Re: [xcat-user] External DNS/makedns/update server

2020-01-06 Thread Thomas HUMMEL

On 12/21/19 11:25 PM, Vinícius Ferrão via xCAT-user wrote:

Hello, Thomas. I’m not from the xCAT development team but I’m curious with your 
question.


Hello Vinícius [I was on vacation]



In your example you mean that the name server info on the SOA field is not 
available on the NS fields of a given zone, right?


Yep.


If this is correct I think xCAT is doing right to notify only the name servers 
on the NS fields, because the name server on the SOA record may be unreachable 
from the client sending the updates, depending on the architecture of the 
domain. A truly hidden master is only reachable from its slaves.


That's what I reckoned also.

In my setup, the master is half hidden indeed as it is listed in the SOA 
record as primary master. I think I did that because of nsupdate which 
man states :


"
Requests are sent to the zone's
master server. This is identified by the MNAME field of the zone's SOA
record.
"

(although I specify the server in the nsupdate input file)

So the 2 software I'm using to feed in my DNS (nsupdate and xCAT) does 
not have the same behavior (but I guess you're right with your explanation).



--
TH




___
xCAT-user mailing list
xCAT-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xcat-user


Re: [xcat-user] External DNS/makedns/update server

2019-12-21 Thread Vinícius Ferrão via xCAT-user
Hello, Thomas. I’m not from the xCAT development team but I’m curious with your 
question.

In your example you mean that the name server info on the SOA field is not 
available on the NS fields of a given zone, right?

If this is correct I think xCAT is doing right to notify only the name servers 
on the NS fields, because the name server on the SOA record may be unreachable 
from the client sending the updates, depending on the architecture of the 
domain. A truly hidden master is only reachable from its slaves.

Or I’ve lost something that I don’t understand?

Sent from my iPhone

> On 25 Nov 2019, at 11:15, Thomas HUMMEL  wrote:
> 
> Hello,
> 
> From what I've seen, when using an external DNS server, xCAT's 'makedns' 
> command looks for the NS records for the zone and sends them the update.
> 
> What's the reasoning behind this instead of, let's say, sending the update to 
> the MNAME field (primary master) of the SOA record, as for instance the 
> 'nsupdate' command does when no 'server' directive is used ?
> 
> As a matter of fact, in a setup consisting of (with bind/named) :
> 
> - a (hidden) master
> - 2 slaves (only them have the NS records for the zone)
> 
> it leads to unneccessary transfers :
> 
> - xCAT sends update to slave(s)
> - slaves forward the update to the master
> - master notify the slaves
> - records get transfered from master to slaves
> 
> Thanks
> 
> --
> TH
> 
> 
> 
> ___
> xCAT-user mailing list
> xCAT-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/xcat-user

___
xCAT-user mailing list
xCAT-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xcat-user


[xcat-user] External DNS/makedns/update server

2019-11-25 Thread Thomas HUMMEL

Hello,

From what I've seen, when using an external DNS server, xCAT's 
'makedns' command looks for the NS records for the zone and sends them 
the update.


What's the reasoning behind this instead of, let's say, sending the 
update to the MNAME field (primary master) of the SOA record, as for 
instance the 'nsupdate' command does when no 'server' directive is used ?


As a matter of fact, in a setup consisting of (with bind/named) :

- a (hidden) master
- 2 slaves (only them have the NS records for the zone)

it leads to unneccessary transfers :

- xCAT sends update to slave(s)
- slaves forward the update to the master
- master notify the slaves
- records get transfered from master to slaves

Thanks

--
TH



___
xCAT-user mailing list
xCAT-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xcat-user