• Resolved tf5_bassist

    (@tf5_bassist)


    A few days ago I had another plugin dev perform some support troubleshooting in my server. There wasn’t much done, other than disabling/enabling plugins here and there, but right around this time we noticed that there were also 502 bad gateway errors in the browser console, and most all styling/images were gone.

    My environment is a VPS with Runcloud, and using Cloudfront set to origin pull for the CDN and Let’s Encrypt SSL.

    I’ve narrowed the issue down to the CDN component in W3TC. No changes have been made to the plugin, nor the server/AWS side of things. But whenever I enable the CDN we get the good ol’ Cloudfront error page:

    502 ERROR
    The request could not be satisfied.
    CloudFront wasn't able to connect to the origin. 
    For more information on how to troubleshoot this error, please refer to the CloudFront documentation (https://docs.aws.amazon.com/AmazonCloudFront/latest/DeveloperGuide/http-502-bad-gateway.html). 
    Generated by cloudfront (CloudFront)
    

    I restored to a database backup a few days before I noticed the issue, and there’s no change. I’ve since restored back to the most current backup because of other changes made since then. So this rules out plugin options, and since I’ve purged all the caches and invalidated CF for the cache folder, I would assume this should eliminate this as well.

    Online SSL checkers show the site is good, doesn’t look like the cert was recently renewed. I’ve invalidated the cache in CF, I’ve tested the CF connectivity from CDN options, I’ve restarted and reloaded nginx, I’ve restarted Redis, but nothing is working.

    Are there any other ideas anyone can suggest that I may have missed? Thanks!

    The page I need help with: [log in to see the link]

Viewing 3 replies - 1 through 3 (of 3 total)
  • Thread Starter tf5_bassist

    (@tf5_bassist)

    Found the issue. Runcloud performed a system-wide application rebuild process affecting all the deployed apps for all customers. This involved disabling TLS 1.0. For some odd reason, my AWS CF distribution origin was set to SSLv3/TLS 1.0, with TLS 1.1 and 1.2 disabled. wtf. Enabling TLS 1.1 and 1.2, killing 1.0 and SSLv3 on the origin settings resolved the issue.

    Runcloud not only didn’t tell anyone the rebuild was disabling TLS 1.0, they didn’t tell anyone the rebuild was happening. So, that’s that. Hooray for excellent communication with customers. :/

    Plugin Contributor gidomanders

    (@gidomanders)

    Runcloud is not the only one, CloudFlare also changed their SSL settings to only include TLS < 1.3 for paying customers, without telling anyone. This made websites fail on older browsers and operating systems which don’t support TLS 1.3.

    Thread Starter tf5_bassist

    (@tf5_bassist)

    …. only TLS 1.3? That’s not even really supported by a majority of browsers… lol. That’s pretty dumb haha.

    https://caniuse.com/#feat=tls1-3

Viewing 3 replies - 1 through 3 (of 3 total)
  • The topic ‘502 Bad Gateway on CDN minified scripts, images, etc’ is closed to new replies.