upcp breaking named?

[hosteurolive]

named Failed

hi

since the release update last night
named is down on all my servers

rndc: connect failed: connection refused

Starting named: [FAILED]
[email protected] [~]#


Jul 21 08:16:48 ns23 named[7330]: using 1 CPU
Jul 21 08:16:48 ns23 named[7330]: loading configuration: failure
Jul 21 08:16:48 ns23 named[7330]: exiting (due to fatal error)
Jul 21 08:16:48 ns23 named: named startup failed
Jul 21 08:22:54 ns23 named[7570]: starting BIND 9.2.4 -u named
Jul 21 08:22:54 ns23 named[7570]: using 1 CPU
Jul 21 08:22:54 ns23 named[7570]: loading configuration: failure
Jul 21 08:22:54 ns23 named[7570]: exiting (due to fatal error)
Jul 21 08:22:54 ns23 named: named startup failed


I dont know what the config problem is...

thx

 

[Salman75]

DNS Failures

Yes, we are also seeing this on all servers running RHEL3.

CPanel needs to check this issue immediately.

 

[heavypredator]

same here :/

i have added ticket - waiting :/

 

[Spearow]

This is due to the rpm update scripts excluding php, but not squirrelmail, which depends on php. This prevents most rpm's from ever being updated. Removing the squirrelmail rpm and running "up2date -u" should resolve this.

 

[heavypredator]

error: package squirrelmail is not installed

looks like its not there

 

[cwihost]

We are seeing this error here as well on almost all Cpanel boxes.

 

[hosteurolive]

[email protected] [~]# rpm -qa |grep squirremail
[email protected] [~]#




exact
no squirremail rpm installed...

and I dont see the rapport between named and php....

 

[heavypredator]

it fixed one of the servers so if your is still not fixed find out why your up2date is not updating and try to fix it

 

[cjscott69]

Seeing this on our RHEL server too...

EDIT: also don't have squirrelmail RPM installed

 

[proton2112]

I just talked to a tech at my host (servermatrix) who got me fixed up right away. Apparently there was a bad update (DOH).

He said something about changing my resolv.conf to point to a server other than myself so 'up2date -u bind-libs' could run. Then he changed resolv.conf back.

 

[janderk]

I had the same problem and servermatrix fixed it for me (within minutes of filing a support ticket. Thanks guys). up2date did not work because the name server was down, on which up2date depended. Adding another outside working name server and made up2date work again. For those interested here is what their support ticket said:

We have added the following line to the top of /etc/resolv.conf:

nameserver [ip of working theplanet name server here]

That line allowed your server to use the planets DNS server to resolve the IP for up2date. We have then updated the bind-lib and bind is back up and running.

This problem is becuase of a cpanel update that failed to install bind-libs and was resolved by running:

up2date -u bind-libs

and restarting bind.

Everything appears to be up and running now.

Installing...
1:bind-libs ########################################### [100%]
[email protected] [~]# service named start
Starting named: [ OK ]
[email protected] [~]# nslookup google.com 127.0.0.1
Server: 127.0.0.1
Address: 127.0.0.1#53

Non-authoritative answer:
Name: google.com
Address: 64.233.167.99
Name: google.com
Address: 64.233.187.99
Name: google.com
Address: 72.14.207.99

Jan Derk

 

[*JPO*]

Hello, I have a few named's that wont start after upcp ran tonight. They were working and went down after midnight (upcp cron). Is there anybody having the same problem? If yes, did you fix it and how?

Thanks a lot!

 

[Jeff-C]

Just encountered this as well tonight - adding squirrelmail* to the up2date skiplist #20 and running up2date solved it for me as well and named is now started again.

 

[cjscott69]

Running

up2date -u bind-libs

and then

service start named

did it for me. Thanks to those who posted the fix here!

 

[flash7]

Primary and secondary DNS down, I can't run up2date !!!!!!!!!

 

[monsterhosting]

This happened on one server and that fix worked w/o playing with the resolve.conf.

I tried a manual upcp on another RHE server but it did not cause the problem so hopefully it is fixed.

 

[rpmws]

i seem to be stuck ..can't get named to resolve any update servers. I tried messing with resolv.conf but named won't run so it can't up2date

 

[chirpy]

You need to make sure that the first nameserver entry in /etc/resolv.conf is a working DNS resolver. For example, it would be a good idea to use the DNS resolver that your datacenter provides. If you don't know what they are, you'll need to seek help from your server provider, unless it's your own hardware and OS, of course.

It's a salutary reminder that you do need to have disperate nameservers defined in resolv.vonf incase the local resolver goes down, as it is in these cases.

 

[rpmws]

You need to make sure that the first nameserver entry in /etc/resolv.conf is a working DNS resolver. For example, it would be a good idea to use the DNS resolver that your datacenter provides. If you don't know what they are, you'll need to seek help from your server provider, unless it's your own hardware and OS, of course.

It's a salutary reminder that you do need to have disperate nameservers defined in resolv.vonf incase the local resolver goes down, as it is in these cases.

actually it is a good nameserver but I may have zone transfers off without trusted setup ..thanks for the tip Chirpy!!! I can fix it now .. 5am blurr makes an old brain think much slower

fixed!!! thanks chirpy! a nice slutty resolver that will put out to anyone did the trick for me

 

[adamint]

I was in the sameboat as Janderk, got a message on my phone saying your servers down, came on. ip was up. dns was down.

restarted it, it failed.

logged a ticket with theplanet, was open less than five minutes
and fixed with the solution Janderk posted.

luckily it didnt affect to many people, but hopefully it wont happen again!