dnsjava-users — The dnsjava mailing list

Re: DNSSEC

[Brian W. <bwe...@xb...>]

On Nov 22, 2010, at 5:04 AM, mod63 wrote:

> 
> Hi there,
> 
> I’m currently working on a project that requires DNSSEC, the current
> situation is basically that I get a request from a client and based on that
> I generate a response i.e. Message object and send it of to the client, very
> basic stuff, the problem is I need to add DNSSEC to that, and that’s where
> I’m currently stuck.
> 
> I couldn’t find any good resources about the subject and the explanation in
> the dnsjava documentation wasn’t adequate
> 
> Any help would be appreciated; I need to resolve this problem as quickly as
> possible.

If you want to write a DNSSEC-compliant name server, there is no resource better than the DNSSEC RFCs.  There are a whole lot of them, but the important ones are probably 4034 and 4035.

You're not going to find much in the dnsjava documentation, as dnsjava doesn't include a DNSSEC-compliant name server.  I think the library is complete enough that one could be written, but it definitely would not be trivial, and would likely take a considerable amount of time.

Brian

DNSSEC

[mod63 <mod...@gm...>]

Hi there,

I’m currently working on a project that requires DNSSEC, the current
situation is basically that I get a request from a client and based on that
I generate a response i.e. Message object and send it of to the client, very
basic stuff, the problem is I need to add DNSSEC to that, and that’s where
I’m currently stuck.

I couldn’t find any good resources about the subject and the explanation in
the dnsjava documentation wasn’t adequate

Any help would be appreciated; I need to resolve this problem as quickly as
possible.

Thanks.


-- 
View this message in context: http://old.nabble.com/DNSSEC-tp30278312p30278312.html
Sent from the dnsjava-users mailing list archive at Nabble.com.

Re: DNS Dynamic Update problem

[Brian W. <bwe...@xb...>]

On Nov 3, 2010, at 1:51 PM, Rafael Diniz wrote:

> As i have read, problems due to the code must be sent through this list, so, here we go! :)
> 
> Well, using the Eclipse Galileo for windows over here.
> I´m coding a small void main trying to test the DNS Dynamic update example from the oficial website.
> 
> My idea is that when my DNS fails this small code could use another DNS so that my application keep working properly.
> 
> Resume:
> If the default DNS fails, the dynamic update could bring me, in some sort of a real time, another DNS that works (8.8.8.8. google dns maybe).

I think you're confused.

There is no concept of "the default DNS" when sending dynamic updates; you need to specify where the updates are sent.  If you specify a server that doesn't accept dynamic updates, it should fail, and the root servers do not (and should not) accept dynamic updates.

There's no way to fall back to another server which would accept them, because there aren't any DNS servers that accept dynamic updates from anyone.  The Google servers certainly don't.

I don't know why you believe that any DNS server on the Internet would trust you enough to let you make changes to its data.  If you want to send dynamic updates, you need to either set up your own DNS server and zone that you're permitted to make updates to, or have the administrator of some other server and zone grant you update privileges.

