Beta Branch how to fix cmupdate and dnf update timeouts?

Please fill in any relevant information that applies to you:

• CentOS Version: AlmaLinux 10.2
• Centmin Mod Version Installed: 141.00beta01.
• Nginx Version Installed:
  Code (Text):

  1.31.8/CODEB]
  [*][B]PHP Version Installed:[/B] [CODEB]8.3.36

• MariaDB MySQL Version Installed:
  Code (Text):

  10.11.19

• When was last time updated Centmin Mod code base ? : just now, sadly not sucessfully. Last sucessful run was end of october.
• Persistent Config:
  
  Code (Text):

  SET_DEFAULT_MYSQLCHARSET='utf8mb4'
  SELFSIGNEDSSL_ECDSA='y'
  PHPFINFO='y'
  PHP_OVERWRITECONF='n'
  PYTHON_INSTALL_ALTERNATIVES='y'
  LETSENCRYPT_DETECT='y'
  NGINX_GEOIPTWOLITE='y'
  NGXDYNAMIC_GEOIPTWOLITE='y'
  # MaxMind license key is OPTIONAL - Centmin Mod downloads GeoLite2
  # databases from its own mirror by default (no account needed).
  # Only set MM_LICENSE_KEY if you prefer direct MaxMind downloads:
  MM_LICENSE_KEY='key removed'
  

I have two vps' with 141.00beta1, both on Alma 10.2 and both in the same data center. Both used to run flawlessly. However, short to two weeks ago one of them has out of nowhere stopped updating. Luckily it is the dev environment that is affected, not the productive one.
It may be that it is caused by myself changing some relevant config file - however, I am not aware of doing something like that.

On login, motd takes forever and then states centmin would be up to date - which is a lie, sadly.

---

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
[OK] Centmin Mod 141.00beta01 — up to date
[OK] CSF 16.32 (matches mirror)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Output of cminfo versions is:

# cminfo versions
Warning: smem not available for EL10, skipping...

last 10:
141.00beta01.b243 #Sat Jul 4 02:28:53 PM CEST 2026
141.00beta01.b243 #Sat Jul 4 05:23:22 PM CEST 2026
141.00beta01.b243 #Sat Jul 4 09:11:45 PM CEST 2026
141.00beta01.b245 #Wed Jul 15 09:19:09 PM CEST 2026
141.00beta01.b245 #Wed Jul 15 09:22:55 PM CEST 2026
141.00beta01.b265 #Sat Sep 5 12:25:57 PM CEST 2026
141.00beta01.b328 #Wed Oct 21 06:47:03 PM CEST 2026
141.00beta01.b341 #Thu Oct 22 09:19:17 AM CEST 2026
141.00beta01.b397 #Mon Oct 26 06:33:18 PM CET 2026
141.00beta01.b407 #Wed Nov 4 05:43:06 PM CET 2026

The trouble started at the end of October, so I am somewhat surprised to find b407 there, the more with the date Nov 4th, b/c for one the issues started earlier and secondly on my other VPS b415 was there already on Oct. 31th. So something seems to have gone out of sync here.

My working vps states:

last 10:
141.00beta01.b403 #Wed Oct 28 06:21:49 PM CET 2026
141.00beta01.b403 #Wed Oct 28 06:58:56 PM CET 2026
141.00beta01.b415 #Sat Oct 31 03:28:10 PM CET 2026
141.00beta01.b415 #Sat Oct 31 03:36:02 PM CET 2026
141.00beta01.b415 #Sat Oct 31 05:46:39 PM CET 2026
141.00beta01.b415 #Sat Oct 31 11:02:04 PM CET 2026
141.00beta01.b415 #Sat Oct 31 11:26:33 PM CET 2026
141.00beta01.b415 #Sun Nov 1 12:55:01 AM CET 2026
141.00beta01.b415 #Sun Nov 1 01:53:11 AM CET 2026
141.00beta01.b420 #Sun Nov 8 10:02:16 AM CET 2026

Running cmupdate on the troubled VPS manually results in a timeout

# cmupdate

-------------------------------------
Updating Centmin Mod code
-------------------------------------

* 141.00beta01
Checking for updates on origin/141.00beta01 ...
fatal: unable to access 'https://github.com/centminmod/centminmod.git/': Failed to connect to github.com port 443 after 19581 ms: Could not connect to server
git_sync: ERROR — git fetch failed for origin/141.00beta01 (network or auth)
Latest commit: bec86aa (2026-10-29)
add cmsec ZDI-CAN-33367 skb_zerocopy kernel check in 141.00beta01

-------------------------------------
Updated Centmin Mod code
-------------------------------------

Note the

Latest commit: bec86aa (2026-10-29)

which seems to indicate when the issue started and that the Nov 4th stated on cminfo versions seems questionable.

Running dnf update results in a timeout, too.

host github.com respectively host mirrors.almalinux.com both show the correct IP-addresses (the same as on my second box that is working fine). Curl https://github.com respectively https://mirrors.almalinux.com runs in a timeout. Pinging these two addresses results in 100% data loss. So it sounds like a networking issue on first look.

However: Curling or pining any other host succeeds. As to my knowledge there are no csf rules that would stop those outgoing connections.

As only these two hosts to be seem affected (and not just one host or all hosts) a networking issue seems less likely, therefor I suspect, that there may be some kind of caching involved or some other mechanism tied to centmin that hinders the update from working. The strange thing is that yum update/dnf update is affected, too.

Any ideas how to find the root cause and how to get the thing back flying?