[ntp:questions] ntp broadcast not working with IPv6

2011-11-30 Thread rakesh v
Hi,
Iam facing issues with ntp broadcast  in case of IPv6.
Iam using the latest ntp 4.26p3 package.
Can I have a use case scenario for ntp broadcast in IPv6?
As I see in the documentation, we can have only multicast addresses in IPv6.

Here are my configurations:
Client side:

*# Created by IMI. /etc/ntp.conf*
*authenticate yes*
*
*
*# Broadcastclient mode. *
*disable auth *
*#broadcastclient*
*
*
*multicastclient ff02::101*
*
*
*# Drift file*
*driftfile /etc/ntp/drift*
*
*
*# Keys file. *
*keys /etc/ntp/keys*
*#trustedkey 100*


Server Side:

*# Created by IMI. /etc/ntp.conf*
*server 127.127.1.0  #Local clock*
*fudge 127.127.1.0 stratum 2 #Not disciplined*
*
*
*broadcast ff02::101 iburst*
*#authenticate yes*
*
*
*# Drift file*
*driftfile /etc/ntp/drift*
*
*
*# Keys file. *
*keys /etc/ntp/keys*
*#trustedkey 100*

Iam able to see the broadcast packets being received in the client but time
synchronization does not happen in the client.
I have even enabled the authentication on both the sides also. But no use.
Can somebody help me out from this?


-- 
Thanks & Regs,
Rakesh V
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions


Re: [ntp:questions] ntp broadcast not working with IPv6

2011-12-06 Thread rakesh v
Hi Danny,
Thanks for your valuable reply. Please see my comments/observations inlined


On Wed, Dec 7, 2011 at 8:54 AM, Danny Mayer  wrote:

> On 11/30/2011 3:50 AM, rakesh v wrote:
> > Hi,
> > Iam facing issues with ntp broadcast  in case of IPv6.
> > Iam using the latest ntp 4.26p3 package.
> > Can I have a use case scenario for ntp broadcast in IPv6?
> > As I see in the documentation, we can have only multicast addresses in
> IPv6.
>
> Where did you see that? It's not true, multicast also works with IPv4.
>

*RV: Iam not saying multicast does not work with IPv4. I was saying for
IPv6, we can give only multicast addresses as part of broadcast command. *

> >
> > Here are my configurations:
> > Client side:
> >
> > *# Created by IMI. /etc/ntp.conf*
> > *authenticate yes*
> > *
>
> What is this? I don't recall such an option.
>
*RV: This is to enable authentication.* *However in my case, I was trying
without authentication. *

>
> > *
> > *# Broadcastclient mode. *
> > *disable auth *
> > *#broadcastclient*
> > *
> > *
> > *multicastclient ff02::101*
> > *
> > *
> > *# Drift file*
> > *driftfile /etc/ntp/drift*
> > *
> > *
> > *# Keys file. *
> > *keys /etc/ntp/keys*
> > *#trustedkey 100*
> >
> >
> > Server Side:
> >
> > *# Created by IMI. /etc/ntp.conf*
> > *server 127.127.1.0  #Local clock*
> > *fudge 127.127.1.0 stratum 2 #Not disciplined*
> > *
> > *
> > *broadcast ff02::101 iburst*
>
> iburst is invalid.
>
*RV: Yes, It is invalid, because it doesn't send burst of packets in case
of broadcast. It was a misconfiguration.*

>
> > *#authenticate yes*
> > *
> > *
> > *# Drift file*
> > *driftfile /etc/ntp/drift*
> > *
> > *
> > *# Keys file. *
> > *keys /etc/ntp/keys*
> > *#trustedkey 100*
> >
> > Iam able to see the broadcast packets being received in the client but
> time
> > synchronization does not happen in the client.
> > I have even enabled the authentication on both the sides also. But no
> use.
> > Can somebody help me out from this?
>
> run ntpd -D2 and see what the output looks like.
>

*RV: I couldn't run the ntpd with -D2 option as it is not available.*
*These are the options available for ntpd (alphabetically)*

*-b no  bcastsync  Allow us to sync to broadcast servers*
*-c Str configfile configuration file name*
*-f Str driftfile  frequency drift file name*

