Lorixen.
EN

Redirect chains and loops: finding and fixing them

Short answer: A redirect chain is when one old URL redirects to another URL that redirects again before reaching the real page; a loop is when the redirects never reach a page at all. Chains slow visitors down and, past a point, stop search engines following them: Google's crawlers follow up to 10 hops by default. Loops break the page completely. The fix for both is the same principle: every old URL should redirect once, directly to its final address, and only one system (server, CDN or plugin) should be in charge of each rule.

Redirects pile up quietly. A site moves from http to https, then from non-www to www, then gets a new URL structure in a redesign, then moves to a new domain. Each change was handled correctly at the time, but the rules stack. Years later an old link might bounce through four addresses before landing. This guide works from what you see, to why it happens, to how to fix it.

How to see what's actually happening

Before fixing anything, trace a few URLs. Use any online redirect checker, a browser extension that shows redirect paths, or the command line (curl -I -L shows each hop). Test:

  • the bare domain with http, e.g. http://example.com.au;
  • the same with www, and with https;
  • two or three old page URLs from before your last redesign;
  • an old URL with and without a trailing slash.

Write down each hop and its status code. Google's documentation explains that 301 and 308 mean permanently moved, and that Google uses a permanent redirect as a signal that the target should be the canonical URL. 302 and 307 mean temporary.

What you see: http://example.com.au/services.html → https://example.com.au/services.html → https://www.example.com.au/services.html → https://www.example.com.au/services/.

Cause: separate rules added at different times, each handling one change: protocol, host, then URL structure.

Fix: rewrite the rules so each old URL goes directly to its final address in one hop. In practice, that means:

  1. Put the specific page mappings (old path → new path) first, pointing straight at the final https, www (or non-www) address with the correct trailing slash.
  2. Then a general rule that sends anything else to the right protocol and host in one step, instead of two separate rules.

Google's crawlers follow up to 10 redirect hops by default, according to its HTTP status code documentation, and the content of redirecting URLs is ignored in favour of the final target. Even well under that limit, every hop adds a delay for visitors on mobile connections. Note too that the same documentation says Google's inspection tools don't follow redirects, so test the final URL directly when using them.

Symptom 2: "Too many redirects" or "This page isn't working"

What you see: the browser gives up with an error. MDN describes a redirection loop as additional redirects following one already followed so that no page is ever found. Servers that detect it may return a 500 error; where the loop spans several systems, the browser detects it and shows its own error message.

Common causes:

  • HTTPS fighting between layers. The CDN or proxy talks to your server over http, the server redirects to https, the CDN passes the request again over http, and so on. Common after enabling a CDN's "flexible" encryption mode with a server that also forces https.
  • www rule conflicts. The host's control panel forces www while a plugin or .htaccess rule forces non-www.
  • CMS address mismatch. WordPress's Site Address set to one version while the server redirects to the other.
  • Trailing slash disagreement. One rule adds a slash and another removes it.
  • A redirect plugin rule pointing to itself, often a pattern rule that matches its own target.

Fix: find the two systems disagreeing and make one of them the owner. Pick one canonical host and protocol, set it in the CMS, and enforce it in exactly one place (server or CDN), then remove the duplicate rule elsewhere. Clear caches in between; browsers cache permanent redirects, so test in a private window.

MDN notes that loops often appear soon after a server configuration change. If the error started right after you touched settings, that change is the first suspect.

Symptom 3: Redirects go to the homepage instead of the right page

What you see: every old URL lands on the homepage.

Cause: a catch-all rule used instead of page-by-page mapping during a migration.

Fix: build a mapping of old URLs to their closest new equivalents, especially pages that had traffic or links. Use the catch-all only for genuinely leftover URLs, and consider letting truly gone pages return 404 rather than forcing them home.

Symptom 4: Some redirects are temporary that should be permanent

What you see: 302 or 307 codes on moves that are final.

Cause: default settings in a plugin or host panel.

Fix: switch to 301 or 308 for moves you won't reverse. Google recommends permanent server-side redirects whenever possible when a URL change should show in search results, and lists meta refresh and JavaScript redirects as alternatives when server-side isn't possible.

What you see: your own menus and content trigger redirects.

Cause: links weren't updated after the move; the redirects hide the problem.

Fix: update menus, buttons and in-content links to the final URLs, and regenerate your sitemap so it lists only final addresses. Redirects are for outside links and bookmarks; your own site shouldn't depend on them.

A clean-up order that works

  1. Decide the canonical form: https, www or not, trailing slash or not.
  2. List every place rules live: host panel, .htaccess or server config, CDN, CMS settings, redirect plugins.
  3. Export existing rules and remove duplicates.
  4. Rebuild: specific page mappings first, then one general host/protocol rule.
  5. Test your list of sample URLs again; each should be one hop or none.
  6. Update internal links and the sitemap.
  7. Recheck in a few weeks; old rules have a habit of being re-added by plugins.

FAQ

How many redirects is too many?

Aim for one. Google's crawlers follow up to 10 hops by default, but every hop slows visitors and adds room for errors.

Do chains lose SEO value?

Google treats permanent redirects as a signal about the canonical URL and processes the final target's content. Short, direct redirects are simply more reliable and faster.

Should I delete old redirects after a while?

Keep redirects for URLs that still get visits or have external links. Remove only those you're sure nobody uses, and update internal links rather than relying on redirects.

Why does it work for me but not for others?

Browsers cache permanent redirects. Test in a private window or another device after changing rules.

If old links bounce through several addresses or end in an error, send us a few example URLs and we'll trace the hops and suggest a single set of rules.

Sources

WhatsAppLet’s talk