Brian

> this is the code:
> 
> 
> 	public static void DNSDynamicUpdate() throws Exception
> 	{
> 		System.out.println("\n" + "DNSDynamicUpdate" + "\n");
> 		String selfHost = "192.168.0.156";
> 		String hostIP = "128.63.2.53";//one of the root dns  //google dns 8.8.8.8
> 		String gateway = "192.168.0.254";//my wireless gateway
> 		
> 		if(QueryRemoteserver(hostIP))
> 		{
> 			Name zone = Name.fromString("root-servers.net.");
> 			Name host = Name.fromString("h", zone);
> 			Update update = new Update(zone);
> 			update.replace(host, Type.A, 3600, gateway);
> 	
> 			Resolver res = new SimpleResolver(hostIP);
> 			//com md5
> 			res.setTSIGKey(new TSIG(host, base64.fromString("1234")));
> 			//res.setPort(53);
> 			res.setTCP(true);
> 	
> 			Message response = res.send(update);
> 	
> 			System.out.println("\n" + response + "\n");
> 		}
> 	}
> 
> 	public static boolean QueryRemoteserver(String hostIP) throws Exception
> 	{
> 		if(hostIP == null)
> 			hostIP = "128.63.2.53";
> 		System.out.println("\n" + "QueryRemoteserver" + "\n");
> 		Lookup l = new Lookup("version.bind.", Type.TXT, DClass.CH);
> 		l.setResolver(new SimpleResolver());
> 		l.run();
> 		if (l.getResult() == Lookup.SUCCESSFUL)
> 		{
> 			System.out.println(l.getAnswers()[0].rdataToString());
> 			return true;
> 		}
> 		else
> 		{
> 			System.out.println(l.getAnswers()[0].rdataToString());
> 			return false;
> 		}
> 		
> 	}
> response:
> ;; ->>HEADER<<- opcode: UPDATE, status: NOTIMP, id: 48592
> ;; flags: qr ; qd: 0 an: 0 au: 0 ad: 0 
> ;; TSIG invalid
> ;; ZONE:
> 
> ;; PREREQUISITES:
> 
> ;; UPDATE RECORDS:
> 
> ;; ADDITIONAL RECORDS:
> 
> ;; Message size: 12 bytes
> 
> 
> 
> Thanks in advance...
> 
> PS: NOTAUTH was the most received message in my "testes"
> ------------------------------------------------------------------------------
> The Next 800 Companies to Lead America's Growth: New Video Whitepaper
> David G. Thomson, author of the best-selling book "Blueprint to a 
> Billion" shares his insights and actions to help propel your 
> business during the next growth cycle. Listen Now!
> http://p.sf.net/sfu/SAP-dev2dev_______________________________________________
> dnsjava-users mailing list
> dns...@li...
> https://lists.sourceforge.net/lists/listinfo/dnsjava-users

DNS Dynamic Update problem

[Rafael D. <raf...@ki...>]

As i have read, problems due to the code must be sent through this list, so,
here we go! :)

Well, using the Eclipse Galileo for windows over here.
I´m coding a small void main trying to test the DNS Dynamic update example
from the oficial website.

My idea is that when my DNS fails this small code could use another DNS so
that my application keep working properly.

Resume:
If the default DNS fails, the dynamic update could bring me, in some sort of
a real time, another DNS that works (8.8.8.8. google dns maybe).



this is the code:


public static void DNSDynamicUpdate() throws Exception {
System.out.println("\n" + "DNSDynamicUpdate" + "\n"); String selfHost =
"192.168.0.156"; String hostIP = "128.63.2.53";//one of the root dns
//google dns 8.8.8.8 String gateway = "192.168.0.254";//my wireless gateway
if(QueryRemoteserver(hostIP)) { Name zone = Name.fromString("
root-servers.net."); Name host = Name.fromString("h", zone); Update update =
new Update(zone); update.replace(host, Type.A, 3600, gateway); Resolver res
= new SimpleResolver(hostIP); //com md5 res.setTSIGKey(new TSIG(host,
base64.fromString("1234"))); //res.setPort(53); res.setTCP(true); Message
response = res.send(update); System.out.println("\n" + response + "\n"); } }
public static boolean QueryRemoteserver(String hostIP) throws Exception {
if(hostIP == null) hostIP = "128.63.2.53"; System.out.println("\n" +
"QueryRemoteserver" + "\n"); Lookup l = new Lookup("version.bind.",
Type.TXT, DClass.CH); l.setResolver(new SimpleResolver()); l.run(); if
(l.getResult() == Lookup.SUCCESSFUL) {
System.out.println(l.getAnswers()[0].rdataToString()); return true; } else {
System.out.println(l.getAnswers()[0].rdataToString()); return false; } }
response:
;; ->>HEADER<<- opcode: UPDATE, status: NOTIMP, id: 48592
;; flags: qr ; qd: 0 an: 0 au: 0 ad: 0
;; TSIG invalid
;; ZONE:

;; PREREQUISITES:

;; UPDATE RECORDS:

;; ADDITIONAL RECORDS:

;; Message size: 12 bytes



Thanks in advance...

