Error after restarting the discourse

AI-generated summary

A user (11145) reported that Discourse fails to start after restarting, with logs pointing to an issue involving the discourse-rss-poster plugin. Falco suggested removing the plugin, which temporarily resolved the issue. However, robbyoconnor clarified that the root cause is likely being throttled by RubyGems due to rate limits — not the plugin itself — especially if hosted on AWS. robbyoconnor and later robbyoconnor emphasized that the error logs show RubyGems rate-limiting, and advised contacting RubyGems or switching hosts (e.g., DigitalOcean or Rackspace). cpradio suggested waiting 24 hours before retrying. JagWaugh noted similar issues with pre-built AMIs (like Intuz) on AWS, recommending a clean Ubuntu 16 AMI. Despite updates and re-enabling the plugin, 11145 continued to face startup failures, prompting further speculation about the plugin’s compatibility or the hosting environment. robbyoconnor proposed testing with another plugin to confirm if the throttling is systemic. The user confirmed the issue persists only when the plugin is enabled, but the consensus remains that it’s a RubyGems rate-limiting issue tied to the hosting provider, not Discourse or the plugin per se. The user’s hosting details were shared privately 11145.

Tell me, what could be the error?
After the restart, the discourse stopped running.
Logs can be seen by reference https://pastebin.com/DzUPNhAT

Remove the discourse-rss-poster plugin and try again.

He’s getting throttled by Rubygems. Where are you hosted?

Thanks huge for the help, after removal of a plug-in discourse-rss-poster, discourse has successfully restarted. @LeoMcA what can be problematic with your plugin? You can view?

If you read the errors, you’re being throttled by rubygems. So your issue wasn’t that plugin per se, but that you got throttled.

Tell me, how can I fix this?

Well, you can wait 24 hours and try re-enabling the plugin and doing another rebuild, maybe by then rubygems might not be throttling you (though to be honest, I’m not sure I see it as a throttling issue, considering the error happens between two rubygem downloads/installs… that or I’m reading the output wrong).

The issue is that your ip block is likely hammering the rubygems server…the solution is likely to use something like DigitalOcean, I rarely have this issue. Something is weird but this is not a discourse issue, you should contact rubygems – you can read more about it here. I never actually saw this issue. I use Rackspace for the discourse instance I maintain…I’d bet money this is an AWS issue.

Tried to plugin the plugin discourse-rss-poster , but the same error
https://pastebin.com/H8dSU9CF

You are being throttled due to hitting the rate limit by rubygems. I repeat: this is not a discourse issue.

Look at line 639 onwards in your paste and read it ;).

One other thing you could do: install another plugin in place of discourse-rss-poster – does that cause you to hit the rate limit as well? I suspect it will.

Are you using a plain Ubuntu AMI, or did you take a prebuilt one from the marketplace?

Last week someone was having similar problems with an Intuz pre-built AMI - the issues went away when they moved to a plain Ubuntu 16 AMI and migrated their instance. See here:
https://meta.discourse.org/t/upgrade-error-on-aws-free-tier-t2-micro/63756

And why if I disabled the plug-in ā€œdiscourse-rss-posterā€ then everything works.
But is it worth it to include, then download discourse fails?

Updated the discourse to the latest version, when the plug-in is activated, the discourse never starts. Is this problem in the discourse-rss-poster plug-in?

@11145 – where is this hosted?

Answered in a personal