Re: [SLUG] 'named' problems (dns)
quote who=Jamie Wilkinson And you shouldn't need to restart BIND to make any changes, if you have rndc set up correctly then 'rndc reconfig' to reload any changes to named.conf and 'rndc reload domain' to reload the zone without interrupting nameservice. Jamie, I have # rpm -q bind bind-9.2.1-1.7x.2 # rndc status rndc: connect failed: connection refused # rndc stop rndc: connect failed: connection refused # rndc start rndc: connect failed: connection refused # ps -A | grep named 17708 ?00:00:00 named 17710 ?00:00:00 named 17711 ?00:35:45 named 17712 ?00:01:04 named 17716 ?00:03:52 named accoring to Webmin, bind is working what am I missing -- Voytek -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Voytek Eymont wrote: # rndc status rndc: connect failed: connection refused # rndc stop rndc: connect failed: connection refused # rndc start rndc: connect failed: connection refused connection refused is the result of a client trying to connect to a port on which nobody is listening. I don't have bind9 installed on my system so I can't check this directly, but the docs I found via google (http://www.nominum.com/content/documents/bind9arm.pdf) suggest that you have to have an rndc.conf file to tell both named and rndc where to find each other and which shared secret key to use. Also here is a link with a sample file: http://www.tacktech.com/display.cfm?ttid=323 Hope this helps some. --Amos -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
quote who=[EMAIL PROTECTED] Voytek Eymont wrote: connection refused is the result of a client trying to connect to a port on which nobody is listening. Hope this helps some. thanks, Amos, yes, it did I looked up RH docs which does have a similar section on how to set it up, as the BSD docs you pointed out (mea culpa, RTFM again, I assumed that the original setup took care of it...) /just when I almost knew how to use 'ndc' they take 'ndc' away and replace it with 'rndc', oh well/ -- Voytek -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
RE: [SLUG] 'named' problems (dns)
Bingo. 'message' dump = ... ... ... Aug 25 16:32:51 linux01 named[5353]: lame server resolving 'agreserves.com.au' (in 'agreserves.com.au'?): 203.2.192.124#53 Aug 25 16:32:51 linux01 named[5353]: lame server resolving 'agreserves.com.au' (in 'agreserves.com.au'?): 203.2.193.124#53 Aug 25 16:32:51 linux01 named[5353]: lame server resolving 'agreserves.com.au' (in 'agreserves.com.au'?): 203.2.192.124#53 Aug 25 16:32:51 linux01 named[5353]: lame server resolving 'agreserves.com.au' (in 'agreserves.com.au'?): 203.2.193.124#53 Aug 25 16:52:22 linux01 named[5353]: message.c:809: REQUIRE(*rdataset == ((void *)0)) failed Aug 25 16:52:22 linux01 named[5353]: exiting (due to assertion failure) - The full file is 600kb+ dated from Aug 22. - The last few rolled logs (from whenever to Aug 22) are about 16k in size! Those are just the last few lines (obviously!) Does this help? Bind version: bind-9.2.0-8 If it can be fixed without re-installing / upgrading bind or named, then fantastic! =) I sense progress ;) I also looked at the zone files as per Plameras advice. NONE of them seemed to have Tabs.? And from memory, this was always the case. I changed a couple of the more important ones to have tabs anyway... I'll give an update tomorrow, in the meantime, if someone sees something else, or finds something in the log extract above, feel free to post it ;) Ta. Regards, Jared Pritchard The Waterexchange Pty Ltd Ph: (02) 6881 6316 Fax:(02) 6881 6318 -Original Message- From: David Gillies [mailto:[EMAIL PROTECTED] Sent: Wednesday, 25 August 2004 1:54 PM To: Jared Pritchard Cc: [EMAIL PROTECTED] Subject: Re: [SLUG] 'named' problems (dns) Jared Pritchard wrote: I assume just by running it...? When I do it reports nothing. =) I'm not confident/experienced enough to make 'major' changes like updating named, because this server is very critical to our livelihood. I can't risk taking it offline. =( on my server, named logs all its output to /var/log/messages If you run 'grep named /var/log/messages', what does it come back with? Any other thoughts as to why it just suddenly started doing this without any apparent reason, or anything I can do to fix it? Also, I can't manage to figure out how to find its distribution. It's not in /etc/issue, and wasn't installed as a package (ie. rpm doesn't list it) try running rpm -q bind. What does it come back with? The dns server package under Redhat linux is called bind. It must be some sort of configuration error because it still does it after a reboot (mind you, I haven't tried a cold boot) - but we made If a warm reboot doesn't work, then a cold boot isn't going to do anything for you under linux unless it was specifically a hardware issue. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
This one time, at band camp, Jared Pritchard wrote: I assume just by running it...? When I do it reports nothing. =) It checks the config in (by default) /etc/named.conf and reports errors. If that returned nothing then there's nothing wrong with the configuration. The other thing to try (based on the assertion failure in your other mail) is to use named-checkzone on the zone files to see if there's an error there. named-checkzone domain path/to/zonefile -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
This one time, at band camp, Jared Pritchard wrote: Aug 25 16:32:51 linux01 named[5353]: lame server resolving 'agreserves.com.au' (in 'agreserves.com.au'?): 203.2.192.124#53 These are mostly harmless and can be ignored, if you want to chuck them out of the logs totally then the following logging { category lame-servers { null; }; }; in named.conf will hush it up. Aug 25 16:52:22 linux01 named[5353]: message.c:809: REQUIRE(*rdataset == ((void *)0)) failed Aug 25 16:52:22 linux01 named[5353]: exiting (due to assertion failure) I think you have a bad zonefile somewhere. And you shouldn't need to restart BIND to make any changes, if you have rndc set up correctly then 'rndc reconfig' to reload any changes to named.conf and 'rndc reload domain' to reload the zone without interrupting nameservice. -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Jamie Wilkinson wrote: This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. When did this specs change. With BIND 4 this is the case, namely: that delimiters on zone files must be TABS and not SPACE. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
On Wed, 25 Aug 2004, O Plameras wrote: Jamie Wilkinson wrote: This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. When did this specs change. With BIND 4 this is the case, namely: that delimiters on zone files must be TABS and not SPACE. I've never used tabs I've been using BIND since 4. Maybe I was lucky :) -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
I used what I'm confident is confirmed knowledge, namely: Delimiter SPECS on zone files are 'TABS'. This is the case with at least BIND4. I stayed with that SPECS since then for as long as the SPECS work for me. I have O'Reilly BIND Book for BIND 4. I know there are latter editions. But I have no need for them now. SPACE delimiters worked for you; stay with it. TABS work for me. I'd be grateful if you can point to me an RFC, Book, or Online Internet Document to say about enhancements to this specs. TABS are still OK because it works for me. But I do not rely on luck when it comes to operating computers and networks. I have to satisfy myself that it is part of the SPECS. That's the rule I follow. David wrote: On Wed, 25 Aug 2004, O Plameras wrote: Jamie Wilkinson wrote: This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. When did this specs change. With BIND 4 this is the case, namely: that delimiters on zone files must be TABS and not SPACE. I've never used tabs I've been using BIND since 4. Maybe I was lucky :) -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Thanks for both copies of your email. This one time, at band camp, O Plameras wrote: Jamie Wilkinson wrote: This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. When did this specs change. With BIND 4 this is the case, namely: that delimiters on zone files must be TABS and not SPACE. BIND 9, and I'm fairly certain my RHEL 2.1 boxes use BIND 8 which also have numerous spaces sprinkled liberally throughout them. I find your use of BIND 4 in this modern age extraordinary! -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
I thank you again for both copies of your email. This one time, at band camp, O Plameras wrote: TABS are still OK because it works for me. But I do not rely on luck when it comes to operating computers and networks. I have to satisfy myself that it is part of the SPECS. That's the rule I follow. Exciting developments in the combined fields of DOMAIN NAME SERVICE and CONFIGURATION MANAGEMENT have led to two tools which I described elsewhere in this thread -- named-checkconf and named-checkzone. -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Jamie Wilkinson wrote: Thanks for both copies of your email. This one time, at band camp, O Plameras wrote: Jamie Wilkinson wrote: This one time, at band camp, O Plameras wrote: Obviously, when you find spaces, replace with TABS. That's false. BIND copes just fine with spaces or tabs. When did this specs change. With BIND 4 this is the case, namely: that delimiters on zone files must be TABS and not SPACE. BIND 9, and I'm fairly certain my RHEL 2.1 boxes use BIND 8 which also have numerous spaces sprinkled liberally throughout them. I find your use of BIND 4 in this modern age extraordinary! The SPECS I'm familiar with says at least one (1) TAB as delimiter. I have BIND 9 on Fedora and I'll experiment. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Jamie Wilkinson wrote: BIND 9, and I'm fairly certain my RHEL 2.1 boxes use BIND 8 which also have numerous spaces sprinkled liberally throughout them. I find your use of BIND 4 in this modern age extraordinary! That's the DNS software I have in my piles of DISKETTES. I am am using a 386 Computer with 128K Memory(not Pentium and not 128Meg) for my DNS running the old version of Slackware. This is a hangover from back over 10 years. I have to chuck it out soon, I suppose. And I just copy the zone files from there for my recent installations and modify a bit. I had a big problem one day with my DNS; and I learned about this SPECS the hard way when everything seems OK but my mail servers will not allow customers to check their emails. Imagine to have about five thousand customers trying to connect to their mail and can't. The phone rings hot and 'abuses' flies one way, to my directions. This was back in 1996. I know that SPEC works because it worked well for me since then. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
