I'm having trouble getting dhcpd to update DNS records in bind9. I'm not seeing any journal files created for bind9. The following appears in syslog when a client machine connects to the network:
Code:
Jun 26 04:07:22 ubuntu dhcpd: DHCPDISCOVER from 00:23:df:44:78:83 via eth0
Jun 26 04:07:23 ubuntu dhcpd: DHCPOFFER on 10.0.0.112 to 00:23:df:44:78:83 (Computer1) via
[code]....
I've configured my dhcp server to update dynamically the DNS zones of BIND9 name server. Everything works ok, except for reverse nslookup:
nslookup 10.141.50.3
Server:127.0.0.1
Address:127.0.0.1#53
[code]....
it appears i haven't quite grasped the meaning of dhcpd.conf despite the numerous tutorials i went through in the internet and the detailed read of dhcpd.conf that i did. what i basically want to put you in perspective, is to have all the pc's in my home network (desktops and laptops) regardless of any other settings, to send their computer name (hstname basically) to the dhcpd (during the handshake)... then the dhcpd will use that name as the hostname and append the known domain name, thus creating an entry (an A record and a PTR record) in BIND....
so say a laptop comes into wifi range of the domain "myhome.org."... i take it that during handshake, the laptop will send its hostname (i don't know what that will be for winXP and win7 dhcp clients, but i presume it will send the output of command 'hostname' for linux clients - or at least thats what i would like it to send....).say "nass-notebook" Then the server will append "myhome.org." and it will update the DNS with an entry looking like[URL].. corresponding to some ip address.
[Code]...
I've found a lot of pages of how to fix the issue on older releases of dhcpd and ubuntu.
#>dhcpd
#> ...
#> Can't create PID file /var/run/dhcpd.pid: Permission denied.
I have tried creating the file and giving it permissions to be accessed, but that isn't solving the issue. I tried to go through some of the older fixes, and one was really close to working but lacked the same folders as I have.
EDIT: one solution that worked was to use ln -s /var/run/dhcp3-server/dhcpd.pid /var/run/dhcpd.pid dhcpd3 -cf /etc/dhcp3/dhcpd.conf at0
Im install bind9 in ubentu 10 with this turotial [url]my dns server is work in linux but my client user cannot use dns when client run nslookup in cmd Cannot find server name for address my_ip_linux :server faild Server: Unknown. im on linux valid ip
View 5 Replies View RelatedI want to make BIND9 aware of my TLD so I need to give it an A record. Then I wanted to create CNAME for ftp, mail and so on as time permits
I have webmin so I need to work from that.
my TLD is [URL]
I wanted to bolt on www, ftp for now
www is port 80, ftp port 21 etc
when i started bind9, it showed me this error, Can anyone help me in this, according to me i dont have a file sysklogd so i can not restart that. that is why i think iam getting error while starting bind9. How to get that file and how to get started Bind9.
View 4 Replies View Relatedusing Ubuntu Server 10.10 x86_64 on this machine. It is used as a master DNS server for my domain. We have migrated it to Ubuntu from Gentoo. The problem is that AppArmor is spamming /var/log/syslog
Code:
Jan 3 14:38:40 hydra kernel: [2154828.893409] type=1400 audit(1294061920.141:660146): apparmor="DENIED" operation="mknod" parent=1 profile="/usr/sbin/named" name="/var/log/named_querylog" pid=15397 comm="named" requested_mask="c" denied_mask="c" fsuid=103 ouid=103
The zone files reside in /etc/bind/ and we have not changed anything in /etc/apparmor.d/usr.sbin.named .
We don't want to just uninstall apparmor, but how do we adjust its settings correctly?