PS: NOTAUTH was the most received message in my "testes"

Re: Cache miss statistics

[Brian W. <bwe...@xb...>]

On Nov 1, 2010, at 2:46 AM, Norman Maurer wrote:

> is it possible to retrieve some statistics about how many requests
> could get handled via the Cache and how many need to get lookup ? We
> at JAMES currently build JMX stuff around our DNSService which use
> DNSJava and would like to show some statistics here.

No, there are no such statistics tracked.

Brian

Cache miss statistics

[Norman M. <no...@ap...>]

Hi there,

is it possible to retrieve some statistics about how many requests
could get handled via the Cache and how many need to get lookup ? We
at JAMES currently build JMX stuff around our DNSService which use
DNSJava and would like to show some statistics here.

Thx,
Norman

Re: dnsjava IXFR fails if the serial number has rolled over on the server

[Brian W. <bwe...@xb...>]

On Sep 9, 2010, at 4:30 AM, <Ale...@us...> <Ale...@us...> wrote:

> I am seeing an issue where an IXFR using dnsjava's IXFR client will fail if the serial number on the server has rolled over since the last serial on the client.
> 
> So for example, if I am running a DNS server with a serial number for example.com of 4294967295 and then send it an update so that the serial number rolls over to 1. Then I call dnsjava's:
> 
> ZoneTransferIn transfer = ZoneTransferIn.newIXFR( new Name( "example.com" ), 4294967295, true, "127.0.0.1", null);
> transfer.run();
> 
> the transfer fails.
> 
> If I modify ZoneTransferIn.java and change the line in the parseRR method from:
> 
> if (qtype == Type.IXFR && end_serial <= ixfr_serial) {
> 
> to instead be:
> 
> if (qtype == Type.IXFR && (Serial.compare(end_serial, ixfr_serial) <= 0 )) {
> 
> then everything works great and the IXFR completes successfully. Note that I ran into this with the 1.6.4 release, but I did take a look at the code in 2.1.0 and the bug seems to still be there. Let me know if that appears to be a reasonable fix.

This fix looks correct.  I'll commit it now, but I'm not sure when the next release will be.

Thanks!

Brian

dnsjava IXFR fails if the serial number has rolled over on the server

[<Ale...@us...>]

I am seeing an issue where an IXFR using dnsjava's IXFR client will fail if the serial number on the server has rolled over since the last serial on the client.

So for example, if I am running a DNS server with a serial number for example.com of 4294967295 and then send it an update so that the serial number rolls over to 1. Then I call dnsjava's:
 
ZoneTransferIn transfer = ZoneTransferIn.newIXFR( new Name( "example.com" ), 4294967295, true, "127.0.0.1", null);
transfer.run();

the transfer fails.

If I modify ZoneTransferIn.java and change the line in the parseRR method from:

if (qtype == Type.IXFR && end_serial <= ixfr_serial) {

to instead be:

if (qtype == Type.IXFR && (Serial.compare(end_serial, ixfr_serial) <= 0 )) {

then everything works great and the IXFR completes successfully. Note that I ran into this with the 1.6.4 release, but I did take a look at the code in 2.1.0 and the bug seems to still be there. Let me know if that appears to be a reasonable fix.

Thanks.

---Alexis MacFarlane

Re: Patch for two issues

[Brian W. <bwe...@xb...>]

On Jul 22, 2010, at 10:30 AM, Kocik, William wrote:

> In this case, I’ve clearly misunderstood the purpose of sendAsync(). I was using it to send NOTIFY messages to a group of slave name servers, and when it was discovered that it was only sending to the first name server on the list, this was the “fix” I implemented.
> 
> Is there some built-in mechanism to send a message to a bunch of name servers (synchronously, now that I think of it) or should I thread that on my own?

