Rails app on Heroku with a verified custom domain issues sandbox subdomain error when sending emails. See updated SMTP config:
# Production SMTP Settings
SmtpConfig.set do |config|
config.server_address = ENV["MG_HOST"]
config.server_port = ENV["MG_PORT"]
config.user = ENV["MG_USER"]
config.password = ENV["MG_PWD"]
config.domain = "custom-domain.xyz"
config.auth_method = :plain
end
I encountered a similar problem with Mailgun on Heroku where the custom domain setup was causing unexpected errors. What helped in my case was verifying that the domain was properly set up in Mailgun and that the necessary DNS records had fully propagated. I also repeated the process of checking that all the corresponding environment variables were correctly set up in Heroku. It turned out to be a configuration oversight that only manifested when the sandbox mode was inadvertently active. Careful scrutiny of both your Mailgun and Heroku settings usually resolves this issue.
Based on my experience, the root of the problem was that the Mailgun configuration was not fully aligned with the settings in Mailgun itself. I discovered that after confirming the domain verification on Mailgun’s side, I had to revisit the Heroku environment variables and ensure they were fully up-to-date. What made a difference was taking the time to inspect the detailed logs provided by both Mailgun and Heroku to pinpoint misconfigurations, particularly issues related to sandbox mode or SPF and DKIM records. A thorough cross-check between both platforms helped resolve the issue.
hey, try double-checking that your custom domain records are apropriatly set in mailgun. i had similar issues, and once i updated the settings & cleared any sandbox mode, it worked. also check env vars in heroku, sometimes they get out of sync.
During my recent deployment, I encountered an issue that felt similar and was quite frustrating initially. I discovered that even when the DNS records seemed correct, a small region mismatch in the SMTP server address was causing the emails to route through the sandbox configuration. In my case, switching to the proper regional endpoint provided by Mailgun resolved the error. Additionally, I audited my Heroku environment variables and confirmed there were no hidden typos in the custom domain entries. This taught me the importance of verifying platform-specific settings side by side.
During a recent project, I experienced a similar email delivery issue on Heroku with Mailgun. The problem turned out to be a subtle configuration mismatch. It was essential to review the exact environment variable values being used for the SMTP configuration. I found that even a minor discrepancy in variable names or an outdated entry caused Mailgun to revert to sandbox mode unexpectedly. Meticulously comparing these values with the Mailgun account settings, particularly for domain authentication records, helped uncover the root cause. Verifying regional endpoints and any potential discrepancies in DNS entries also proved valuable in resolving the delivery problem.