Since I have not gotten anything on the "General Help" I am posting this here: When a Ubuntu 10.10 I have starts up apache2, MySQL and postfix start properly but bind9 doesn't. Once booted is I run 'sudo /etc/init.d/bind9 start' it starts. The only thing odd on this system is I have a "inet6 v4tunnel" interface defined in my /etc/network/interfaces. From booting in the syslog there is:
Code:
/var/log/syslog:Feb 28 19:02:42 ubuntu named[1029]: starting BIND 9.7.1-P2 -u bind -d 9
/var/log/syslog:Feb 28 19:02:42 ubuntu named[1029]: built with '--prefix=/usr' '--mandir=/usr/share/man' '--infodir=/usr/share/info' '--sysconfdir=/etc/bind'
[code]....
Even with the bind debug level at set at 90 I do not get any more in the logs that what I posted before. The lack of errors in the logs is giving me no idea where to start. I commented out the IPv6 tunnel interface and rebooted and still no help. bind still is not starting at boot. OK, it is not a permissions issue, I have changed the config and zone files to be owned by bind:bind, root:bind, root:root and bind:root and it has made no difference. Doing a fresh install on a Virtual Box VM and configure it in a similar manner bind9 starts normally. (I am almost to the point where I will rebuild the box)
I have googled for hours and can not find an answer to what this means.
dig @172.16.255.254 twitter.com.
; <<>> DiG 9.7.0-P1 <<>> @172.16.255.254 twitter.com.
; (1 server found)
;; global options: +cmd
;; Got answer:
[Code]...
And I have a very long debuging log file I will not post unless requested. I also have added a ufw allow from all to all rule for testing purposes only.
I am migrating my server from Kloxo (lxadmin) to Ubuntu (webmin/virtualmin), and I already had my Nameservers on my register (Godaddy) to go to [URL] and [URL] along with my IP. (I set the ttl to 60 cuz I thought that might be a problem) Now, I do a dig [URL], it gives me this output.
Code:
; <<>> DiG 9.6.1-P2 <<>> thedomz.com
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 40276
[code]...
I can go to my website only when I change my hosts file (on my windows machine).
I operate a home network with Ubuntu Server 10.04 with services including DHCP3, Bind9, Apache, and so on. Since I host several dozen websites from home, I have to run Bind DNS. All Ubuntu boxes on my network operate fine. However, all Windows boxes on the network seem to forget to look internally for DNS after a couple of page loads on my internal sites. The network settings still indicate that my internal domain name server is the first lookup and everything seems normal.
View 2 Replies View RelatedI am using samba4 and bind9 (9.7.2rc1) on an ubuntu server 10.04.
Everything works fine, but bind9 ist unable to update DNS entries sent by clients.
I followed the tutorial [URL] but it doesn't work for me.
It doesn't look like a file permission problem:
Code:
-rw-r--r-- 1 root root /usr/local/samba/private/dns.keytab
-rw-r--r-- 1 root root /usr/local/samba/private/dns_update_list
-rw-r--r-- 1 root root /usr/local/samba/private/named.conf
-rw-r--r-- 1 root root /usr/local/samba/private/named.conf.update
[Code].....
I am trying to configure Sarg to read my Squid access.log file.When I run Sarg from either the command line or from Webmin I get the following error:
Code:
SARG: Records in file: 8343, reading: 100.00%
SARG: No records found
[code]....
I am trying to install dhcpd on my Fedora 13 VM. I'm using VirtualBox and I installed using the i686 image. I did yum update and everything was fine. When I do yum install dhcpd , I get the following error.
Quote:
Loaded plugins: presto, refresh-packagekit
Setting up Install Process
No package dhcpd available.
Nothing to do
I also tried yum clean to clear cache but that didn't help either. I installed Fedora 12 in a different VM and tried but with no luck. Also tried installing using i386 DVD but same error. I should also mention that what I'm trying to do is that I'm trying to set up a relay agent on my Fedora gateway.