There is no method for sending a message to multiple servers, and I'm not sure how that would even work in a general way.  When dealing with multiple messages, there are a number of different ways to handle errors, and the correct one will generally be application-specific.

Brian

> On 7/22/10 1:18 PM, "Brian Wellington" <bwe...@xb...> wrote:
> 
> ExtendedResolver.java – sendAsync() was not using its multiple resolvers to send a message, only the first one on the array.
> This is the intended behavior.  Multiple requests will be sent if necessary (that is, if the first request fails or times out), but in the normal case, a single request sent using the first resolver (or a random resolver if setLoadBalance() has been called) is sufficient.
> 
> Brian
> 
> 
> 
> 
> 
> -- 
> Bill 
> 

Re: Patch for two issues

[Kocik, W. <wk...@ve...>]

In this case, I¹ve clearly misunderstood the purpose of sendAsync(). I was
using it to send NOTIFY messages to a group of slave name servers, and when
it was discovered that it was only sending to the first name server on the
list, this was the ³fix² I implemented.

Is there some built-in mechanism to send a message to a bunch of name
servers (synchronously, now that I think of it) or should I thread that on
my own?

> On 7/22/10 1:18 PM, "Brian Wellington" <bwe...@xb...> wrote:
> 
>> 1. ExtendedResolver.java ­ sendAsync() was not using its multiple resolvers
>> to send a message, only the first one on the array.
> This is the intended behavior.  Multiple requests will be sent if necessary
> (that is, if the first request fails or times out), but in the normal case, a
> single request sent using the first resolver (or a random resolver if
> setLoadBalance() has been called) is sufficient.
> 
> Brian
> 




-- 
Bill 

Re: Patch for two issues

[Brian W. <bwe...@xb...>]

On Jul 22, 2010, at 9:12 AM, Kocik, William wrote:

> Hi -
> 
> In the course of working with DNSJava I’ve encountered two issues which are addressed by the patch below my signature:
> 
> LOCRecord.java – a LOC record with a negative altitude would cause a NumberFormatException from parseFixedPoint() when reading a zone file, because it used a regular expression that did not account for the potential presence of a ‘-’ character at the beginning of the number
This is a problem.

I modified the regex in your patch to use "-?" instead of "-{0,1}" (as they're identical), and fixed it to properly handle negative non-integers (the fractional part needs to be negated, or the result is off by one), and committed.
> ExtendedResolver.java – sendAsync() was not using its multiple resolvers to send a message, only the first one on the array.
This is the intended behavior.  Multiple requests will be sent if necessary (that is, if the first request fails or times out), but in the normal case, a single request sent using the first resolver (or a random resolver if setLoadBalance() has been called) is sufficient.

Brian

Patch for two issues

[Kocik, W. <wk...@ve...>]

Hi -

In the course of working with DNSJava I¹ve encountered two issues which are
addressed by the patch below my signature:

1. LOCRecord.java ­ a LOC record with a negative altitude would cause a
NumberFormatException from parseFixedPoint() when reading a zone file,
because it used a regular expression that did not account for the potential
presence of a Œ-¹ character at the beginning of the number
2. ExtendedResolver.java ­ sendAsync() was not using its multiple resolvers
to send a message, only the first one on the array.

Hopefully my fixes are acceptable, but if they aren¹t I¹ll gladly help find
more appropriate solutions.

