DPDK Software Engineer

Hi,

Hope you are doing well !

Please find the below urgent role.

Position: DPDK Software Engineer

Location : Remote

Type of Hire :– Long Term Contract

Role Description:

Focus: Build, optimize, and containerize a highly efficient Virtual Service Gateway (vSG) running at the edge, supporting UBB metering, QoE analytics, and DPI classification.

Role overview:

• We are looking for a hands-on contractor to help design, implement, and production-harden a containerized, high-performance vSG that runs at the network edge.
• The solution must process traffic at high scale with strict performance and correctness requirements, while also meeting operational needs such as observability, health management, safe upgrades, and repeatable deployment.
• This role is heavily focused on the packet/flow data plane (C++/DPDK) plus practical containerization and Linux performance tuning to reliably run at the edge.

What you will do:

• Build and optimize vSG data plane (C++/DPDK)
• Implement high-throughput packet processing pipelines in C++ using DPDK (PMD model, multi-queue RX/TX, mempools, rings, batching).
• Design multi-core scaling strategies: queue-to-core mapping, flow hashing, worker pipelines, and backpressure.
• Optimize for pps/Gbps per core, tail latency, cache locality, lock avoidance, and NUMA correctness.
• Implement core service functions (fast path + correctness)
• UBB metering: accurate per-subscriber and per-service byte accounting (up/down), rollups, counter lifecycle, and reconciliation-friendly outputs.
• QoE analytics: flow/service KPIs and telemetry outputs (loss/jitter/RTT/throughput indicators where applicable), scalable aggregation, and safe handling of high-cardinality.
• DPI integration: classification pipeline integration, metadata tagging, signature update lifecycle hooks, and performance isolation/guardrails (timeouts, sampling, degradation modes).
• Containerize and operationalize for edge deployment
• Package the vSG as a containerized service with clean runtime configuration (env/config files), explicit resource requirements, and predictable startup/shutdown.
• Work with platform teams on Kubernetes/edge orchestration readiness:
• CPU pinning and isolation, huge pages, SR-IOV/VFIO device access where applicable
• liveness/readiness probes, graceful termination, restart behavior
• Build health, telemetry, and diagnostics:
• Clear counters and drop reasons, pipeline health metrics, internal watchdogs
• Debug tooling for production triage (safe packet/flow tracing, snapshots, crash dumps)
• Performance validation and regression prevention
• Create repeatable performance benchmarks and test harnesses (traffic patterns, sizing assumptions, soak tests).
• Establish performance and correctness regression checks and document host/kernel tuning prerequisites.

Required skills and experience:

• C++: 5+ years building production-grade systems software.
• DPDK: 3+ years implementing and tuning user-space packet processing pipelines.
• Strong Linux performance experience:
• NUMA, huge pages, CPU affinity/isolation, memory allocation behavior
• perf/flame graphs, gdb, core dumps, troubleshooting under load
• Strong networking fundamentals:
• L2/L3 basics, MTU/fragmentation, TCP/UDP behavior, packet timing effects
• Practical containerization experience:
• Docker/OCI images, Linux capabilities, host networking considerations, resource constraints
• Experience deploying high-performance workloads in containers (preferred: Kubernetes)

Nice to have:

• Prior work on any of: UBB/accounting, QoE telemetry, DPI engines/integration, high-scale metering systems.
• Experience with edge constraints: limited CPU, strict latency, noisy neighbors, upgrade windows.
• SR-IOV/VFIO, vhost-user/virtio, multi-NIC/multi-queue tuning.
• Observability pipelines (Prometheus/Open Telemetry) and metric design for high throughput systems.
• Experience building safe degradation strategies (sampling, feature toggles, overload behavior).

Deliverables / outcomes (what we expect):

A containerized vSG service that:

• Meets target throughput/latency and scales predictably with cores/queues
• Produces accurate UBB accounting outputs and stable QoE/DPI telemetry
• Is production operable: health checks, metrics, logs, diagnostics, safe restart/upgrade
• Performance baseline report and tuning guide:
• Host requirements (CPU pinning, huge pages, NUMA)
• Benchmark results and regression methodology
• Clear documentation for deployment and troubleshooting

Diverse Lynx LLC is an Equal Employment Opportunity employer. All qualified applicants will receive due consideration for employment without any discrimination. All applicants will be evaluated solely on the basis of their ability, competence and their proven capability to perform the functions outlined in the corresponding role. We promote and support a diverse workforce across all levels in the company.