Overview
Beeks Analytics as an Appliance is built on Dell PowerEdge servers, offering a range of CPU, storage, and expansion options to meet your specific requirements.
For high-performance financial market data processing, appliances can be configured with FPGA capture and compression cards. The system supports a distributed architecture, with separate dedicated appliances recommended for maximum scalability. Beeks Analytics is compatible with multiple CPU architectures, with Rocky Linux as the preferred operating system. The architecture is modular and scalable, allowing components to run on separate or combined appliances depending on your needs.
Features & Benefits
Distributed Architecture
- Supports a distributed setup with separate dedicated appliances for VMX-Analysis and VMX-Capture.
- Can receive data from multiple VMX-Capture appliances across different locations.
- Both VMX-Capture and VMX-Analysis can run concurrently on the same appliance with careful resource management.
Server Chassis Variants
- Appliances are typically built on Dell PowerEdge servers.
- Available in 1U and 2U chassis with single or dual CPU slots.
- Support for PERC hardware NVMe RAID, with RAID configurations recommended for performance and resiliency.
CPU Architecture
- Beeks Analytics utilises AMD EPYC4 CPUs, with continued support for EPYC2, EPYC3, and Intel Xeon Scalable CPUs (2nd, 3rd, 4th Generation).
- Appliances are specified in sizes from S – XXL, starting at 16 cores/64GB RAM.
- Best cost: clock-speed CPUs selected for each variant, with scalability options.
Capture Cards
- Napatech cards are mandatory for any VMX-Capture installation, providing
- High fidelity capture with deep buffers.
- Ability to handle PTP and/or PPS time signals.
- Capability to separate traffic into distinct streams, offloading work from the server CPU and improving scalability.
- Card Options
- NT200A02
- Fully-featured with PTP hardware timestamping and 2 QSFP28 ports.
- NT50A01
- Lower cost, lacks hardware time sync, uses OS clock sync, and has 2 SFP28 ports.
- NT200A02
- Standard setup uses a single Napatech card, but up to 3 NT200A02 cards can be used in 2U chassis.
Compression Cards
- Optional for VMX-Capture deployments.
- Compress data on-the-fly before writing to disk, enabling handling of higher data volumes.
- Offload packet capture decompression tasks to the FPGA, speeding up tasks and reducing server CPU load.
- Supported Cards
- AMD Alveo U50
- Bitware Agilex IA-440i (standard option)
- Multiple cards can be used in an appliance and scale linearly based on throughput needs.
- Combined capture + analysis appliances usually have one compression card, with larger setups or dedicated 2U capture boxes recommended to have at least 2 compression cards.
Storage Variants
- Beeks Analytics turnkey appliances support various storage configurations with RAID controllers.
- Dedicated RAID controllers recommended for VMX-Capture, VMX-Analysis, and mdPlay functions.
- NVMe drives offer high performance, with RAID5 recommended for large capture volumes.
Operating System
- Supported OS’s: RHEL8, RHEL9, and their derivatives.
- Rocky Linux 9 from version 12 onwards is preferred for 25% better performance.
- Easy setup with minimal install; the installer configures repositories and system settings automatically.
Why Choose Beeks Analytics?
Beeks Analytics is designed with three core principles in mind
Open Architecture
We integrate easily with your existing systems, giving you full ownership of your data. Beeks Analytics can also run on your hardware, offering flexibility and control.
Open Scaling
As your infrastructure grows or load fluctuates, Beeks Analytics scales up or down seamlessly with commercially available server hardware advancements.
Open Consumption
Our modular design ensures you only pay for what you use. Whether you need packet capture for network performance monitoring or more advanced analytics, our transparent pricing reflects actual usage, with cost drivers like core count clearly outlined.
Public Documentation
As part of our commitment to transparency, most of our Beeks Analytics documentation is publicly accessible at Beeks Technical Documentation.