*Are you mentioning about -b option here? How to give that option?*


> Danny
> ___
> questions mailing list
> questions@lists.ntp.org
> http://lists.ntp.org/listinfo/questions
>



-- 
Regs,
Rakesh Vanam
+91 9900 024 002
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions


Re: [ntp:questions] ntp broadcast not working with IPv6

2011-12-06 Thread rakesh v
Hi David,
Iam using the ntp package that is available for Debian 64.
http://packages.debian.org/stable/ntp
http://packages.debian.org/squeeze/amd64/ntp/download

Not sure whether this is Openntp?
One query though it is not intended to discuss here:
So, if the option(-D2) is not present can't we run ntp multicast client in
Openntp/SNTP? Is there any alternative?




On Wed, Dec 7, 2011 at 11:35 AM, David Woolley
 wrote:

> rakesh v wrote:
>
>  *RV: I couldn't run the ntpd with -D2 option as it is not available.*
>> *These are the options available for ntpd (alphabetically)*
>>
>> *-b no  bcastsync  Allow us to sync to broadcast servers*
>> *-c Str configfile configuration file name*
>> *-f Str driftfile  frequency drift file name*
>>
>
> From a fairly old man page:
>
> SYNOPSIS
>   ntpd [ -46aAbdDgLnNqx ] [ -c conffile ] [ -f driftfile ] [ -i
> jaildir ]
>   [ -k keyfile ] [ -l logfile ] [ -p pidfile ]  [  -P  priority  ]  [
>  -r
>   broadcastdelay  ]  [  -s statsdir ] [ -t key ] [ -u user[:group] ] [
> -v
>   variable ] [ -V variable ]
>
>
> Are you possibly using openntp?  There is very little knowledge of that
> here and, at least at one time, this was actually an SNTP implementation,
> not a full NTP one.
>
>
> __**_
> questions mailing list
> questions@lists.ntp.org
> http://lists.ntp.org/listinfo/**questions<http://lists.ntp.org/listinfo/questions>
>



-- 
Regs,
Rakesh
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions


Re: [ntp:questions] ntp broadcast not working with IPv6

2011-12-07 Thread rakesh v
Hi Danny,
Actually I ran the ntpd with Debug option before with a downloaded source
code (ntp-4.26p4).


These are the logs Iam seeing on the client side...

*6 Dec 06:11:52 ntpd[2884]: set_process_priority: Leave priority alone:
priority_done is <2>*
* 6 Dec 06:11:52 ntpd[2884]: proto: precision = 0.208 usec*
* 6 Dec 06:11:52 ntpd[2884]: ntp_io: estimated max descriptors: 1024,
initial socket boundary: 16*
* 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 0 v4wildcard 0.0.0.0 UDP 123
*
* 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 1 v6wildcard :: UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 2 lo 127.0.0.1 UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 3 eth1 10.16.34.6 UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 4 lo ::1 UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 5 eth2 2000::2 UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 6 eth1
fe80::a00:27ff:fe68:7b4c UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: Listen normally on 7 eth2
fe80::a00:27ff:fe38:605 UDP 123*
* 6 Dec 06:11:52 ntpd[2884]: peers refreshed*
* 6 Dec 06:11:52 ntpd[2884]: Listening on routing socket on fd #24 for
interface updates*
* 6 Dec 06:11:52 ntpd[2884]: bind(25) AF_INET6 ff02::101#123 (multicast)
flags 0x0 failed: Invalid argument*
* 6 Dec 06:11:52 ntpd[2884]: multicast address ff02::101 using wildcard
interface #1 v6wildcard*
* 6 Dec 06:11:52 ntpd[2884]: Joined :: socket to multicast group ff02::101*
*
*
Actually Iam able to see only one Broadcast packet being received on client
side in wireshark. Packets are not received.
Here Iam not sure why the socket bind is failing for ntpd? Any inputs on
this?


Regs,
RV.







