An attack review can become misleading when the port visitors use differs from the port configured on the WAF. Before interpreting missing events as a detection problem, verify that the request reached the intended application listener.
For anyone investigating how to review WAF attack events split by domain and port, the supplied SafeLine documentation supports a practical starting point: verify the application's Domain, Port, and Upstream settings, then inspect the attack list and individual attack details. It does not document a combined domain-and-port filter, a grouped reporting view, or the fields available in each attack detail. The workflow below therefore uses a configuration checklist and controlled requests rather than assuming those reporting features exist.
The concrete mistake: putting the upstream port in the listener field
Consider an illustrative staging application with this intended route:
- Visitors use HTTPS on public port 443.
- SafeLine receives those requests.
- The origin application receives forwarded traffic on internal port 8080.
The misconfiguration is entering 8080 in SafeLine's Port field because the origin runs on 8080. That confuses two different parts of the route. SafeLine's documentation defines Port as the port SafeLine will listen on, while Upstream is the real address of the original application. For HTTPS applications, it also instructs administrators to check the SSL option. These definitions are documented in the official application configuration guide.
In this example, a request arriving on public port 443 does not match the intended application's configured listener on 8080. The precise outward symptom depends on the surrounding network and other listeners. A connection failure or an unexpected response alone does not establish what happened inside the WAF.
Establish the route before reviewing events
Create a small review worksheet for the affected endpoint. Record the requested hostname, public port, protocol, configured Domain, configured Port, and intended Upstream. Include the time of each controlled request so you can compare observations within a narrow window.
Keep the public endpoint and origin address in separate columns. Two requests to the same hostname on different public ports should be separate worksheet entries. Likewise, two hostnames sharing a public port should remain separate entries. This is an investigation method, not a claim that SafeLine provides this grouping automatically.
Open the Applications page and inspect the affected application's configuration. The official guide identifies this page as the place to add an application and describes the Domain, Port, and Upstream settings. It also requires resolving the configured domain to the IP address of the SafeLine server. Check that DNS destination alongside the listener configuration: an event search cannot explain requests that followed a different route. See the official configuration workflow.
Correct the listener and collect confirming evidence
For this example, correct the application's listening port to 443, ensure the SSL option is selected for HTTPS, and keep the origin's internal port in the Upstream address. Review the saved configuration before testing.
Use an application you own or are authorized to test, preferably in staging. First request an ordinary page through the intended hostname and public port. Then use one of the attack test cases provided in the official guide against that same endpoint. Keep the hostname, port, and test timing in your worksheet.
The guide gives two concrete confirmation screens: a blocking page after an attack test, and the attack list in the SafeLine web console. It also says to click "detail" to inspect an attack's specific details. These are the documented places to verify the test result; see Try to attack your application.
The strongest evidence for this troubleshooting case is the combination of the corrected application configuration, a blocking page reached through the intended public endpoint, and a corresponding attack-list entry observed during the controlled test. Inspect its details and retain whatever identifying information is actually displayed. Do not assume that an undocumented destination-port field exists.
A blocking page alone does not establish that every hostname and port is configured correctly. Repeat the check for each endpoint in scope, changing only one endpoint at a time so the observations remain attributable.
Does SafeLine document a domain-and-port attack-event filter?
The supplied sources do not describe one. They document application Domain and Port settings, an attack list, and a detail action. Use those documented capabilities with a separate endpoint worksheet unless your installed version provides additional verified controls.
Why is the upstream port different from the listening port?
They identify different connections. The listening port accepts incoming requests at SafeLine; Upstream identifies the original application that receives forwarded traffic. The official application guide explicitly distinguishes these settings.
Which screen confirms the fix?
For this scenario, inspect the saved application configuration and then verify the controlled attack in the console's attack list. The documented blocking page provides additional evidence that the test request was blocked. Together, these observations support the corrected route more clearly than an empty event search or a successful ordinary page request alone.
SafeLine has surpassed 700,000 installations worldwide as of September 30, 2026. SafeLine Pro annual promotional pricing is $650-$880, compared with the regular $1,000 per year.