Understanding redirect loops: what causes them and how to fix them
A redirect loop is one of the more stubborn errors the modern web throws at visitors. The page never finishes loading, the address bar keeps flickering between two URLs, and the browser eventually gives up with a message like "too many redirects" or "the page isn't redirecting properly". Tapping refresh does nothing, opening the link in a new tab does nothing, and rebooting the router only delays the inevitable. The destination server is caught in a cycle it cannot resolve, and your job is to figure out whether the fault lies on your end, somewhere in the middle, or at the website itself.
The complication in Australia is the length of the middle. NBN traffic can hop through a fibre-to-the-premises segment in a Brisbane suburb, a fixed wireless tower outside Wagga Wagga, or a satellite link covering remote parts of the Northern Territory, and each hop adds a chance for a transparent HTTP to HTTPS rewrite or a caching layer to interfere.
What a redirect loop actually looks like
Technically, a redirect is just an instruction. A web server replies with a status code in the 3xx family — usually 301 for a permanent move, 302 or 307 for a temporary one — and includes a Location header that tells the browser where to look next. Under normal conditions the browser follows the chain, hits the last entry, and renders a page. A loop forms when at least two URLs keep handing the request back to each other. A common example is a site that sends visitors from the bare domain to the www subdomain, which then sends them right back, round and round, until the browser hits its built-in cap and aborts.
Most browsers impose a sanity limit of around twenty hops, designed to prevent a runaway server from locking up clients forever. Chromium, Firefox, and Safari each enforce this guard, which is why the same error appears whether you are at a desktop in Adelaide or on a phone moving between Flinders Street and Southern Cross. During the failed loading process, the address bar flashes between two values, and reading those values carefully is the first clue about which two endpoints are stuck.
Why redirect loops happen
Most loops begin with a server-side configuration error rather than a deliberate design choice. The textbook cause is the migration from HTTP to HTTPS, where the new secure version forwards everyone to a www or non-www host, while a leftover rule in the original config still points back to the secure URL. Once both rules are live they reference each other endlessly. Content management systems such as WordPress, Shopify, and Magento expose their own redirect managers, and a plugin update can quietly add new entries while old ones stay in place.
Content delivery networks tend to amplify the problem rather than calm it down. Sites behind Cloudflare, Fastly, or Akamai hit the edge first, and any HTTPS rule set on the origin creates a second chance for the misconfiguration to fire. Australian e-commerce brands running on Shopify occasionally see loops after enabling HSTS preload, only to discover that an old subdomain rule was never deleted from the storefront settings. Rebrands from a .com.au address to a .com can also leave stale DNS records pointing at the wrong bucket.
A subtler source of loops is the affiliate or analytics layer. Affiliate links often pass through two or three tracking hops before reaching a destination, and any hop that returns a relative URL instead of an absolute one can land the request back at an earlier stage. Sites that run aggressive A/B testing sometimes see their flag fall into the chain, sending requests between variation URLs that share cached assets in a way the server did not expect.
How Australian networks make the problem harder to spot
A loop that should take a few seconds to identify often looks like an outright outage in regional areas. In towns that rely on fixed wireless or satellite NBN services, the standard reflex is to check the modem lights, reboot the router, or call the ISP, and that reflex is reasonable. Telstra, Optus, TPG, and Aussie Broadband each handle their own HTTP to HTTPS rewrites differently, and those rewrites can extend a redirect chain in a way the origin server never intended. The same URL can produce different chains on different ISPs, which complicates reporting back to the website owner.
Habits that look sensible tend to fall short with this particular fault. Tapping refresh a few times, opening the same URL in a second tab, or switching browsers all funnel the request through the same broken path. Reading the address bar carefully during a refresh usually gives the first real clue: the URL flickers between two values without ever settling, which is the visual signature of a closed loop.
Cultural habits around sharing links also feed the cycle. Australians frequently paste URLs into group chats on WhatsApp or Signal, and chat apps sometimes append their own preview redirect on top of any redirect the website itself uses. Two layers of redirection can look identical to a single-loop error to a user who cannot see either layer. Copying the link from the original source, rather than from the chat preview, is the safest move.
Quick checks you can run from your browser
Before assuming the website is down, run a short sequence of browser-side fixes that resolve a surprising share of redirect issues without touching any server settings. The whole routine takes under a minute and is worth trying before calling the ISP or chasing the site owner.
Browser checks worth trying:
- Open the page in a private or incognito window so stored cookies and the shared cache are bypassed.
- Clear the cache for that specific site only, or for the whole browser if the private window still fails.
- Test the same URL on a phone using mobile data, so a different network stack handles the request.
- Disable browser extensions temporarily, especially privacy tools that rewrite requests on the fly.
If the URL loads correctly on mobile data but stalls on home Wi-Fi, the ISP or local router is inserting the extra hop and the answer lies with the network. If it fails on every network, the loop lives on the server side and the next section is more useful.
Fixing the loop from the server side
Site owners reading this from a Sydney or Melbourne office can take a more direct approach by editing the redirect rules that started the cycle. The aim is to leave only one canonical path between the origin and the final destination, and to remove any rule that points back to a previous hop. Running an audit after every CMS update is wise, because plugins and themes tend to leave behind redirects they installed during setup.
Server fixes worth running:
- Audit the .htaccess, nginx.conf, or web.config file for duplicate RewriteRules that bounce between www and non-www hosts.
- Check the CMS redirect manager for rules imported from an older SEO plugin that were never deleted.
- Review the CDN configuration, especially the Always Use HTTPS toggle, which often clashes with the origin's own redirects.
- Flush the CDN cache and the application cache after each fix so stale responses do not replay the old loop.
Once changes are live, run curl -I from a workstation, or paste the address into a free online header inspector. A clean final response should return a 200 OK after at most two redirects. Anything longer, especially a chain that ends where it began, tells you the loop is still alive. For a worked example of a domain that tidied up its redirect setup, the classic slots walkthrough on the same hosting account shows the kind of article-level configuration that survives these audits.
When nothing on your end moves the needle, the simplest next step is to wait a few minutes and try again, since maintenance windows and mid-deploy glitches often resolve themselves. For sites that remain stuck, the cyberschooldps holding page gives a clear checklist of the usual suspects — from network interruptions and DNS filtering to caching layers — and helps visitors understand whether the fault sits with their line, their device, or the destination itself. The same checklist can also point server administrators toward the audits covered above.