CentOS 5 Server :: Name Resolution Of Host Names Using Bind?
Mar 21, 2011
I installed bind & did not install chroot. I set up a fictional domain kelly.local. I am able to resolve FQDN (example: angus.kelly.local) in both dig & nslookup & ping on linux boxes.
I want to be able to resolve bare host names (example: angus) using bind. I get mixed results.
(1) linux boxes resolve bare host names & FQDNs just fine using nslookup & ping, but not dig. dig gives error:
I want to use single label host names on my local LAN, without using any domain (at least no registered one). My machines should be named in a fashion similar to 'myserver', 'mydesktop' etc. so that i from a browser on any LAN machine may write 'http://myserver', and get the webserver on 'myserver' (so NO domain part, e.g. myserver.domain).
Do i need to create a DNS zone for each host or is there a way to put all host in a single zone, if so; which? Would such a name be considered a root domain?
BackgroundI have registered a domain name "abc.mobi" with domainit. abc.mobi is pointed to our mail server which is running at 199.2.2.2 (fake address).All mails that are being delivered to kazmi@abc.mobi have no problem getting delivered. We can also send emails out to other domains.I have recently configured BIND 9.3.6 on the same server (running CentOS) that is running our mail server. The purpose is to make sub domain dynamically without getting domainit involved. These subdomains will have incoming and outgoing emails as well. For example, I would want to make a usable subdomain "xyz" under abc and have kazmi@xyz.abc.mobi email address.
I have configured abc.mobi.hosts file as follows (/var/named/chroot/var/named/abc.mobi.hosts) $ttl 38400
I have an RHEL5 server running BIND. I'd like to be able to resolve internal, non-FQDN names such as "wiki" for the company internal wiki. in my named.conf file...Named-checkzone returns no errors, and nslookup works when specifying the server explicitly, but no browser resolves the name when entered in the address line. I know this by no means a standard use of BIND, but I'm sure someone out there's pulled this off. I did see when googling someone suggest "load the root zone (.) as master, and add your "hosts" as TLDs, using CNAMEs", but tried that and probably did it wrong since it didn't work.
I'm new to using a BIND DNS server. I currently have one set as a master zone on my LAN. My question is how can I add a record for my local lan that will resolve without typing in the FQDN? For example I want to be able to type "server" instead of "server.domain.com".
dammit... all that typing and I hit the wrong "submit" button. *sigh*
Ok, I have a couple of SMTP servers for our infrastructure. They are running Postfix. I have them configured so that specific email addresses such as support@mydomain.com and billing@mydomain.com all go to a new support server that I am building with osTicket. Lets call that server SUPP1.
SUPP1 runs sendmail from the default install of CentOS 5 i386. At this point everything runs great. New emails get added into the osTicket system via a pipe in sendmail. Here's where the problem comes in. In order to accept mail, sendmail has to have the domain listed in local-host-names and the addresses in virtusertable. That works just dandy. But in doing so, sendmail believes it is the destination SMTP server for "mydomain.com". That means that I can't send mail from that server back into my normal SMTP servers. So things like the LogWatch, cron jobs, etc can't send notifications. Is there a way to work around that? For sendmail to ignore local-host-names for outbound email or something?
What I am trying to do is log IP Address, MAC Address and the host name for records. So I can have a record of what PC got which IP address and when they got it. So far my search has yielded no results. It would not be that big of a deal as I could always look at the dhcp file, but it is generated dynamically (3rd party application using RADIUS and each user gets a host).
i am trying to run bind in centos 6 and bind keeps giving me errors every time i check all of the configurations. the named.conf file works fine but the zone files keep giving me errors heres the contents of my named.conf file
Am running the latest CentOS5 with Bind. Bind will run for a period(time period unknown) the shutsdoown. We need to every day start the service. When the service starts there are no errors given.
I had centos 5-5 server with 6G of RAM and 4 core cpu 3GHZ i installed bind 9.7.1-p2 on my server with multi thread support there are a lot of dns requests on my server , about 2500 Packets/sec and 3Mbit UDP traffic but my server response week to most of them.
For example when i use nslookup or dig command to query Yahoo.com the response from server maybe deliver about 5 Sec or become timed out , but sometimes response time less than 1 sec!