-- 
Bill 


Index: org/xbill/DNS/LOCRecord.java
===================================================================
--- org/xbill/DNS/LOCRecord.java    (revision 205)
+++ org/xbill/DNS/LOCRecord.java    (revision 207)
@@ -76,9 +76,9 @@
 private double
 parseFixedPoint(String s)
 {
-    if (s.matches("^\\d+$"))
+    if (s.matches("^-{0,1}\\d+$"))
         return Integer.parseInt(s);
-    else if (s.matches("^\\d+\\.\\d*$")) {
+    else if (s.matches("^-{0,1}\\d+\\.\\d*$")) {
         String [] parts = s.split("\\.");
         double value = Integer.parseInt(parts[0]);
         double fraction = Integer.parseInt(parts[1]);
Index: org/xbill/DNS/ExtendedResolver.java
===================================================================
--- org/xbill/DNS/ExtendedResolver.java    (revision 205)
+++ org/xbill/DNS/ExtendedResolver.java    (revision 207)
@@ -132,7 +132,9 @@
     public void
     startAsync(ResolverListener listener) {
         this.listener = listener;
-        send(0);
+                for(int i = 0;i < this.resolvers.length;i++) {
+                    send(i);
+                }
     }
 
     /*

Re: Problem transferring large zones

[Kocik, W. <wk...@ve...>]

Thanks for your fast response, Brian. I think you¹re right, my understanding
of the problem was incorrect. What you describe matches the behavior we saw.

I applied  your changes and it¹s working beautifully. I¹m glad I asked. :)

Speaking of which, I have a patch to submit that addresses some other
issues; I¹ll put that in another thread.


-- 
Bill 

Re: Problem transferring large zones

[Brian W. <bwe...@xb...>]

On Jul 21, 2010, at 3:20 PM, Kocik, William wrote:

> I’ve discovered that when transferring in a zone that is too large to fit in a single message and using a TSIG key to secure it, the transfer fails with “last message must be signed.” This is because the TSIG.StreamVerifier correctly calls key.verify() on the first and every 100th (at least) message’s signatures, but doesn’t necessarily do this with the last message, so when ZoneTransferIn.doxfr() notices this is the last message and checks response.isVerified(), it fails. The trouble is that StreamVerifier can’t possibly know when it’s operating on the last message to know that it must call key.verify(), and by the time doxfr() figures it out it’s too late.
> 
> I’m willing to make the code fix, but I’ve been fiddling with this all afternoon and have yet to come up with a workable solution (other than parsing all the records before checking the TSIG record, so that state == END, and then telling StreamVerifier.verify() that it must verify this record via an added boolean parameter). I’m looking for ideas...

It's possible that I'm missing something, but I think the problem is different than what you're describing.  It looks like this change, added in 2.0.6:

        - The TSIG verification routines (TSIG.verify,
          TSIG.StreamVerifier.verify() now update the Message object with the
          status of the verification in addition to returning the status.

missed several places in StreamVerifier, and the state is never being updated for messages other than the first in the stream.  The released code is unable to complete any TSIG-signed zone transfer longer than a single message, regardless of whether or not every message is signed.

I just committed a fix for this, and my test program passes now.  Could you test the current (svn) code and see if it works now?

I don't think what you described is an issue - what should happen is that StreamVerifier verifies any message that contains a TSIG, whether the TSIG is required or not.  It checks to make sure that there haven't been more than 100 unsigned messages in a row, but it doesn't (and can't) skip any signed messages.

Brian

Problem transferring large zones

[Kocik, W. <wk...@ve...>]

Hi -

I¹ve discovered that when transferring in a zone that is too large to fit in
a single message and using a TSIG key to secure it, the transfer fails with
³last message must be signed.² This is because the TSIG.StreamVerifier
correctly calls key.verify() on the first and every 100th (at least)
message¹s signatures, but doesn¹t necessarily do this with the last message,
so when ZoneTransferIn.doxfr() notices this is the last message and checks
response.isVerified(), it fails. The trouble is that StreamVerifier can¹t
possibly know when it¹s operating on the last message to know that it must
call key.verify(), and by the time doxfr() figures it out it¹s too late.

I¹m willing to make the code fix, but I¹ve been fiddling with this all
afternoon and have yet to come up with a workable solution (other than
parsing all the records before checking the TSIG record, so that state ==
END, and then telling StreamVerifier.verify() that it must verify this
record via an added boolean parameter). I¹m looking for ideas...


