altayevrim
Verified User
- Joined
- Oct 6, 2019
- Messages
- 44
Hi all,
This is not a typical "the service is down, and I don't know what to do" situation. I already fixed the Dovecot issue, which was quite obvious. I had disabled SSL for a domain behind Cloudflare because CF handles it, and DirectAdmin was constantly trying to get Let's Encrypt certificates and failing.
Then, I updated the kernel two nights ago and didn't realize that disabling an SSL certificate could lead to this error
It was an easy fix (just moved the /etc/dovecot/conf/sni/mydomain.com.tr.conf file and restarted Dovecot). However, the reason I'm opening this topic is that since Dovecot was down, I couldn't receive the 38 important message notifications DirectAdmin was trying to send.
https://shottr.cc/s/15tG/SCR-20260706-kyt.png (For those who prefer not to open external links, this is a screenshot of my messages screen showing 38 auto-generated notifications from DirectAdmin stating that The service dovecot on server myserver.mysever.com is currently down, which I never received because Dovecot itself was failing.)
So, what should I do in such situations? The mistake on my end was being in a rush; I just did the regular updates and didn't check if all the services were running properly afterward. That is obviously my fault. To reduce the risk of this happening again, I'm planning to implement a workaround, but I wanted to get your opinions first.
I am planning to create a /usr/local/directadmin/scripts/custom/user_message_post.sh script to send admin notifications via Pushover, which I find to be very reliable.
(CloudFlare blocks my request when I try to paste this code so I pasted it to pastebin: https://pastebin.com/w8AVVf7f)
Any other ideas or suggestions would be greatly appreciated.
This is not a typical "the service is down, and I don't know what to do" situation. I already fixed the Dovecot issue, which was quite obvious. I had disabled SSL for a domain behind Cloudflare because CF handles it, and DirectAdmin was constantly trying to get Let's Encrypt certificates and failing.
Then, I updated the kernel two nights ago and didn't realize that disabling an SSL certificate could lead to this error
Code:
doveconf: Fatal: Error in configuration file /etc/dovecot/conf/sni/mydomain.com.tr.conf line 2: ssl_server_cert_file: open(/usr/local/directadmin/data/users/myuser/domains/mydomain.com.tr.cert.combined) failed: No such file or directory
https://shottr.cc/s/15tG/SCR-20260706-kyt.png (For those who prefer not to open external links, this is a screenshot of my messages screen showing 38 auto-generated notifications from DirectAdmin stating that The service dovecot on server myserver.mysever.com is currently down, which I never received because Dovecot itself was failing.)
So, what should I do in such situations? The mistake on my end was being in a rush; I just did the regular updates and didn't check if all the services were running properly afterward. That is obviously my fault. To reduce the risk of this happening again, I'm planning to implement a workaround, but I wanted to get your opinions first.
I am planning to create a /usr/local/directadmin/scripts/custom/user_message_post.sh script to send admin notifications via Pushover, which I find to be very reliable.
(CloudFlare blocks my request when I try to paste this code so I pasted it to pastebin: https://pastebin.com/w8AVVf7f)
Any other ideas or suggestions would be greatly appreciated.
Last edited: