RE: Comments on draft-nadas-vrrp-unified-spec-00.txt

Hi Steve,

With regard to number 4:
In section 6.4.3 when a master receives a higher priority advertisement then
the master accepts the new advertisement interval, needs to recompute the
Master_Down_Interval, and transition to backup.

Totally unrelated but far less important is the typo at 5.2.6, which should
be Rsvd.

Steve

-----Original Message-----
From: Stephen Nadas (RL/TNT) [mailto:[email protected]] 
Sent: Monday, August 20, 2007 5:58 AM
To: Steve Bates; [email protected]
Subject: RE: Comments on draft-nadas-vrrp-unified-spec-00.txt 

Hi Steve, 

Thank you for reading and sending comments.  As to the specifics, please see
inline.  

Regards,
Steve 

> -----Original Message-----
> From: Steve Bates [mailto:[email protected]]
> Sent: Friday, August 17, 2007 11:51 AM
> To: Stephen Nadas (RL/TNT); [email protected]
> Subject: Comments on draft-nadas-vrrp-unified-spec-00.txt
> 
> Steve,
> 
> Overall I like the new draft, but I would like to offer a few
> comments:
> 
> 1) In the abstract and introduction you mention VRRP groups.  
> I assume what you mean by this is that within a VRRP router the 
> virtual routers in each address family are a domain unto themselves 
> and do not overlap.  It might just be me, but the word group implies 
> some new organizational structure that we're not really adding.  
> Particularly in the second paragraph of section 3 where you mention a 
> group number.

I agree with your point and I'm not intending any "new organizational
sttructure"; I'll reword this to avoid the term. 

> 
> 2) In section 6.4.1 when we transition from the initialize state to 
> the backup state don't we want to set the Master_Down_Timer to the 
> Master_Down_Interval (not the Adver_Timer)?
> 

Yes. This appears to be a merge error on my part; I'll fix this. 

> 3) I think we should also mention in the two cases (both from the 
> backup and master state) when we set the Master_Adver_Interval to 
> Adver Interval contained in the ADVERTISEMENT that we need to 
> recompute the Master_Down_Interval.
> 

I see the text you mention in 6.4.2 (backup).  Unless there are objections,
I am okay adding a bullet here to "recompute the Master_Down_Interval."
However, I don't find this text in the master state - Did you mean backup
and initialize state (where the same text will appear once I fix previous
merge error?) 

> 4) In section 7.1 we now have a contradiction.  Since we are accepting 
> the Master's advertisement interval we can no longer reject a 
> mismatch.
> 

Right.  Need to fix.  (Also looks like a merge error from 3768.) 

> 5) In A.1 bullet item 2 you have a reference to VRRPv4.
> 

Sure, this is a typo from pasted email text. 

> Thanks,
> Steve
> 
> 


_______________________________________________
vrrp mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/vrrp