I don't know why, perhaps kernel works week so i decided to do the following :
But the problem didn't solved
I previously had Freebsd with same version of bind and same configuration and everything worked fine.
don't find box of BIND so i was posting into this boxI have some problem when i config bind DNS for my domain and then i can't start named,this's error message
[root@server1 named]# service named restart Stopping named: [ OK ] Starting named:
I installed BIND 9.7.3 from source on Centos 5.5, and chrooted it, and I'm getting an error when I run 'service named status' I get a reply, but at the end it says 'named dead but pid file exists'
Here's the entire output:
[root@ns etc]# service named status version: 9.7.3 number of zones: 23 debug level: 0 xfers running: 0 xfers deferred: 0 soa queries in progress: 0 query logging is OFF recursive clients: 0/0/1000 tcp clients: 0/100 server is up and running named dead but pid file exists Nothing is logged in /var/log/messages. Named is running and responding correctly. How can I get rid of this error?
i have xp and virtual guest centos 5.i install apache, bind, squid and webmin through xp, i can access URL...but when i start my squid, put the ip in the browser proxy settings, when i logon its ip turn to URL...I believe my bind is working. Though theres some little glitch. I dont know where.and i already put my servers bind ip to my winxp primary dns server settings.
I can't get the client server to authenticate to the openldap server. I can authenticate on the server itself though. I can su to, login and shh into the openldap server and become a ldap user. I just can't become an ldap user on the client.I didn't setup TLS/SSL. I can do that after I have it working. I'm using hashed passwords though. I don't have replication setup. I'm am tying to setup the most basic openldap environment then build from there. I have read the openldap section in the admin guide.
My setup at home.
Openldap server � light.deathnote.net -- 10.0.1.21 client server � vm-centos01.deathnote.net � 10.0.1.7 -- VM on virtualbox Virtualbox host � L (OS MAC) � 10.0.1.2 router (apple airport extreme) / default gatway � 10.0.1.1
All computer can reach the internet and ping each other. When I installed centos I disabled SELinux.I used these guids to setup my openldap.
[URL]
Below I have included some output from the files I'm using with openldap.
[root@vm-centos01 ~]# tail /var/log/messages Jul 2 09:25:33 vm-centos01 xfs: nss_ldap: reconnecting to LDAP server (sleeping 16 seconds)... Jul 2 09:25:49 vm-centos01 xfs: nss_ldap: failed to bind to LDAP server ldap://light.deathnote.net: Can't contact LDAP server Jul 2 09:25:49 vm-centos01 xfs: nss_ldap: failed to bind to LDAP server ldap://10.0.1.21/: Can't contact LDAP server
I have setup two BIND9 servers as slaves for an internal Windows domain. I receive messages in my logs about a Windows server not being the master for the slave domain on BIND. I have placed the allow-notify statement in the global options section of named.conf, as well as setting the IP address in the masters section of the zone. I'm confused as to why I'm still getting this error message.
We installed CentOS 5 at our school and are attempting to get 2 simple DNS servers online. These are simple name resolution servers for our network. We installed the needed caching nameserver packages via Yum and all seems to be well, but for some reason when the server is online, it fails to resolve DNS records except when we direct the requests to its 127.0.0.1 (localhost) interface. As for the server specs, the DNS server is running on a Virtual Machine, and the ESXi server has 2 quad-core Xeon CPU's with 48 Gigs of RAM (in short.. hardware is adequate.)
Here's more details: When the box boots up, we can confirm that the named service is running and it resolves DNS records when we direct queries to its localhost address: [root@ns1 ~]# dig hotbot.com @localhost ; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> hotbot.com @localhost ;; global options: printcmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 55510 ;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 4, ADDITIONAL: 0 ..... ; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5_4.2 <<>> hotbot.com @66.186.176.2 ;; global options: printcmd ;; connection timed out; no servers could be reached
What would cause the Named service to not answer DNS queries from the physical IP but only from it's internal IP?
I've been looking for a good tutorial for setting up a BIND DNS server for my local network. What I want to do is..Have BIND running on my home server receiving all DNS requests.Have certain zones (my.zone.lan) pointing to custom IP addresses (I.E. server.lan points to 192.168.{server IP})Zones that don't exist should be passed on to OpenDNS for processing.
Don't work nslookup from clients guest OS.I have LinuxMint 7 and I'm installed VirtualBox on her. I created three guests OS. Two CentOS and XP
Name The first CentOS linux1.starline.ca The second CentOS centos.starline.ca The third XP xp2.starline.ca[code].....
