This CLI reference guide is generated from the output of running the weka command with the help option. It provides detailed descriptions of available commands, arguments, and options. >>>
Added support for GCP regions asia-southeast2 and europe-central2 in Terraform configuration.
Added Dell PowerScale S3 (version 9.8.0.0 and higher) to the certified object stores.
4.2.15
Added a verification step for LLQ and WC in the upgrade workflow. To ensure proper LLQ functionality after upgrades, verify that Write Combining (WC) is enabled in the igb_uio driver.
Extended support for RHEL/Rocky Linux 8.10 on backends and clients.
4.2.14
Extended support for operating systems on:
Clients: RHEL/Rocky Linux 9.4, AlmaLinux 9.4, 8.10, Debian 12.
Backends: RHEL/Rocky Linux 9.4.
Update: Object store types now included in Analytics reports on WEKA Home.
New topic:
4.2.12.92
Extended support for Linux kernel to Ubuntu 22:
5.19, 6.2, 6.5.
4.2.12
Extended support for operating systems on clients: RHEL/Rocky Linux 9.3, Oracle Linux 9.
Extended support for Las_v3 machine types for backends on Microsoft Azure.
4.2.11
Customers can now see originating IP address information in S3 logs via the X-Forwarded-For header.
NFS Floating IP failover now contains a timeout as a mitigating factor during prolonged outages.
Limit increases for drives (20k), processes (20k), and NUMA nodes (20).
Added Rocky Linux 9.3 to supported operating systems on backends.
4.2.10
For both files and directories, relatime now produces reliable atime updates on large clusters. Previously, some conditions caused atime values to revert.
Customers can now transition between custom and predefined S3 bucket policies with one command.
4.2.9
Path optimization for S3 requests.
Custom bucket policies now accept only valid JSON through REST API.
4.2.8
Active-active network port usage in HA configurations using RDMA.
Better handling for cgroups with small memory footprints.
More graceful client upgrades with clusters that have been scaled in, preventing communication with backends that no longer exist.
STS session duration option available for IAM AssumeRole use with Amazon S3.
4.2.7
Initial support for the namespace feature in Hashicorp Vault.
Enhanced performance serving S3 GET requests using byte-range fetching.
Faster cleanup of failed multi-part upload parts.
Added support for Ubuntu 22.04.3 point release with 6.2-based kernel.
SMB-W now supports creating local mappings for AD users and groups using the rid ID-mapping, alongside the existing RFC2307 support.
Added support for using AssumeRole with Amazon IAM STS tokens when using Amazon S3.
Added support for floating IPs in AWS when using NFS.
4.2.6
Added IMDSv2 to the supported Amazon EC2 instances.
Certified Broadcom BCM957508-P2100G Dual-Port 100 Gb/s QSFP56 as supported.
Added support for IOMMU on WEKA backend servers with Mellanox NICs.
Certified OFED 23.10-0.5.5.0.
Added support for RHEL/Rocky Linux 9.2 operating system.
Updated the default provider of SMB services to SMB-W, replacing the legacy SMB.
4.2.5
Alerts are in place for the use of duplicate IP addresses.
Reduced likelihood of receiving false alerts for full quotas.
Better error handling for aborted S3 MultiPartUpload requests.
The weka cluster failure-domain CLI command is enhanced with more component fields.
No more umount/mount cycling for SCMC clients during upgrades.
4.2.4
Certified CX7 (Infiniband) as supported.
Non-disruptive upgrades of clients in more configurations.
When using NFS, the df command now reports WEKA quotas as expected.
ILM Policy deploys now use less memory when scaled to a billion objects.
Deleting empty buckets via S3 API works as expected.
To upgrade to 4.2.4, the source version must be 4.1.2.
If the S3 protocol is configured, contact Customer Success to confirm that ETCD (internal key-value store) has been upgraded to KWAS.
4.2.3
Added operating systems supported on clients: RHEL/Rocky Linux 8.8 and SuSe 15 SP4.
Added support for Vault 1.14 (certified from WEKA release 4.2.1).
NFS client connections are steady during floating address migration.
The weka local upgrade command now supports servers without frontend containers.
The weka status command output newly reflects the unavailable capacity.
The Synchronized Snapshot feature (the ability to perform incremental snapshots downloaded from an object store) is temporarily disabled. This feature will be reinstated upon subsequent release.
4.2.2
N/A
4.2.1
IAM support on GCP: You can access Google Cloud Storage using a service account with the required permissions granted by the IAM role.
NDU improvement: The non-disruptive upgrade process is now improved by upgrading the compute containers one at a time (rolling upgrade) while the remaining containers continue serving the clients.
E810 NIC support on MCB: The Intel E810 NIC is now supported in the multi-container backends (MCB) architecture.
Added support for Rocky 9.0, 9.1, and Ubuntu 22.04 on backends.
4.2.0
Mount filesystems from multiple clusters on a single client: You can mount filesystems from up to seven clusters in parallel on a single WEKA client for enhanced performance and use cases.
Snapshots improvements:
Quickly download a previously taken snapshot to another cluster. Following that operation, metadata is auto-prefetched.
Allow IO operations continuity to the filesystem while restoring a snapshot (using a preserved snapshot name).
Added abort/pause snapshot download functions.
WEKA CSI Plugin enhancements:
Can control WEKA mount options through the storage class.
Snapshots and volume cloning.
Added support for k8s fsGroups.
Increased organizations support: Increased the maximum number of supported organizations to 256 per cluster.
New GUI improvements: The GUI is improved with new features and operations, such as the insights page with top processes usage, drives load, and latency.
Azure cloud enhancement:
Improved performance using DPDK networking and higher MTU.
Added support for auto-scaling.
Interoperability updates:
Added support for Mellanox OFED version 5.9-0.5.6.0.
Added support RHEL/Rocky Linux 9.1/9.0 and Ubuntu 22.04 for clients only (only use applications set with 2MB hugepages).
HashiCorp Vault: Added support for HashiCorp Vault up to version 1.13.
Breaking changes and deprecations:
Removed the auth_token mount option.
Single protocol type per server support: A single WEKA server can now only have one protocol server (S3, SMB, or NFS). Adding an additional protocol process/server is no longer allowed. For clusters being upgraded, distribute the various (S3, SMB, or NFS) protocols across all the backend servers before running the upgrade.
Only the multi-container backend (MCB) architecture is supported: The legacy Single Container Backend (SCB) architecture, where each server in the cluster includes a single container with all the processes running on it, is deprecated in 4.1. To upgrade to 4.2, the source cluster must be in MCB architecture. Contact the WEKA Customer Success team to convert a 4.1 cluster in a legacy architecture to the MCB architecture.
ETCD replacement for S3 protocol improvement: The ETCD component, which stores the IAM format, policies, service accounts, users, STS, and policy mappings, is replaced by a more robust mechanism. When upgrading a cluster running with ETCD from V4.1, the cluster continues to run with ETCD in V4.2, and an alert is raised to migrate to the new mechanism. Contact the WEKA Customer Success team to perform this update.