On Wed, Dec 7, 2011 at 8:54 AM, Danny Mayer  wrote:

> On 11/30/2011 3:50 AM, rakesh v wrote:
> > Hi,
> > Iam facing issues with ntp broadcast  in case of IPv6.
> > Iam using the latest ntp 4.26p3 package.
> > Can I have a use case scenario for ntp broadcast in IPv6?
> > As I see in the documentation, we can have only multicast addresses in
> IPv6.
>
> Where did you see that? It's not true, multicast also works with IPv4.
> >
> > Here are my configurations:
> > Client side:
> >
> > *# Created by IMI. /etc/ntp.conf*
> > *authenticate yes*
> > *
>
> What is this? I don't recall such an option.
>
> > *
> > *# Broadcastclient mode. *
> > *disable auth *
> > *#broadcastclient*
> > *
> > *
> > *multicastclient ff02::101*
> > *
> > *
> > *# Drift file*
> > *driftfile /etc/ntp/drift*
> > *
> > *
> > *# Keys file. *
> > *keys /etc/ntp/keys*
> > *#trustedkey 100*
> >
> >
> > Server Side:
> >
> > *# Created by IMI. /etc/ntp.conf*
> > *server 127.127.1.0  #Local clock*
> > *fudge 127.127.1.0 stratum 2 #Not disciplined*
> > *
> > *
> > *broadcast ff02::101 iburst*
>
> iburst is invalid.
>
> > *#authenticate yes*
> > *
> > *
> > *# Drift file*
> > *driftfile /etc/ntp/drift*
> > *
> > *
> > *# Keys file. *
> > *keys /etc/ntp/keys*
> > *#trustedkey 100*
> >
> > Iam able to see the broadcast packets being received in the client but
> time
> > synchronization does not happen in the client.
> > I have even enabled the authentication on both the sides also. But no
> use.
> > Can somebody help me out from this?
>
> run ntpd -D2 and see what the output looks like.
>
> Danny
> ___
> questions mailing list
> questions@lists.ntp.org
> http://lists.ntp.org/listinfo/questions
>



--
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions


Re: [ntp:questions] ntp broadcast not working with IPv6

2011-12-07 Thread rakesh v
Thanks for your inputs Danny.

Hi Dave,
Can you give your inputs on this below issue?
Is the socket bind failing due to the reuse of the udp port 123. But I
assume for ntp multicast client also we use the same port for listening. Or
Is it due to some other issue?



On Wed, Dec 7, 2011 at 6:34 PM, Danny Mayer  wrote:

> On 12/7/2011 4:53 AM, rakesh v wrote:
> > Hi Danny,
> > Actually I ran the ntpd with Debug option before with a downloaded
> > source code (ntp-4.26p4).
> >
> >
> > These are the logs Iam seeing on the client side...
> >
> > /6 Dec 06:11:52 ntpd[2884]: set_process_priority: Leave priority alone:
> > priority_done is <2>/
> > / 6 Dec 06:11:52 ntpd[2884]: proto: precision = 0.208 usec/
> > / 6 Dec 06:11:52 ntpd[2884]: ntp_io: estimated max descriptors: 1024,
> > initial socket boundary: 16/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 0 v4wildcard 0.0.0.0 UDP
> > 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 1 v6wildcard :: UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 2 lo 127.0.0.1 UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 3 eth1 10.16.34.6 UDP
> 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 4 lo ::1 UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 5 eth2 2000::2 UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 6 eth1
> > fe80::a00:27ff:fe68:7b4c UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 7 eth2
> > fe80::a00:27ff:fe38:605 UDP 123/
> > / 6 Dec 06:11:52 ntpd[2884]: peers refreshed/
> > / 6 Dec 06:11:52 ntpd[2884]: Listening on routing socket on fd #24 for
> > interface updates/
> > / *6 Dec 06:11:52 ntpd[2884]: bind(25) AF_INET6 ff02::101#123
> > (multicast) flags 0x0 failed: Invalid argument*/
> > / 6 Dec 06:11:52 ntpd[2884]: multicast address ff02::101 using wildcard
> > interface #1 v6wildcard/
> > / 6 Dec 06:11:52 ntpd[2884]: Joined :: socket to multicast group
> ff02::101/
> > /
> > /
> > Actually Iam able to see only one Broadcast packet being received on
> > client side in wireshark. Packets are not received.
> > Here Iam not sure why the socket bind is failing for ntpd? Any inputs on
> > this?
>
> That sounds like a bug. I haven't looked at that area of the code
> recently but being unable to bind to the socket would prevent it
> receiving packets. Dave Hart has touched this area recently may will
> likely know more. I do remember some issues with multicast but I don't
> remember this one.
>
> Danny
>



-- 
Thanks & Regs,
Rakesh
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions


Re: [ntp:questions] ntp broadcast not working with IPv6

2011-12-08 Thread rakesh v
Hi Danny,
Have you got a chance to go through the code for this?
Here though the socket bind is failing the multicast registration is
happening successfully (which are shown by the debug traces here.)
Does that have anything to infer here?