I am trying to get bind 9 up and running on a new installation of server 10.04 and keep getting errors when starting the daemon -
* /etc/bind/named.conf.local:19: unknown option 'zone'
* /etc/bind/named.conf.local:20: unknown option 'zone'
* /etc/bind/named.conf.default-zones:3: unknown option 'zone'
* /etc/bind/named.conf.default-zones:11: unknown option 'zone'
* /etc/bind/named.conf.default-zones:16: unknown option 'zone'
* /etc/bind/named.conf.default-zones:21: unknown option 'zone'
* /etc/bind/named.conf.default-zones:26: unknown option 'zone'
The onlything I could find after scrubbing the internet is once ina while there's a missing bracket. But I have checked and triple checked my config files and the syntax and can't find anything wrong. I have another installation of bind9 running on Fedora and the same exact syntax works just fine.
[Code]....
Im using bind9 as DNS server on my LAN, but it does not seem to translate its own hostname correctly for some reason. Other hosts is translated correctly, the problem only seems to apply to the DNS host itself. if i "ping <server_hostname>" from the server, it translates correctly. But if i "ping <server_hostname>" from the client it only says "unknown host" The client has the correct DNS-server assigned. How can i start troubleshooting this?
[Code]...
After upgrading from 9.10 to 10.04, a hundred or so of (for example)
Code:
error (network unreachable) resolving '14.42.117.203.in-addr.arpa/PTR/IN': 2001:500:13::c7d4:35#53: 1 Time(s)
messages appear in syslog each day. Each URL has from 2 - 6 attempts at various ipv6 addresses. My question is why is bind9 trying to resolve ipv6 addresses? I have done nothing to enable or disable ipv6 and thought that if not explicitly enabled I would not have to be concerned with it.
I am having trouble with setting up BIND9 for 6 virtual servers that use ubuntu x64 v10.10. I have main server running ubuntu as well. host name is xeonserver I would like to explain my setup first.
my router ip: 192.168.1.1/24 host server for VMs ip: 192.168.1.2/24 Then on qemu my virtual machines are in 10.0.0.0/24 network, gateway to my router is 10.0.0.1
1. kerberos.xeonserver (not configured yet) 10.0.0.2
2. dns.xeonserver (the one I have trouble with) 10.0.0.3
3. mysql.xeonserver (not configured yet) 10.0.0.4
4. apache.xeonserver (not configured yet) 10.0.0.5
5. ftp.xeonserver (not configured yet) 10.0.0.6
6. mail.xeonserver (not configured yet) 10.0.0.7
To configure it I followed instructions found on [URL]
[Code]...
I have installed bind9 for DNS. I have also installed Apache2, PHP5, MySQL, and correctly configured these together. When I view domain1.co.uk it currently works fine as required. I would like to change it to:
* /var/www/domain1.co.uk/prod/ - becomes the root dir of domain1.co.uk
* /var/www/domain2.co.uk/prod/ - becomes the root dir of domain2.co.uk
Nevertheless, I have created a the following files; these are copies of the "default" file and correctly amended. the required destination directories are in place!
* /etc/apache2/sites-available/domain1.co.uk
* /etc/apache2/sites-available/domain2.co.uk
I then executed:
Code:
sudo ln -s domain1.co.uk ../sites-enabled/domain1.co.uk
Code:
sudo ln -s domain2.co.uk ../sites-enabled/domain2.co.uk
It is important to understand domain1.co.uk and domain2.co.uk represent 2 Existing domain names I own, and both point to my server.
domain1.co.uk currently works in /var/www
Why I would like multiple domain names on the server.
For some weird reason dhcpd fails to startup, but when I installed gadmin's dhcpd GUI configuration tool and click activate to start the DHCP server. It works, but for some reason I can't get it to start like a regular service. (Service dhcpd start) Fails without a error code or message.
View 14 Replies View RelatedIs there a way to use dhcpd to assign vpn clients their IP and configuration?
I'd like to have dhcpd listen on br0 (already set as 10.0.0.1) so that devices conecting through eth1 and tap0 (should I be using tun?) will get the same address pool.
