Using Cloudflare Always Online for limited outage fallback

Confirm the relevant website traffic is proxied through Cloudflare. Review whether archiving public pages is appropriate for your site, because the current feature integrates with the Internet Archive and may share popular public URLs for crawling.

1. Open the domain in Cloudflare and locate the caching Configuration area containing Always Online.
2. Read the feature’s current behaviour and enable it only for a site where public-page archiving is acceptable.
3. Keep ordinary backups, monitoring and origin recovery in place. The fallback depends on available cached or archived content and does not cover every page or every origin error.
4. During an actual incident, check which pages visitors can access and make outage communications accurate. Do not claim the whole application is operational because a cached landing page loads.
5. After the origin recovers, verify current content, forms and account functions. Disable the feature if its archive behaviour no longer fits the website.

The setting should be saved for the intended zone, and your recovery plan should distinguish read-only public fallback from normal dynamic service. Avoid deliberately interrupting production solely to test the feature.

If a page is unavailable during an outage, it may not have usable cached or archived content. Restore origin service rather than repeatedly purging the cache. Never treat Always Online as protection for private records or as a replacement for a tested database and file backup.

• Checking whether Cloudflare is active for a hostname
• Creating a backup plan you can actually restore
• Moving DNS away from Cloudflare and removing a zone
• Checking why cookies prevent or alter Cloudflare caching

• Cloudflare documentation