>
> On Thu, Dec 8, 2011 at 9:58 AM, Danny Mayer  wrote:
>>
>> On 12/7/2011 11:15 PM, rakesh v wrote:
>> > Thanks for your inputs Danny.
>> >
>> > Hi Dave,
>> > Can you give your inputs on this below issue?
>> > Is the socket bind failing due to the reuse of the udp port 123. But I
>> > assume for ntp multicast client also we use the same port for listening.
>> > Or Is it due to some other issue?
>> >
>>
>> I think it is trying to bind to the incorrect address but I'd have to
>> look carefully at the code to figure out exactly what is going on.
>>
>> Danny
>>
>> >
>> >
>> > On Wed, Dec 7, 2011 at 6:34 PM, Danny Mayer > > <mailto:ma...@ntp.org>> wrote:
>> >
>> >     On 12/7/2011 4:53 AM, rakesh v wrote:
>> >     > Hi Danny,
>> >     > Actually I ran the ntpd with Debug option before with a downloaded
>> >     > source code (ntp-4.26p4).
>> >     >
>> >     >
>> >     > These are the logs Iam seeing on the client side...
>> >     >
>> >     > /6 Dec 06:11:52 ntpd[2884]: set_process_priority: Leave priority
>> >     alone:
>> >     > priority_done is <2>/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: proto: precision = 0.208 usec/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: ntp_io: estimated max descriptors:
>> > 1024,
>> >     > initial socket boundary: 16/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 0 v4wildcard
>> >     0.0.0.0 UDP
>> >     > 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen and drop on 1 v6wildcard ::
>> >     UDP 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 2 lo 127.0.0.1 UDP
>> >     123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 3 eth1 10.16.34.6
>> >     UDP 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 4 lo ::1 UDP 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 5 eth2 2000::2 UDP
>> >     123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 6 eth1
>> >     > fe80::a00:27ff:fe68:7b4c UDP 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listen normally on 7 eth2
>> >     > fe80::a00:27ff:fe38:605 UDP 123/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: peers refreshed/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Listening on routing socket on fd #24
>> > for
>> >     > interface updates/
>> >     > / *6 Dec 06:11:52 ntpd[2884]: bind(25) AF_INET6 ff02::101#123
>> >     > (multicast) flags 0x0 failed: Invalid argument*/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: multicast address ff02::101 using
>> >     wildcard
>> >     > interface #1 v6wildcard/
>> >     > / 6 Dec 06:11:52 ntpd[2884]: Joined :: socket to multicast group
>> >     ff02::101/
>> >     > /
>> >     > /
>> >     > Actually Iam able to see only one Broadcast packet being received
>> > on
>> >     > client side in wireshark. Packets are not received.
>> >     > Here Iam not sure why the socket bind is failing for ntpd? Any
>> >     inputs on
>> >     > this?
>> >
>> >     That sounds like a bug. I haven't looked at that area of the code
>> >     recently but being unable to bind to the socket would prevent it
>> >     receiving packets. Dave Hart has touched this area recently may will
>> >     likely know more. I do remember some issues with multicast but I
>> > don't
>> >     remember this one.
>> >
>> >     Danny
>> >
>> >
>> >
>> >
>> > --
>> > Thanks & Regs,
>> > Rakesh
>> >
>> >
>>
>
>
>
> --
>
>
>
>
> --
> Regs,
> Rakesh Vanam
> +91 9900 024 002
>
>
>
>
> --
>



-- 
Regs,
Rakesh V
___
questions mailing list
questions@lists.ntp.org
http://lists.ntp.org/listinfo/questions