-- 
Bill 

Re: update

[Jihane <jbe...@ho...>]

Hello Brian, 
 
Thanks for responding to my email. Now, i need to add a resource record to
the dnsjava server. the jnamed server is running on my computer. To
accomplish that, i have developed the following code: 
 
********************************************************************
 

System.out.println(AAAAsyn);
Update up_add =new Update(Name.fromString("movie.edu.")); // zone
up_add.add(AAAAsyn);
Resolver res = new SimpleResolver("10.192.20.6"); // the IP address of my
computer
res.setTSIGKey(new TSIG(Name.fromString("xbill.org."),
base64.fromString("1234")));// xbill.org. is the name of the key
res.setTCP(true);
Message response =res.send(up_add);
System.out.println(response);


**********************************************************************
 
the result is : 
 

robocop.movie.edu. 79210 IN 2001:db8:85a3:8d3:1319:8a2e:370:7334
;; ->>HEADER<<- opcode: UPDATE, status: NOTIMP, id: 61392
;; flags: ; qd: 0 an: 0 au: 0 ad: 0 
;; TSIG invalid
;; ZONE:
;; PREREQUISITES:
;; UPDATE RECORDS:
;; ADDITIONAL RECORDS:
;; Message size: 12 bytes
 
As you ca see, the response message is empty. 
 
Have you a solution for this problem? 
 
and when i add an RR in the server. It can appears in .db file referenced in
jnamed.conf?
 
Thanks in advance for your help, 
 
cheers, 
 
Jihane


Brian Wellington wrote:
> 
> On Mar 29, 2010, at 12:55 AM, Jihane wrote:
> 
>> 
>> Hello Brian, 
>> 
>> Thank you for taking time to respond to my email. Yes it works for AAAA
>> record. But, it doesn't work for the new type. I have implemented a class
>> (
>> new type) which extends Record class and added the initialization to the
>> type class as you said. when i have run jnamed class i got the following
>> error: 
>> 
>> org.xbill.DNS.Tokenizer$TokenizerException: movie.db:3: invalid unknown
>> RR
>> encoding
>> 
>> Have you any idea to solve this problem? 
> 
> You probably missed adding the type initializer in the Type class.  If
> not, there's not much I can tell you, since I don't have your code.
> 
> Brian
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> dnsjava-users mailing list
> dns...@li...
> https://lists.sourceforge.net/lists/listinfo/dnsjava-users
> 
> 

-- 
View this message in context: http://old.nabble.com/Developping-a-new-record-type-in-dnsjava-tp28040758p28085293.html
Sent from the dnsjava-users mailing list archive at Nabble.com.

Re: Developping a new record type in dnsjava

[Brian W. <bwe...@xb...>]

On Mar 29, 2010, at 12:55 AM, Jihane wrote:

> 
> Hello Brian, 
> 
> Thank you for taking time to respond to my email. Yes it works for AAAA
> record. But, it doesn't work for the new type. I have implemented a class (
> new type) which extends Record class and added the initialization to the
> type class as you said. when i have run jnamed class i got the following
> error: 
> 
> org.xbill.DNS.Tokenizer$TokenizerException: movie.db:3: invalid unknown RR
> encoding
> 
> Have you any idea to solve this problem? 

You probably missed adding the type initializer in the Type class.  If not, there's not much I can tell you, since I don't have your code.

Brian

Re: Developping a new record type in dnsjava

[Jihane <jbe...@ho...>]

Hello Brian, 

Thank you for taking time to respond to my email. Yes it works for AAAA
record. But, it doesn't work for the new type. I have implemented a class (
new type) which extends Record class and added the initialization to the
type class as you said. when i have run jnamed class i got the following
error: 

org.xbill.DNS.Tokenizer$TokenizerException: movie.db:3: invalid unknown RR
encoding

Have you any idea to solve this problem? 

