Re: [pim] Publishing draft-ietf-pim-dr-improvement-08

[<[e-mail]>]

Hi Gyan,






As co-author, thank you very much for your review! :-)


The motivation of this draft is from the actual PIM deployment of our customers. And the solution works well.


From RFC7761, the default value of DR priority is 1. For the interface which never becomes DR you may set the DR priority to 0 by configuration.






Thanks,


Sandy











原始邮件



发件人：GyanMishra <[e-mail]>
收件人：Gyan Mishra <[e-mail]>;
抄送人：徐本崇10065053;[e-mail] <[e-mail]>;[e-mail] <[e-mail]>;[e-mail] <[e-mail]>;[e-mail] <[e-mail]>;
日 期 ：2019年10月12日 22:40
主 题 ：Re: [pim] Publishing draft-ietf-pim-dr-improvement-08




PIM WG

In reviewing the draft I agree this is a much needed for PIM stability to prevent preemption from occurring when a PIM neighbor comes up with a higher priority.


I have seen this scenario come up many times and had caused many multicast outages over the years with PIM being enabled accidentally on a terminal server or L2 switch for a vlan and had a higher IP address and became the DR or had a higher priority configured and became the DR.


Is there a way to have a DR priority 0 to make the device non DR eligible similar to ospf DR priority 0.


I had posted PIM BFD spec from the BFD WG which helps tune down PIM timers to sub millisecond convergence which helps to get hitless convergence for LAN multipoint scenario and for MVPN, however this draft is critical to provide LHR DR stability and prevent DR preemption from occurring.  Also adding the role priority DR/BDR state I think also adds to faster convergence when the DR goes down.


In environments where MLAG or like Cisco proprietary vPC is used forwarding happens to both DR and BDR active/active state which is very common have PIM MRIB MFIB state on both routers so is that addressed how that would work in the draft. 


Also was wondering in the case of active backup for faster convergence is it possible for both DR and BDR to have their MRIB MFIB state table built for ASM SPT tree and with PIM BFD tuned down timers so that convergence becomes hitless.


Thank you 


Gyan Mishra 
Verizon Communications 
Cell 301 502-1347


Sent from my iPhone


On Oct 11, 2019, at 8:58 PM, Gyan Mishra <[e-mail]> wrote:





I agree that PIM SM DR process and failover does need improvements of which I would like to bring to everyone’s attention the BFD WG draft for PIM BFD for fast hitless convergence.


https://tools.ietf.org/html/draft-mirsky-pim-bfd-p2mp-use-case-02


Gyan Mishra
Verizon Communications 
Cell 301 502-1347


Sent from my iPhone


On Oct 8, 2019, at 11:39 PM, <[e-mail]> <[e-mail]> wrote:









Hi, Stig


I don't know any other undisclosed IPR.


Thanks


Benchong Xu












_______________________________________________
pim mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/pim








发件人：StigVenaas <[e-mail]>
收件人：[e-mail] <[e-mail]>;[e-mail] <[e-mail]>;
日 期 ：2019年10月09日 06:06
主 题 ：Re: Publishing draft-ietf-pim-dr-improvement-08


Hi

Looks that there are no concerns moving ahead with publication. Can
the authors, and any others that might know
anything about IPR, let us know if they are aware of any other IPR
claims than what has been filed already?

Stig

On Tue, Oct 1, 2019 at 11:18 AM Stig Venaas <[e-mail]> wrote:
>
> Hi
>
> An earlier version of this draft passed WGLC many months ago. However,
> as discussed in the WG we agreed it should have some text pointing out
> how it relates to draft-mankamana-pim-bdr-02 that we are considering
> for adoption. This text has now been added, as well as some other
> minor changes.
>
> Please look at revision 08 and speak up if you have any concerns.
>
> Authors, and others, please let us know if you are aware of any other
> IPR claims than the one that already has been filed.
>
> I'll wait a week or so to give people a chance to respond. We need at
> least all the authors to respond regarding IPR before proceeding
> though.
>
> Thanks,
> Stig