[Dnsmasq-discuss] how to set NS for a zone?

2011-03-04 Thread Carlos Carvalho
I've put dnsmasq as the authoritative server for a zone. All seems to be working fine, except that I didn't find how to answer NS queries. Is there a way? In principle this should be defined by the parent zone but it's strange that the authoritative server for a zone doesn't know about it...

[Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Michael P. McDonnell
So I'd like to use vendor class identifer to create two different ranges in dnsmasq, and I'm not entire sure how to go about that. dhcp-vendorclass=set:custom,Sample Vendor Class dhcp-range=tag:custom,192.168.0.50,192.168.0.150,12h dhcp-range=192.168.0.151,192.168.0.254,12h However - if I take a

Re: [Dnsmasq-discuss] how to set NS for a zone?

2011-03-04 Thread Matthias Andree
Am 04.03.2011 01:01, schrieb Carlos Carvalho: I've put dnsmasq as the authoritative server for a zone. All seems to be working fine, except that I didn't find how to answer NS queries. Is there a way? In principle this should be defined by the parent zone but it's strange that the

Re: [Dnsmasq-discuss] Announce: dnsmasq-2.57

2011-03-04 Thread Helmut Hullen
Hallo, Simon, Du meintest am 18.02.11: dnsmasq 2.57 is now available at http://www.thekelleys.org.uk/dnsmasq/dnsmasq-2.57.tar.gz Binary for Slackware (current/13.x): http://helmut.hullen.de/filebox/Linux/slackware/n/dnsmasq-2.57-i486-1hln.txz Seems to run as expected - thank you! Viele

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
Are you releasing the lease after changing the client configuration, or renewing an existing lease? On Thu, Mar 3, 2011 at 11:31 PM, Michael P. McDonnell bzaks1...@gmail.comwrote: So I'd like to use vendor class identifer to create two different ranges in dnsmasq, and I'm not entire sure how

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Michael P. McDonnell
I'm was just doing a dhclient eth0 on the client. So I guess that would make sense not killing the lease all together... so I tried doing a dhclient -r eth0, then doing a dhclient eth0. No luck :-/ On Fri, Mar 4, 2011 at 10:12, richardvo...@gmail.com richardvo...@gmail.com wrote: Are you

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
Does dnsmasq have dhcp logging enabled? Can you paste the few lines associated with giving the client an address from the wrong block? On Fri, Mar 4, 2011 at 10:15 AM, Michael P. McDonnell bzaks1...@gmail.comwrote: I'm was just doing a dhclient eth0 on the client. So I guess that would make

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Michael P. McDonnell
I understand this isn't EXACTLY the scenario I need to fix for - but I kind of want to learn what I'm doing wrong as well. (This will help me learn more about dhcp in general) Here's my dnsmasq.conf: ###DNSMASQ CONF no-resolv no-poll dhcp-vendorclass=set:custom,Magic Server

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Simon Kelley
Michael P. McDonnell wrote: I understand this isn't EXACTLY the scenario I need to fix for - but I kind of want to learn what I'm doing wrong as well. (This will help me learn more about dhcp in general) Here's my dnsmasq.conf: ###DNSMASQ CONF no-resolv no-poll

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
It looks like the dhcp-range is setting a tag instead of matching it. If you took out tag:custom, from the dhcp-range and restart dnsmasq, would the query still show tags: tag:custom Wait a second. I don't think the tags line is supposed to contain tag:. What version are you running? On

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
The behavior you're seeing exactly matches the old description, when tag: wasn't a keyword. *-F, --dhcp-range=[[net:]network-id,]start-addr,end-addr[[,netmask],broadcast][,default lease time]*Enable the DHCP server. Addresses will be given out from the range start-addr to end-addr and from

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Simon Kelley
richardvo...@gmail.com wrote: The behavior you're seeing exactly matches the old description, when tag: wasn't a keyword. Richard's right: you need to upgrade to version = 2.53, or use the old, confusing tag syntax. I strongly recommend upgrading. Cheers, Simon.