On the clients guest OS nslookup don't work. It write : timed out; no servers could be reached .What is going on? Why nslookup don't work from clients guest OS?On client machine in the file /etc/resolv.conf have record ameserver 168.135.88.2
I have installed bind with yum install bind bind-chroot.I am having query timeouts due to no ipv6 connectivity.Is there a way to re-compile and not loose the chroot structure?Or is there another way to disable ipv6 lookups?Example of issue resulting in ipv4 query timeout:
30-Dec-2010 17:52:03.226 client x.x.x.x#53593: view internal: query: paypal.com.cms.local IN A + 30-Dec-2010 17:52:03.227 client x.x.x.x#53594: view internal: query: paypal.com.cms.local IN AAAA + 30-Dec-2010 17:52:03.228 client x.x.x.x#53595: view internal: query: paypal.com IN A +
I have updated bind using yum on a Centos 5.3 server, after restarting, I have this error now. Error in named configuration: zone localhost/IN: loaded serial 42 /var/named/mydomain.hosts:20: unknown RR type 'SPF' My version of is : bind.x86_64 30:9.3.6-16.P1.el5
when client goes to bind to NFS share on remote server - they are getting access denied when using the mount command; [SERVER] - CentOS 5.3 /etc/exports /mnt/data 192.168.5.199(rw) - implying the client I want to have access
I am setting up a cluster of servers which use Centos Directory Server for control of logins, etc and kerberos for authentication. The basic setup is working fine, I have been able to manually create accounts using the directory console and these accounts seem to work. Now what I want to do is automate the process of creating new accounts. I am writing a perl script which can be run by one of the server administrators, they supply a small number of arguments and it should create a new user in the directory server, and also create a principal in the kerberos.
I want them to be able to do this using their logged-in kerberos credentials, i.e., without having to enter and re-enter their passwords. My first attempt was to use perl modules Net::LDAP and Authen::SASL. I could not get this working so fell back to using ldap command line tools, but even these I cannot seem to get working! When using mozldap tools, as specified in the admin manual, I get the following:
Using openldap tools I strike exactly the same problem: $ ldapmodify -Y GSSAPI -H LDAP://ldaphost.mycompany.com -D uid=eharmic,ou=mydept,dc=mycompany -U eharmic < ../ldapmod.txt SASL/GSSAPI authentication started ldap_sasl_interactive_bind_s: Invalid credentials (49) additional info: SASL(-14): authorization failure:
I believe I have set up the mapping correctly: dn: cn=MyMapping,cn=mapping,cn=sasl,cn=config objectClass: top objectClass: nsSaslMapping cn: MyMapping nsSaslMapRegexString: ^(.+)@MYCOMPANY.COM nsSaslMapBaseDNTemplate: ou=mydept,dc=mycompany nsSaslMapFilterTemplate: (uid=1)
It must be getting reasonably far because after doing the above I can see the LDAP service ticket in my "klist" output.
How do I know the reply was from caching-nameserver, master nameserver or slave nameserver in BIND DNS.? Can I trace the nameserver that replied my dns query? Like I have a master nameserver, a slave nameserver and a caching nameserver in my network. Is it possible to know that which of these servers replied my dns query? Also can i trace back that replying server exactly?
I am bit new to Linux and have setup caching-only name server with Centos 5.5. when i do dig server, it provide resolutions. but when i use the server IP as DNS on my windows client, it says, "connection refused" on the NSlookup output. (IP table didn't enable) My server Ip is 192.168.1.253 and bellow is the configuration of "/var/named/chroot/etc/named.conf"
I have a fairly simple iSCSI setup using two devices, but they have swapped names on different machines. running CentOs 5.3 ia64, and using iscsi-initiator-utils-6.2.0.868-0.18.el5
vm1: [root@vm1 ~]# fdisk -l Disk /dev/xvda: 4194 MB, 4194304000 bytes 255 heads, 63 sectors/track, 509 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes
[Code].....
Any way to get iSCSI to mount the devices as consistent device names ?
I updated bind to version bind-9.3.4-6.0.3.P1.el5_2 today, and then I started getting "Starting named: named: user 'named' unknown" messages. Selinux is enabled. ID named shows: uid=25(named) gid=25(named) groups=25(named) context=root:system_r:unconfined_t:SystemLow-SystemHigh
