DEV Community

禾
禾

Posted on

Public Origin vs Restricted Origin: Where to Stop Plugin-Probing Botnets

A scraping botnet does not need an exploit in its first request. It can start by collecting plugin names, requesting version files, and testing familiar directory paths. Those discoveries help an attacker decide which sites deserve a second pass with an exploit payload.

For a security owner asking how to block scraping botnets probing for outdated plugins, the first decision is architectural: can requests reach the application without passing through its defenses?

Compare two placements: a WAF in front of an origin that remains publicly reachable, and a WAF in front of an origin restricted to approved ingress. Both can filter traffic on the protected route. Only the second design closes the intended bypass route, provided the network restrictions cover every application listener.

Choice A: Proxy the hostname, leave the origin public

SafeLine is a self-hosted WAF that filters and monitors HTTP traffic. Its documentation describes placing it in front of an application as a reverse proxy. The application setup requires a domain, listening port, and upstream address, followed by resolving the domain to the SafeLine server. See the official overview and application configuration guide.

This placement gives requests sent to the configured hostname a filtering path. However, changing DNS does not itself establish a network restriction on the upstream server.

If the origin remains accessible from the Internet, an attacker who discovers its address may send requests directly to it, including the expected hostname. Plugin enumeration and exploitation attempts can then reach a route that bypasses the WAF.

The evidence can be misleading: the WAF records blocked probes while application access logs continue to show similar requests arriving through another path. A successful block demonstrates protection for that request, not complete coverage of the application.

Choice B: Proxy the hostname and restrict origin access

The stronger placement uses SafeLine as the public application entry point and separately restricts origin ingress through the infrastructure's firewall or network controls. This restriction is an architectural recommendation, not a feature automatically enabled by SafeLine's documented application setup.

Before applying it, inventory legitimate upstream callers. Monitoring services, deployment checks, internal applications, and other proxies may need access. Permit the required paths deliberately, then test from an external network that unauthorized direct access fails.

Include IPv4, IPv6, alternate ports, and additional hostnames in that review. An overlooked application listener can preserve the same bypass that the primary firewall rule was meant to remove.

Choose this placement when you control origin networking. If you cannot restrict ingress, document the remaining exposure and prioritize removing vulnerable plugins rather than assuming the proxy covers every request.

Turn probe evidence into concrete decisions

Treat the investigation as a sequence of checks:

  1. Identify the requested resource. Separate ordinary page scraping from requests for plugin metadata, installation scripts, administrative endpoints, or executable files. A recognizable path is evidence of interest, not proof that a vulnerable plugin exists.
  2. Verify the installed software. Check the actual plugin inventory and versions against the maintainers' security guidance. Remove unused components and update affected ones. An old-looking public asset is a lead to investigate, not a reliable inventory.
  3. Compare traffic paths. Correlate WAF and origin records by time, method, path, and response. Investigate requests that reach the application without corresponding evidence on the expected ingress path, accounting for differences in logging.
  4. Look for escalation. A transition from metadata requests to suspicious POST requests, traversal strings, or executable-file access warrants closer review. Unexpected files or process activity should trigger incident investigation; blocking later requests does not establish that earlier attempts failed.

SafeLine's application guide documents an attack list and a detail view for inspecting detected attacks. Use those records alongside application and host evidence rather than treating any single log as the entire incident history. Source: application configuration guide

Configure controls for the traffic you actually see

SafeLine documents IP-based rate limiting and CAPTCHA challenges among its capabilities. Source: official overview

An IP-based threshold can help with repeated requests from one address, but distributed botnets can spread requests across many addresses. Also check whether legitimate visitors share an address before tightening limits.

A browser challenge requires a different decision: whether the protected workflow expects a human using a browser. Test legitimate automation, API clients, and accessibility-sensitive journeys before introducing challenges. The supplied documentation does not establish plugin-version discovery or a dedicated botnet attribution capability, so do not build the response around those assumptions.

Does a plugin probe prove that exploitation succeeded?

No. Verify whether the component exists, whether its version is affected, and whether subsequent application or host evidence indicates execution or unauthorized changes.

Is pointing DNS at SafeLine enough?

It establishes the documented proxy route. Independently test whether the origin still accepts unauthorized direct connections, including connections using the application's hostname.

Should every probing IP be blocked permanently?

Base the decision on behavior and legitimate use. Address rotation limits the value of individual blocks, while shared addresses can make broad blocking disruptive. Review whether the control reduces unwanted requests reaching the origin.

What should be checked after changing placement?

Verify normal application journeys, required integrations, and denial of unauthorized direct-origin access. Then compare request evidence across the WAF and origin. Keep patching and plugin removal in the same work plan: ingress controls reduce exposure, while software maintenance removes the vulnerable target.


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.

Top comments (0)