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
This probably should have gone to the list. On Fri, Mar 4, 2011 at 11:01 AM, richardvo...@gmail.com richardvo...@gmail.com wrote: Simon's latest version is 2.57, you want at least 2.53, a snippet from the changelog: version 2.53 Rationalised the DHCP tag system. Every

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Simon Kelley
richardvo...@gmail.com wrote: I'm on version 2.45 - it was the default with SLES. I'll see if I can find something closer to 2.5.x, is there a minimum just in case I find an RPM. If you can't upgrade, let us know: the functionality is in 2.45, but the configuration

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread Simon Kelley
Michael P. McDonnell wrote: I found 2.55 pretty easily. I'm wondering - is it legit to have my multi-word vendor class set in quotes like that? Yes. I released and renewed, and was still in the wrong range, so I'm looking to other potential problems. Did you see my earlier reply about

Re: [Dnsmasq-discuss] Using Vendor Class to determine range

2011-03-04 Thread richardvo...@gmail.com
On Fri, Mar 4, 2011 at 11:16 AM, Simon Kelley si...@thekelleys.org.ukwrote: Michael P. McDonnell wrote: I found 2.55 pretty easily. I'm wondering - is it legit to have my multi-word vendor class set in quotes like that? Yes. I released and renewed, and was still in the wrong range,

Re: [Dnsmasq-discuss] how to set NS for a zone?

2011-03-04 Thread richardvo...@gmail.com
On Fri, Mar 4, 2011 at 11:48 AM, Carlos Carvalho car...@fisica.ufpr.brwrote: richardvo...@gmail.com (richardvo...@gmail.com) wrote on 4 March 2011 10:15: That prevents it from running as a parent zone. As Carlos has said, the parent zone needs to provide the NS record for subzones,

Re: [Dnsmasq-discuss] how to set NS for a zone?

2011-03-04 Thread Matthias Andree
I wrote: Dnsmasq is not supposed to run as an authoritative server - it doesn't support SOA or NS records, for instance. Richard Voigt concluded: That prevents it from running as a parent zone. As Carlos has said, the parent zone needs to provide the NS record for subzones, otherwise

[Dnsmasq-discuss] dnsmasq and unicode hosts

2011-03-04 Thread Pedro Marcolino
Hi, I'm having problems using dnsmasq with unicode hosts. For example, none of the bellow rules will work: address=/www.\u4ffa\u306e\u5bff\u53f8\u304c\u3046\u307e\u3044\u305e.com/172.23.70.3^M address=/www.俺の寿司がうまいぞ.com /172.23.70.2 is this something that is currently, or will be in the future,

Re: [Dnsmasq-discuss] dnsmasq and unicode hosts

2011-03-04 Thread Simon Kelley
On 04/03/11 21:29, Pedro Marcolino wrote: Hi, I'm having problems using dnsmasq with unicode hosts. For example, none of the bellow rules will work: address=/www.\u4ffa\u306e\u5bff\u53f8\u304c\u3046\u307e\u3044\u305e.com/172.23.70.3^M address=/www.俺の寿司がうまいぞ.com /172.23.70.2 is this

Re: [Dnsmasq-discuss] dnsmasq and unicode hosts

2011-03-04 Thread Pedro Marcolino
Hi, Seems like you're right: dnsmasq -v Dnsmasq version 2.45 Copyright (C) 2000-2008 Simon Kelley Compile time options IPv6 GNU-getopt no-ISC-leasefile no-DBus no-I18N TFTP I need at least version 2.51. Thanks for pointing me in the correct direction. Regards, Pedro 2011/3/4 Simon Kelley

Re: [Dnsmasq-discuss] dnsmasq and unicode hosts

2011-03-04 Thread Simon Kelley
Pedro Marcolino wrote: Hi, Seems like you're right: dnsmasq -v Dnsmasq version 2.45 Copyright (C) 2000-2008 Simon Kelley Compile time options IPv6 GNU-getopt no-ISC-leasefile no-DBus no-I18N TFTP ^^^ I need at least