BIND 9, and I'm fairly certain my RHEL 2.1 boxes use BIND 8 which also have numerous spaces sprinkled liberally throughout them. I find your use of BIND 4 in this modern age extraordinary! The SPECS I'm familiar with says at least one (1) TAB as delimiter. I have BIND 9 on Fedora and I'll experiment. I copied my zone files from my server and changed all TABS and restarted my named. It works just as well. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
RE: [SLUG] 'named' problems (dns)
I assume just by running it...? When I do it reports nothing. =) I'm not confident/experienced enough to make 'major' changes like updating named, because this server is very critical to our livelihood. I can't risk taking it offline. =( Any other thoughts as to why it just suddenly started doing this without any apparent reason, or anything I can do to fix it? Also, I can't manage to figure out how to find its distribution. It's not in /etc/issue, and wasn't installed as a package (ie. rpm doesn't list it) It must be some sort of configuration error because it still does it after a reboot (mind you, I haven't tried a cold boot) - but we made absolutely no changes that could affect that (I'm the only one that touches the server, and all I do is code web pages, PERL MySQL etc.) Could it be a DoS attack or something!? =/ All help is appreciated!! =) Cheers, Jared Pritchard The Waterexchange Pty Ltd Ph: (02) 6881 6316 Fax:(02) 6881 6318 -Original Message- From: Jamie Wilkinson [mailto:[EMAIL PROTECTED] Sent: Tuesday, 24 August 2004 3:02 PM To: [EMAIL PROTECTED] Subject: Re: [SLUG] 'named' problems (dns) This one time, at band camp, Jared Pritchard wrote: Thanks for any help. Maybe see what named-checkconf tells you, if you're running a recent BIND. -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
What Linux are you using? I had RedHat 9, and I got a this problem where each time I restarted 'named', i.e., service named restart I got Sopping Named Failed. The reason was that /var/run/named/named.pid was not being created because /var/run/named was owned by 'root' and 'service named restart' sets user to 'named'. I got that fixed by, chown -R named.named /var/run/named But this is not the reason why your 'named' daemon was failing,of course. One possibility why your named is failing is because there is 'format' or syntax error in one or more lines in your named database. When that line gets referenced during a process that's when your DNS server gives up. In your zone files, fields must be separated by TABS or ^I (not spaces). Because of this the format or syntax error is not obvious to the eye under normal settings. For example, I have a line, server IN A 203.7.132.57 when I load a file with 'vi file'. The raw format looks like this, server^I^IIN^I^IA^I^I203.7.132.57 where ^I represents a TAB-Character. Now you can confirm that you have tabs separating your fields by using 'vi filename' and then issue this colon-command, :set list To restore, issue this colon-command, :set nolist Obviously, when you find spaces, replace with TABS. Jared Pritchard wrote: Hi, Just experiencing some crazy problems with named (dns). Our web server (which is also our gateway to the Internet for our office machines) serves as a DNS and sporadically over the last couple of days, for no known reason, named keeps failing and every now and then (between 15 minutes to a few hours) it will just, stop. I restart it using service named restart and it says that the shutdown of named failed, but startup is OK. Originally (not knowing what the problem was) I reset our web server, so its not just some bug in startup, and we have made NO changes that should effect its workings at all. Just weird. =/ Ive looked through the cronjobs theres nothing that even comes close to the failure times, so I cant see what the problem might be. Has anyone experienced this sort of thing before, or have any idea what it might be? When named fails, it simply wont let us view our website, or any other site on the net using domain names. IPs work fine. Using Redhat 7.3 I think Here is a quick dump of rndc (when named is working unfortunately, havent been here for last few failures) * [root]# rndc status number of zones: 28 debug level: 0 xfers running: 0 xfers deferred: 0 soa queries in progress: 0 query logging is OFF server is up and running * I have no idea how thats going to help, but who knows? Not me = ) Thanks for any help. * Cheers, * Jared Pritchard The Waterexchange Pty Ltd -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
Jared Pritchard wrote: I assume just by running it...? When I do it reports nothing. =) I'm not confident/experienced enough to make 'major' changes like updating named, because this server is very critical to our livelihood. I can't risk taking it offline. =( on my server, named logs all its output to /var/log/messages If you run 'grep named /var/log/messages', what does it come back with? Any other thoughts as to why it just suddenly started doing this without any apparent reason, or anything I can do to fix it? Also, I can't manage to figure out how to find its distribution. It's not in /etc/issue, and wasn't installed as a package (ie. rpm doesn't list it) try running rpm -q bind. What does it come back with? The dns server package under Redhat linux is called bind. It must be some sort of configuration error because it still does it after a reboot (mind you, I haven't tried a cold boot) - but we made If a warm reboot doesn't work, then a cold boot isn't going to do anything for you under linux unless it was specifically a hardware issue. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] 'named' problems (dns)
