Ultra-Low Latency Linux SRE for Trading Infrastructure

QuantFlow; Amsterdam; Sur place; EUR 99 000 - 121 000; Plein temps

Résumé du poste

QuantFlow is seeking an SRE/ Linux Performance Engineer to own its bare-metal trading infrastructure in Amsterdam. You will tune BIOS/firmware, kernel parameters and application performance across the stack to achieve nanosecond precision in time sync and ultra-low latency networking.

You will build and maintain CI/CD pipelines, profile latency at all layers, and optimize the network stack, including DPDK or Onload, while collaborating with quantitative developers and traders.

Qualifications

• 4+ years in a Linux infrastructure, SRE or performance engineering role; HFT or electronic trading experience preferred.
• Deep Linux systems knowledge: kernel tuning (sysctl, cgroups, isolcpus, tickless kernels), scheduling, memory management, storage I/O.
• Hands-on kernel bypass networking: Solarflare/Xilinx Onload, ef_vi, DPDK or similar.
• Solid understanding of PTP (IEEE 1588), PPS and time synchronisation architecture.
• Strong Bash and Python, with Ansible, Salt or similar, and CI/CD (Jenkins, GitLab CI) in bare metal/on-prem setting.
• Comfort with hardware: server and NIC selection, NIC firmware, BIOS tuning and colocation.
• Profiling tools such as perf, ftrace, flamegraphs, eBPF/bpftrace and cyclictest.
• Plus: FPGA networking or hardware timestamping, ITCH/OUCH/FIX, InfiniBand or RoCE.

Responsabilités

• Own and optimise bare metal Linux infrastructure end to end: BIOS and firmware settings, kernel tuning and application level performance.
• Implement and maintain precision time synchronisation across the trading environment with nanosecond accuracy.
• Configure and tune kernel bypass networking for ultra low latency market data and order paths.
• Profile and remove latency at every layer: hardware interrupts, CPU isolation and pinning, NUMA topology, memory allocation and network stack.
• Build CI/CD pipelines and infrastructure as code for repeatable, auditable deployments across physical hosts.
• Build monitoring focused on latency distribution, jitter and tail latency, not just uptime.
• Manage network infrastructure including switch configuration, multicast and colocation connectivity.

Connaissances

Linux performance

SRE

Kernel tuning

Kernel bypass networking

PTP time sync

CI/CD

Bash

Python

Ansible/Salt

CI/CD pipelines

Onprem hardware

Outils

Solarflare/Onload

ef_vi

DPDK

Perf/ftrace

eBPF/bpftrace

Description du poste

QuantFlow is seeking an SRE/ Linux Performance Engineer to own its bare-metal trading infrastructure in Amsterdam. You will tune BIOS/firmware, kernel parameters and application performance across the stack to achieve nanosecond precision in time sync and ultra-low latency networking.

You will build and maintain CI/CD pipelines, profile latency at all layers, and optimize the network stack, including DPDK or Onload, while collaborating with quantitative developers and traders.