01Customer edge / prefixesBGP, protected IPs or inbound handoff
→
02Peeryx mitigation fabricAnalysis, signatures, filtering and upstream relief when required
→
03Peeryx delivery layerCross-connect, GRE, IPIP, VXLAN or router VM
↓
04Customer productionDedicated server, cluster, proxy, backbone or custom logic
The upstream layer protects ports and transit while customer logic handles the rest.
Why some buyers want Peeryx only for the first volumetric layer while keeping their own filtering stack behind it.
This model fits XDP, eBPF, DPDK or application-specific filtering.
Why some buyers want Peeryx only for the first volumetric layer while keeping their own filtering stack behind it.
Commercially, it is often more realistic than pretending one generic filter solves everything.
Why some buyers want Peeryx only for the first volumetric layer while keeping their own filtering stack behind it.
This article explains Building a filtering stack behind volumetric protection in practical terms for teams that need a serious Anti-DDoS model.
The goal is not only to absorb attack volume, but also to preserve legitimate traffic, keep handoff readable and avoid unnecessary architectural mistakes.
Why this matters
Building a filtering stack behind volumetric protection matters because the wrong first layer can saturate links, damage user experience or hide the real operational problem.
A better design starts with visibility, upstream relief where needed and a clean return path for useful traffic.
The upstream layer protects ports and transit while customer logic handles the rest.
This model fits XDP, eBPF, DPDK or application-specific filtering.
Commercially, it is often more realistic than pretending one generic filter solves everything.
Where classic setups fail
Classic setups often fail when they rely on generic blocking, unclear routing or a model that only speaks about raw capacity.
What serious buyers need is a model that explains where traffic enters, where mitigation happens and how clean traffic comes back.
How to design the right model
A credible approach combines upstream volumetric mitigation, a handoff model matched to topology and customer-operated logic where it adds value.
That is why pages about protected transit, router VM, dedicated servers and specialised gaming delivery all matter on the same site.
1
Where will saturation happen first: transit, link, stateful firewall or local server?
2
How will clean traffic be returned: BGP, GRE, VXLAN, cross-connect or an intermediate VM?
3
Which filtering logic stays upstream and which logic remains under customer control?
4
How will latency, observability and operational changes be handled during mitigation?
Questions to ask before choosing a provider
Where will saturation happen first: transit, link, stateful firewall or local server?
How will clean traffic be returned: BGP, GRE, VXLAN, cross-connect or an intermediate VM?
Which filtering logic stays upstream and which logic remains under customer control?
How will latency, observability and operational changes be handled during mitigation?
FAQ
Does this topic only matter during very large attacks?
No. The design choices discussed here also affect smaller incidents, operational cost and the quality of legitimate traffic during normal periods.
Can one generic product solve everything?
Usually not. The cleanest result comes from matching the first protective layer, the handoff model and any customer-owned downstream logic.
Conclusion
Building a filtering stack behind volumetric protection should be understood as part of a broader Anti-DDoS architecture, not as an isolated checkbox.
The strongest commercial position is a realistic one: stop upstream risk, return cleaner traffic and let the design fit the customer instead of forcing a generic model.
Resources
Related reading
To go deeper, here are other useful pages and articles.
