M3UA Loadsharing Case, Withdrawal of ASP

[Ambika P Tripathy]

Hi Folks,

 

I have a doubt in section 5.2.3 of rfc 4666.

5.2.3.  n+k Sparing, Loadsharing Case, Withdrawal of ASP

   Following from the example in Section 5.1.4, ASP1 withdraws from
   service:

        SGP                 ASP1                ASP2                ASP3
          |                   |                   |                   |
          |<----ASP Inact.----|                   |                   |
          |---ASP Inact Ack-->|                   |                   |
          |                   |                   |                   |
          |--NTFY(Ins. ASPs)->|                   |                   |
          |---------------------------------------NOTIFY(Ins. ASPs)-->|
          |                   |                   |                   |
          |                   |                   |                   |
          |<----------------------------------------ASP Act (Ldshr)---|
          |------------------------------------------ASP Act (Ack)--->|
          |                   |                   |                   |
          |-NTFY(AS-ACTIVE)-->|                   |                   |
          |-------------------NOTIFY(AS-ACTIVE)-->|                   |
          |---------------------------------------NOTIFY(AS-ACTIVE)-->|
          |                   |                   |                   |
          |                   |                   |                   |

   For the Notify message to be sent, the SG maintains knowledge of the
   minimum ASP resources required (e.g., if the SG knows that "n+k" =
   "2+1" for a Loadshare AS and "n" currently equals "1").

   Note: If the SGP detects loss of the ASP1 M3UA peer (e.g., M3UA
   heartbeat loss or detection of SCTP failure), the initial ASPs
   Inactive message exchange (i.e., SGP-ASP1) would not occur.

 

Question>>

1. When ASP1 withdraw request from AS and ASP3 becomes active why SG will notify ASP2 with AS-ACTIVE??? As per my understanding the state of ASP2 is not changed.

--
Br,
Ambika Prasad Tripathy
Nethawk Networks
Cell: [phone]

_______________________________________________
Sigtran mailing list
Sigtran@...
https://www.ietf.org/mailman/listinfo/sigtran

[Brian F. G. Bidulock]

Ambika,

The diagram in section 5.2.3 of RFC 3332 is correct.
RFC 4666 is in error here.

--brian

Ambika Tripathy wrote:                                                                                                                                            (Thu, 10 Jul 2008 19:36:19)
>
>    Hi Folks,
>
>    I have a doubt in section 5.2.3 of rfc 4666.
>
<snip>---
>
>    Question>>
>
>    1. When ASP1 withdraw request from AS and ASP3 becomes active why SG
>    will notify ASP2 with AS-ACTIVE??? As per my understanding the state
>    of ASP2 is not changed.
>
<snip>---

--
Brian F. G. Bidulock
bidulock@...
http://www.openss7.org/
_______________________________________________
Sigtran mailing list
Sigtran@...
https://www.ietf.org/mailman/listinfo/sigtran

[David Laight-3]

 
My interpretation is that the NTFY(AS-ACTIVE) reports the state
at the SGP, not at each ASP.
See 5.1.2 where both ASP1 and ASP2 are sent the AS_ACIVE when
ASP1 sends and activate.

In 5.2.3 the NOTIFY(AS-ACTIVE) is sent to all ASPs in order
to reverse the NOTIFY(Ins. ASPs) sent earlier.

        David

> -----Original Message-----
> From: sigtran-bounces@...
> [mailto:sigtran-bounces@...] On Behalf Of Brian F. G. Bidulock
> Sent: 10 July 2008 15:32
> To: Ambika Tripathy
> Cc: sigtran; ambika.tripathy@...
> Subject: Re: [Sigtran] M3UA Loadsharing Case, Withdrawal of ASP
>
> Ambika,
>
> The diagram in section 5.2.3 of RFC 3332 is correct.
> RFC 4666 is in error here.
>
> --brian
>
> Ambika Tripathy wrote:                                        
>                                                              
>                                       (Thu, 10 Jul 2008 19:36:19)
> >
> >    Hi Folks,
> >
> >    I have a doubt in section 5.2.3 of rfc 4666.
> >
> <snip>---
> >
> >    Question>>
> >
> >    1. When ASP1 withdraw request from AS and ASP3 becomes
> active why SG
> >    will notify ASP2 with AS-ACTIVE??? As per my
> understanding the state
> >    of ASP2 is not changed.
> >
> <snip>---
>
> --
> Brian F. G. Bidulock
> bidulock@...
> http://www.openss7.org/
> _______________________________________________
> Sigtran mailing list
> Sigtran@...
> https://www.ietf.org/mailman/listinfo/sigtran
>

Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)  
P Please consider the environment and don't print this e-mail unless you really need to
_______________________________________________
Sigtran mailing list
Sigtran@...
https://www.ietf.org/mailman/listinfo/sigtran

[Brian F. G. Bidulock]

David,

Hmm.  Where did you see that stated in RFC 4666?

I can't really think that AS-ACTIVE reverses Ins. ASPs as
an AS can be AS-ACTIVE and have insufficient ASPs at the
same time.

--brian

David Laight wrote:                           (Thu, 10 Jul 2008 17:19:55)

>  
> My interpretation is that the NTFY(AS-ACTIVE) reports the state
> at the SGP, not at each ASP.
> See 5.1.2 where both ASP1 and ASP2 are sent the AS_ACIVE when
> ASP1 sends and activate.
>
> In 5.2.3 the NOTIFY(AS-ACTIVE) is sent to all ASPs in order
> to reverse the NOTIFY(Ins. ASPs) sent earlier.
>
> David
>

--
Brian F. G. Bidulock
bidulock@...
http://www.openss7.org/
_______________________________________________
Sigtran mailing list
Sigtran@...
https://www.ietf.org/mailman/listinfo/sigtran