For future reference, yes there are a few ways to clear cache, and generally the first steps to showing the problem to us so we can help is to find the value of the x-nf-request-id HTTP response header, of the request for an asset you are seeing that is not up to date. This article describes that process:
While we have occasionally had bugs that lead to stale content, usually there is some other cause and/or you can self-serve on a fix:
- a redeploy will fix most, regardless of the cause (successful production deploys clear production cache)
- most that aren’t fixed by that are caused by custom
Cache-ControlHTTP headers that you (or your framework) as configured - be careful not to do that (details in this article) or by a misconfigured service worker (details in this article). - Or perhaps you did not have the deploy you intended published. This might happen because we only publish commits to your production branch at your production URL, or because you’ve configured locked deploys and we are faithfully serving an old deploy and our UI will say this (you’ll see a “published” notice on an older deploy, and there will be a button to
Start auto-publishingat the top of your deploy logs listings page).
If you see it again, check the details from #2 / #3, and if they don’t seem relevant, provide us with the detail from #1 and we’ll be happy to investigate!