Thank you for your helping, 

cheers, 

Jihane 



Brian Wellington wrote:
> 
> 
> On Mar 26, 2010, at 3:52 AM, Jihane wrote:
> 
>> I'm a student in an engineering school. As a part of my training, I need
>> to
>> develop a new resource record type in dnsjava. Firstly, i have created a
>> class which defines this new type. This latter is extends Record class.
>> After that, I have run the jnaned server on my computer and registered a
>> name and its corresponding information of this new type. Unfortunately,
>> when
>> I did a lookup for this type or the AAAA record type, I didn't get an
>> answer
>> even if these RRs exist in .db.
>> 
>> PS: I added in class Type this new type.
>> 
>> Have you an explanation for this problem?
>> 
>> If the server does not recorgnize this new type.In which class i need to
>> add
>> information about this new type (jnamed for instance)? because it doesn't
>> work also for AAAA Record type.
> 
> All that should be needed to add a new type is to implement a class which
> extends Record and add the appropriate initialization to the Type class. 
> If AAAA records also aren't working, then you're likely doing something
> else wrong, but I have no idea what it is.
> 
> Brian
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> dnsjava-users mailing list
> dns...@li...
> https://lists.sourceforge.net/lists/listinfo/dnsjava-users
> 
> 

-- 
View this message in context: http://old.nabble.com/Developping-a-new-record-type-in-dnsjava-tp28040758p28065987.html
Sent from the dnsjava-users mailing list archive at Nabble.com.

Re: Developping a new record type in dnsjava

[Brian W. <bwe...@xb...>]

On Mar 26, 2010, at 3:52 AM, Jihane wrote:

> I'm a student in an engineering school. As a part of my training, I need to
> develop a new resource record type in dnsjava. Firstly, i have created a
> class which defines this new type. This latter is extends Record class.
> After that, I have run the jnaned server on my computer and registered a
> name and its corresponding information of this new type. Unfortunately, when
> I did a lookup for this type or the AAAA record type, I didn't get an answer
> even if these RRs exist in .db.
> 
> PS: I added in class Type this new type.
> 
> Have you an explanation for this problem?
> 
> If the server does not recorgnize this new type.In which class i need to add
> information about this new type (jnamed for instance)? because it doesn't
> work also for AAAA Record type.

All that should be needed to add a new type is to implement a class which extends Record and add the appropriate initialization to the Type class.  If AAAA records also aren't working, then you're likely doing something else wrong, but I have no idea what it is.

Brian

Developping a new record type in dnsjava

[Jihane <jbe...@ho...>]

Hello Brian,

I'm a student in an engineering school. As a part of my training, I need to
develop a new resource record type in dnsjava. Firstly, i have created a
class which defines this new type. This latter is extends Record class.
After that, I have run the jnaned server on my computer and registered a
name and its corresponding information of this new type. Unfortunately, when
I did a lookup for this type or the AAAA record type, I didn't get an answer
even if these RRs exist in .db.


PS: I added in class Type this new type.

Have you an explanation for this problem?

If the server does not recorgnize this new type.In which class i need to add
information about this new type (jnamed for instance)? because it doesn't
work also for AAAA Record type.

Thank you in advance,

Cheers


Jihane
 

-- 
View this message in context: http://old.nabble.com/Developping-a-new-record-type-in-dnsjava-tp28040758p28040758.html
Sent from the dnsjava-users mailing list archive at Nabble.com.

Query

[Sandeep K. <tks...@gm...>]

Hi All,
          I am new to this group and Thanks for having this group.
 I just downloaded and was trying out few things got this doubt and thought
of getting help on the same
        My requirement is to
            1) specify the DNS server which it needs to query
            2) allow me to use options similar to dig
            3) collect the reponse given by the DNS server.

        Thanks in Advance
Sandy

Re: Re: Problem validating NSEC signatures

[<80...@gm...>]

On , Brian Wellington <bwe...@xb...> wrote:
--- snip ---

> I received a patch to fix this problem earlier today, and it's now  
> applied to the svn repository.


Thanks guys!

