ZONE BRIDGE

Traffic paths

A cache and a traffic filter do different jobs

Follow a request from the reader to the origin, and see which problems a cache can solve before you choose another service.

Optical fibres beside a network socket

Follow one ordinary request

Picture a local club publishing a timetable. Most visitors want the same small page and the same badge image. Some arrive with files already stored in their browser; others need a fresh response. Drawing those two journeys is a useful first step before discussing traffic protection. Mark where the name is resolved, where the secure connection ends and which machine supplies each response. Keep the drawing specific to the service you operate: a general diagram can hide a second hostname or an uncached download.

Understand what reuse changes

A cache can reuse a stored HTTP response when its rules allow it. A browser cache serves one client; a shared cache can serve more than one. Reusing an eligible response can reduce work at the origin, but a request that must reach the application still needs that application to function. Do not assume that a personalised page is safe to share just because it contains a cookie. MDN explains the distinction and the response directives that govern storage and reuse.

Separate an efficient page from a protected route

Optimising the timetable makes an ordinary visit cheaper to serve. It does not establish that every service address is protected against hostile traffic. Draw a second route for requests that reach the origin directly. Is that route required, deliberately restricted or simply forgotten? Changes to origin access need coordination with the host and any monitoring or management services. Avoid applying a copied firewall recipe to a live service whose dependencies you have not mapped.

Ask where filtering would sit

KoDDoS offers remote DDoS protection for services hosted elsewhere. When considering that arrangement, ask how traffic reaches the filtering service and then the existing host, which addresses and protocols are covered, and what changes are required at the origin. Keep certificate handling, client address visibility and the rollback route on the same checklist. A service description is a starting point for those questions; the appropriate setup depends on the actual site and the agreed scope.

Keep a before-and-after record

For a planned change, record one public page, one downloadable file and any operation that cannot use a shared response. Test them before the change and repeat the same checks afterwards, including from outside the office network. Keep the earlier route documented privately until the new arrangement is accepted. If a request fails, the drawing helps you say where the failure begins instead of changing DNS, application settings and caching rules all at once.

Put it into practice

  1. 01 READERWhich files are already in the browser?
  2. 02 EDGEWhich requests can be reused or filtered?
  3. 03 ORIGINWhich requests still reach the host?

For each arrow, write the hostname, its purpose and the person responsible. Keep private addresses and access details off the public diagram.

Sources & further reading

Examples and worksheets on this page are illustrative. Documentation links were checked in November 2026; follow the relevant product’s current instructions.