I just got the problem while starting the dhcp daemon. Its showing failed. I need to start it for my clients to be autoinstalled. I am having CentOS 5.2 linux. I've installed all optional packages also while installing CentOS. How to start the dhcp daemon? I've tried the /etc/init.d/dhcpd start also. But it didnt work. Showing Failed.
View 1 Replies View RelatedI would like to send a specific option 15, domain-name, to a few clients on a network from a specific manufacturer. Usually all clients [URL] from the DHCP server, but when a client with mac address bellonging to manufaturer A asks for an IP address I would like to give them [URL] How would I go about doing this? Feels like it should be possible but I am not sure how.. I remember doing something similar in a microsoft DHCP server using vendor-identifier and passing out a vendow-specific option.
Quote:
class "xxx" {
match if substring (hardware, 1, 3) = 00:00:10;
option domain-name "yes.this.works.com";
}
/Carl
I'm at a loss to why my reverse lookup zone doesn't work for me.I've got two views. One internal and one external. My domain is isp2.datornatverk.se. Public IP: 130.240.133.81.
dig -x @8.8.8.8 130.240.133.81
gives me:
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 2917
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0
I've set it up so that the internal subnets gets the domains resolved to the internal IP-addresses. When querying from external addresses I will get public IP.My named.conf.local file:
Code:
acl internals {
127.0.0.0/8;[code]..........
I don't know whether the views has messed something up. It worked before I added the views.
The problem is that I have set up a test installation with DHCP server and DNS server on Fedora 14 and want to set up dynamic updates from the DHCPD to the NAMED. The problem is that any attempts to make updates of the DNS server fails with a "SERVFAIL" error. I can not even update it through the NSUPDATE tool. This is NAMED.CONF:
Code:
options {
directory "/var/named";
[code]....
I installed dhcpd using:
Code:
Now when I try to run it, I get this weird error:
Code:
The error was:
I changed permissions to 777:
Code:
To no avail!
It's got something to do with AppArmor.
I don't think it is installed on this system:
Code:
I have installed ntp using Code: apt-get -y install ntp.Logging to /var/log/ntpstats has been enabled in /etc/ntp.conf however this directory is not getting populated with anything. I have also tried removing it with apt-get --purge remove ntp and reinstalling but to no avail.
View 9 Replies View RelatedI've been running my little server at home for a few months now, and I've noticed that Webmin has detected that I'm in need of over 100 updates. I'm a bit scared to run the updates because everything is working just fine right now, but a part of me still wants to run some of the updates.Now I know backing up the server would be a good idea to do before this happens. This server is a media server with videos, music and pictures taking up the majority of the hard drive space on the server. I just have one partition on the server, outside of the swap partition. I would like to back up everything on the server except the videos, music, and pictures, because if even an update messes with the server, I could always retrieve those, and the external hard drive I'm going to back up to wont be big enough to hold everything anyways.
if I were to backup up everything except for those directories that hold my videos, music, and pictures, and something were to go wrong, if I were to restore all of those, would I then be back to the state my system was at when I backed up? I've never done a back up and restore in a linux environment before. I just want to make sure that just doing that will be enough, because the last thing I want to do is hose my server after taking several weeks to get it to the working state that its at.
I'm running a virtualised backupserver. This makes backups to an LVM volume on the host. /dev/mapper/system-backupserver is put through as a block device to the KVM-based, libvirt-managed backupserver and there it's mounted (obviously) rw.
I've also mounted the same volume, as ro, on the host, so as to get at my backups more easily. So now the same partition is mounted on two different machines, both thinking they've got sole access.
Now when the backupserver VM alters the contents of this partition, it sees this happening normally on its df report, but on the host machine, this remains mostly unseen. I tried doing a mount -o remount, but that doesn't help, although umount/mount will do it.
Meanwhile I seem to be able to get at all the data just fine, and it's just the total used/free that gets out of whack. Is there any way to regenerate/reread those totals without doing a full umount/mount cycle?