Regards,
Tom

Re: Problem validating NSEC signatures

[Brian W. <bwe...@xb...>]

On Nov 6, 2009, at 3:53 AM, Tom wrote:

> Hi,
>
> I've had a few issues trying to validate NSEC records. I am using
> dnsjava to load a zonefile and then validate signatures. All
> signatures so far have been fine with the exception of RRSIGs for NSEC
> records.
>
> When re-constructing the data for the signature the first record type
> of the NSEC record seems to be missed in the output, eg.
>
> www.tom.        3600    IN      NSEC    www2.tom. A RRSIG NSEC
> www.tom.        3600    IN      RRSIG   NSEC 5 2 3600 20091118194406
> 20091104163457 56229 tom. wVc0nokSM..... ;{id = 56229}
>
> With these records TypeBitmap.toWire() will only add the types RRSIGN
> and NSEC to the RRDATA - in other words, it misses out the A.
>
> If I initialize the mapbase variable in the code to 0 instead of -1 it
> works. I can't really follow the code so this could be a dumb
> suggestion ;-)

I received a patch to fix this problem earlier today, and it's now  
applied to the svn repository.

Thanks for the report!

Brian

Problem validating NSEC signatures

[Tom <80...@gm...>]

Hi,

I've had a few issues trying to validate NSEC records. I am using
dnsjava to load a zonefile and then validate signatures. All
signatures so far have been fine with the exception of RRSIGs for NSEC
records.

When re-constructing the data for the signature the first record type
of the NSEC record seems to be missed in the output, eg.

www.tom.        3600    IN      NSEC    www2.tom. A RRSIG NSEC
www.tom.        3600    IN      RRSIG   NSEC 5 2 3600 20091118194406
20091104163457 56229 tom. wVc0nokSM..... ;{id = 56229}

With these records TypeBitmap.toWire() will only add the types RRSIGN
and NSEC to the RRDATA - in other words, it misses out the A.

If I initialize the mapbase variable in the code to 0 instead of -1 it
works. I can't really follow the code so this could be a dumb
suggestion ;-)


Regards,
Tom

Re: Flags.AD missing on DNSSEC response header

[Brian W. <bwe...@xb...>]

On Sep 30, 2009, at 11:23 PM, Christian Möller wrote:

> Hi Brian
>
>>> I've testing dnsjava lib with DNSsec and it seems that the AD flag
>>> (authenticated data) within a response is not recognized correctly.
>>>
>>> Here is the header section of a response I've received asking
>>> nameserver a.ns.se for A records of google.se (for testing  
>>> purposes):
>>>
>>> ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 24516
>>> ;; flags: qr ; qd: 1 an: 0 au: 6 ad: 1
>>> ;; QUESTIONS:
>>> ;;    google.se., type = A, class = IN
>>> [...]
>>>
>>> But asking response's org.xbill.DNS.Message header for Flag.AD  
>>> fails:
>>>
>>> org.xbill.DNS.Message response = ...
>>> response.getHeader().getFlag(org.xbill.DNS.Flags.AD); // returns
>>> "false"
>>
>> Unless I'm missing something, this is because the AD bit isn't set.
>>
>>> Asking header for other flags (like Flags.QR in this example)
>>> succeeds, so what's going wrong here? Is it a bug or am I missing
>>> something?
>>
>> The QR flag is set.  There are no other flags set, so asking for any
>> other one will return false.
>>
>> Is it possible that you're misreading the dig ouptut?  The "ad" in
>> there refers to the count of records in the additional section, not a
>> flag.
>
> Indeed, you're right! Sorry for any inconvenience and thanks for  
> opening my eyes.
> But what the hell drives them to give different concepts the same  
> abbreviation? Maybe checking attentiveness of the reader? ;-)


No problem.

As to why the abbreviations are the same - that's a good question.   
The flags are usually printed in capital letters, which helps a bit,  
but dig (the original dig, from which dnsjava's dig header format was  
copied) doesn't do that.

Brian