HI there. We were testing a migration to a new server (Rackspace to WP Engine) and in the process my IT guy toggled development mode on and off. We switched our traffic to the new server, had some errors, and decided to switch the traffic back to the original server until we could resolve the problems. Upon that, we saw the site at the old server for perhaps 15 minutes, then we get a timed out error page . It has now been a couple hours. My IT guy says this is due to the toggling of the Dev mode and it will resolve. Is that accurate, and is there anything I can do to speed the process along?
The site appears to be up and running now. Are you still experiencing issues?
Thanks. We found that the DNS was pointing directly at the server instead of going through CF. I’m still unsure if this happened because of Development Mode or if it was human error. The site is up, but not all CF DNS are being implemented. We don;t know why this is the case, either. Does CF need to time to propagate once it has been turned around and then back again?
I’m going to guess at the root problem, please don’t beat me senseless if I’m wrong.
For records which are in the control panel if you do a DNS lookup you’ll see a Cloudflare IP address, not the origin IP address. And in the case of some records (primarily in the CNAME use case) if you use a global query tool to see if an CNAME record has propagated it will say it hasn’t. The reason for that is behind the scenes we’re transforming that to act like an A record instead so the web check tool fails for that specific query (but that’s not how a browser would ask and so it’s really working). If you instead search for foo.example.com as an A record you’ll probably see it fully propagated with a cloudflare IP.
Right now it appears your www site is and the bare root is not. Since there is a rule somewhere (Cloudflare or origin) to redirect www to the bare root it’s effectively bypassing Cloudflare at the moment I think.