I have 3 servers configured to run syslog-ng.

We'll call the servers web, jump, and client.

Web is CentOS 6
Jump and Client are CentOS 7

Jump is an OpenVPN server
Client is an OpenVPN client
Web is on the same private network as Jump

Connectivity is working fine between all 3 servers.
I can SSH and ping directly from Client to either Jump or Web, and vice versa.

I have Syslog-NG configured on Client and Web to log to Jump, and Jump is configured to receive logs from both Web and Client on tcp port 514.

Client is able to log to Jump without issues. Web, however, does not seem to want to log to Jump.

For testing purposes, I disabled the firewall temporarily on Jump.

I also stopped syslog-ng on Web and ran it in debug mode. There were no errors related to connecting to Jump.

I also am able to telnet from Web to Jump on port 514, so I know there is no routing issue or firewall issues with it, and I know the configuration is good.

Given that the only difference between Client and Web is the OS version, is Syslog-NG in CentOS 6 just not compatible with Syslog-NG in CentOS 7? I mean has the syslog protocol changed that much that the older version of Syslog-NG won't work with the newer version?

Client: syslog-ng-3.5.6-3.el7.x86_64
Jump: syslog-ng-3.5.6-3.el7.x86_64
Web: syslog-ng-3.2.5-4.el6.x86_64


Turns out after more investigation that I had 2 issues.

1) When the firewall WAS enabled, I had the internal NIC on jump in the wrong zone. Once that was corrected, I was then able to get connected via telnet with the firewall running

2) I compared the config of client against web again and found that I was missing the log line on web which indicated to log to jump and was present on client.

Once the log line was added on web and the daemon restarted, logging began immediately on jump

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.