Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Configuring a cloud WAF means choosing managed rule sets, tuning them in count mode, adding rate limits and bot controls, and making sure the origin cannot be reached directly.

How Do You Configure a Cloud WAF?
A cloud web application firewall inspects HTTP requests at the provider’s edge and decides whether to allow, block, count, or challenge each one. AWS WAF, Azure Front Door and Application Gateway WAF, Google Cloud Armor, and Cloudflare all work on the same principle. Turning one on takes ten minutes. Configuring it so that it stops attacks without blocking customers is the part that takes a fortnight, and it is the part that determines whether the control is worth anything.
Every provider ships managed rule groups covering the common injection, traversal, and scripting patterns, usually derived from or aligned with the OWASP Core Rule Set. Enable the core set plus the groups relevant to your stack—SQL database, operating system command, PHP or Java, known bad inputs—and set all of them to count rather than block.
Then wait. A week of production traffic in count mode tells you which rules would have fired against legitimate requests. Rich text editors, file uploads, base64 payloads, and any API that accepts JSON containing SQL-like strings are reliable generators of false positives. Review the sampled requests, write exclusions scoped to the specific path, parameter, or header rather than disabling the rule globally, and only then switch to block.
Rule sets that use paranoia levels, such as the OWASP Core Rule Set, let you tighten gradually. Level 1 is the sensible production default; higher levels catch more and block more of your own users, so raise them only for specific high-value paths.
A web application firewall at the edge is bypassed entirely by a request sent straight to the origin address. Attackers find that address through stale DNS records, certificate transparency logs, mail headers, and forgotten staging hosts. Restrict inbound traffic at the origin to the provider’s published ranges, require a secret header injected by the edge, or remove the public address altogether using private connectivity. Verify it by trying to reach the origin yourself from an unrelated network.
Send full WAF logs, including sampled request detail, to your monitoring platform rather than leaving them in the console. Useful detections include a sharp rise in blocks from a single source, a single source triggering many different rules in sequence, which looks like scanning, and blocks on paths that should never receive traffic. Review the top blocked and top counted rules monthly; the pattern tells you both what is being attempted and where your tuning is drifting.
Build the configuration as code alongside the rest of your infrastructure as code, with review on changes. Exclusions added during an incident at 2 am have a way of becoming permanent, and a version-controlled rule set makes that visible.
It is a filter in front of an application, not a substitute for secure code. It cannot see business logic abuse, it rarely helps with broken access control, where the request is perfectly well formed and simply asks for someone else’s record, and determined attackers routinely find encodings that evade signature rules. Treat it as a control that raises the cost of mass exploitation and buys time for patching, and keep fixing the application behind it.
At Privia Security, we test applications and the filtering in front of them through our web application security services.
No. Deploy managed rules in count or detection mode for at least a week of representative traffic, review what would have been blocked, and write scoped exclusions for the legitimate requests that matched. Switching straight to blocking on a live application reliably breaks file uploads, rich text fields, and APIs that carry structured payloads, and the resulting rollback damages trust in the control.
They group rules by how aggressive they are. Level 1 catches clear attack patterns with few false positives and suits most production sites. Higher levels add rules that flag increasingly ambiguous input, catching more genuine attacks but also more legitimate traffic. Raise the level selectively for sensitive paths rather than globally, and expect substantial tuning work at level 3 and above.
Yes, in two ways. Attackers craft encodings, fragmentations, or parameter pollution that the rule set does not match, which is why signature rules degrade over time. More simply, they send requests directly to the origin, skipping the edge entirely. Origin restriction is the higher-value fix, because it closes a complete bypass rather than one evasion technique.
PCI DSS requires public-facing web applications to be protected either by an automated technical solution that detects and prevents web-based attacks or by regular application vulnerability assessment, and from version 4.0 the automated solution became the expected approach. A tuned WAF with logging and review satisfies this. An untuned one in count mode does not, so keep records of configuration and tuning.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.