• Inviting everyone who uses WordPress management tools in Plesk
    The Plesk team is conducting a 60-minute research session that includes an interview and a moderated usability test.
    To participate, please use this link .
    Your experience will help shape product decisions and ensure the tools better support real-world use cases.

Update Broke Bind?

S

Superman857

Guest
I'm running Plesk 10.0.1. Today, I ran a yum update to update all packages on the machine (CentOS) and it seems like it might have broken BIND, perhaps...or something else did.

After the yum update, everything worked fine (or so I thought). However, I just noticed that the site was down and BIND was not running. I think it must have been since the update, but my web proxy at home had cached info for the domain. Tried starting it and obtained this message:

dnsmng: Service /etc/init.d/named failed to start

Trying to start via console provided a bit more information, regarding a zone testdude.com that didn't have records properly. Looking inside /etc/named.conf revealed that the zone was in there (not mine). Commenting it out fixes the problem and BIND starts.

So, my question is this: where did this zone come from? I didn't add it manually or through the interface. Would it be added in during a yum update? Did Plesk do something over the last few hours to add it in? I haven't had this issue before or dealt with BIND too much in the past so I'm not sure what the default zones are or how they are updated really. I'd just like to know where it came from so that I can keep it in mind in the future and avoid sites from going down.
 
Last edited by a moderator:
$ host testdude.com
testdude.com mail is handled by 1 inbound.testdude.com.criticalpath.net.

Parallels has no relation to this zone.
 
On rh/ce zones stored in /var/named/run-root/var, zones includes in /var/named/run-root/etc/named.conf

For bind start with debug info U can call 'sh -x /etc/init.d/named restart'
 
Back
Top