DNS config on a new 2.9 collector

[hgieselman]

I just installed a new 2.9 collector. During initial config, setting the DNS server kept failing so I continued anyway. Now I am in the admin shell and trying:

admin# conf dns -a 192.168.10.4

but it keeps failing with:

execute % Error occured while executing command

Any ideas?

[Contratos Teltec]

I'm in the same situation! 

[Contratos Teltec]

Hello,

I had the same problem, so I installed it again in version 2.8 and the DNS was configured automatically, as expected.

Now we have another problem, but this is apparently a version problem.

[afarrior]

Similar problem with 2.9.0.2; so, it's not fixed yet.

I can define DNS via the CLI and even ping hosts using FQDN from the CLI.

However, in the GUI, the ping utility and SMTP settings among others fails any DNS queries.  If you use IP addresses it works.

  Administration -> Email Settings -> SMTP Server    => Error modifying Mail settings. Smtp Mail Settings <FQDN> does not exist

  Administration -> Diagnostic Tools -> Ping  =>  Invalid Host name: <FQDN>

Updates using the GUI also fail.  Had to upload files and upgrade via CLI.

[sfletcher11]

same issue but manual define DNS in CLI results in

execute % Error occurred while executing command

[Jay Stants]

login to the shell as collectorlogin account

elevate permissions to root using su - 

edit \etc\resolv.conf with vi 

add your name servers and search domain

example

nameserver x.x.x.x

nameserver y.y.y

search corp.somedomain.com

save and exit

verify that this saved by running cat against resolv.conf

example: cat \etc\resolv.conf

you should see your entries listed and be able to resolve dns correctly now with permanent entries that persist after reboots. 

hopefully this helps, please mark if this solution works for you

[sfletcher11]

this definitely helps. with this i was able to specify a dns server. However the file resolv.conf actually didn't exist until i manually created it.

Now the problem is the nameserver lines blank out after a reboot.  the resolv.conf file stays, and now i can definitely set it though from the CPSC gui , so progress anyways.

[Jay Stants]

Has your collector been patched yet? I just installed mine today and patched it then did the above steps.

[afarrior]

issue still exists.

from the CLI:

[root@ciscocollector collectorlogin]# cat /etc/resolv.conf
# Generated by NetworkManager
search victoriacollege.edu
nameserver 172.16.x.y
nameserver 172.16.y.z
[root@ciscocollector collectorlogin]# ping www.google.com
PING www.google.com (172.217.9.164) 56(84) bytes of data.
64 bytes from dfw25s27-in-f4.1e100.net (172.217.9.164): icmp_seq=1 ttl=116 time=8.34 ms

# show version -d
Build-name : Collection Platform Software 2.9.0.2

Package-type : ServicePack
Version : sp-30.2.0-2-0-lnx64
Component : CSPC Base
Version : 2.9.0.2
Package-type : JeOS
Version : jeos-30.2.0-2-lnx64
Component : AdminShell
Version : 1.9.3
Component : LCM
Version : 1.9.3
Component : CASP
Version : 2.9.0.2
Component : ConcsoTgw
Version : 1.7.9
Component : Hardened CentOS
Version : 8.1 patch#0

[Jay Stants]

strange I've run the same test and don't have any issues with performing this test either in the Gui or at the CLI. are you able to ping something on the local lan by ip and fqdn? 

[afarrior]

I get same results with local IP/FQDN

From CLI:

ping internal IP address: success

ping internal FQDN: success

From GUI

ping internal IP address: success

ping internal FQDN: fail

What's odd in the GUI is viewing the "Appliance Config", doesn't show any settings for DNS, Timezone, NTP, Time, or hostname.  I'd expect it to at least display what it's using.  When I enter something in Appliance Config->DNS, it'd expect it to update the /etc/resolv.conf file, but it does not.  The confirmation prompt shows it updated DNS, but nothing displays in Appliance Config->DNS.

It's almost as if the java tools are using a separate set of settings from the OS and the gui tools don't have permission to update those corresponding files.