Hi, Just experiencing some crazy problems with named (dns). Our web server (which is also our gateway to the Internet for our office machines) serves as a DNS and sporadically over the last couple of days, for no known reason, named keeps failing and every now and then (between 15 minutes to a few hours) it will just, stop. I restart it using service named restart and it says that the shutdown of named failed, but startup is OK. Originally (not knowing what the problem was) I reset our web server, so its not just some bug in startup, and we have made NO changes that should effect its workings at all. Just weird. =/ Ive looked through the cronjobs theres nothing that even comes close to the failure times, so I cant see what the problem might be. Has anyone experienced this sort of thing before, or have any idea what it might be? When named fails, it simply wont let us view our website, or any other site on the net using domain names. IPs work fine. Using Redhat 7.3 I think Here is a quick dump of rndc (when named is working unfortunately, havent been here for last few failures) * [root]# rndc status number of zones: 28 debug level: 0 xfers running: 0 xfers deferred: 0 soa queries in progress: 0 query logging is OFF server is up and running * I have no idea how thats going to help, but who knows? Not me = ) Thanks for any help. Cheers, Jared Pritchard The Waterexchange Pty Ltd -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
On 24/08/2004, at 2:52 PM, Jared Pritchard wrote: Just experiencing some crazy problems with named (dns). Our web server (which is also our gateway to the Internet for our office machines) serves as a DNS and sporadically over the last couple of days, for no known reason, named keeps failing and every now and then (between 15 minutes to a few hours) it will just, stop. I restart it using service named restart and it says that the shutdown of named failed, but startup is OK. Originally (not knowing what the problem was) I reset our web server, so its not just some bug in startup, and we have made NO changes that should effect its workings at all. Just weird. =/ Ive looked through the cronjobs theres nothing that even comes close to the failure times, so I cant see what the problem might be. Has anyone experienced this sort of thing before, or have any idea what it might be? Be useful to see some logs from the server. /var/log/daemon.log is a good place to start, depending on your syslog config. Sounds to me like bind (named) is dying. -- Tony Green [EMAIL PROTECTED] -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
This one time, at band camp, Jared Pritchard wrote: Thanks for any help. Maybe see what named-checkconf tells you, if you're running a recent BIND. -- [EMAIL PROTECTED] http://spacepants.org/jaq.gpg -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] 'named' problems (dns)
On Tue, 2004-08-24 at 14:52, Jared Pritchard wrote: st1\:*{behavior:url(#default#ieooui) } -- -- /* Style Definitions */ table.MsoNormalTable {mso-style-name:Table Normal; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:Times New Roman;} -- Hi, Just experiencing some crazy problems with named (dns). Our web server (which is also our gateway to the Internet for our office machines) serves as a DNS and sporadically over the last couple of days, for no known reason, named keeps failing and every now and then (between 15 minutes to a few hours) it will just, stop. I restart it using service named restart and it says that the shutdown of named failed, but startup is OK. I queried this yesterday on another list. When I did an A query it would come back as the root DNS servers being the authority. I gather there was some really screwed routing problems somewhere, perhaps with our favourite telco. Seems like they might be continuing. Originally (not knowing what the problem was) I reset our web server, so its not just some bug in startup, and we have made NO changes that should effect its workings at all. Just weird. =/ Ive looked through the cronjobs theres nothing that even comes close to the failure times, so I cant see what the problem might be. Has anyone experienced this sort of thing before, or have any idea what it might be? When named fails, it simply wont let us view our website, or any other site on the net using domain names. IPswork fine. Using Redhat 7.3 I think Here is a quick dump of rndc (when named is working unfortunately, havent been here for last few failures) * [root]# rndc status number of zones: 28 debug level: 0 xfers running: 0 xfers deferred: 0 soa queries in progress: 0 query logging is OFF server is up and running * I have no idea how thats going to help, but who knows? Not me = ) Thanks for any help. Cheers, Jared Pritchard The Waterexchange Pty Ltd __ -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html -- Howard. LANNet Computing Associates; Your Linux people http://www.lannetlinux.com -- When you just want a system that works, you choose Linux; when you want a system that just works, you choose Microsoft. -- Flatter government, not fatter government; Get rid of the Australian states. -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html