<div dir="ltr"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">My speculative guess would be that OOB access to a few outbound-facing<br>routers per DC does not help much if a configuration error withdraws the<br>infrastructure prefixes down to the rack level while dedicated OOB to<br>each RSW would be prohibitive.<br></blockquote><div><br></div><div>If your OOB has any dependence on the inband side, it's not OOB. </div><div><br></div><div>It's not complicated to have a completely independent OOB infra , even at scale. <br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Oct 5, 2021 at 8:40 AM Hauke Lampe <<a href="mailto:lampe@hauke-lampe.de">lampe@hauke-lampe.de</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On 05.10.21 07:22, Hank Nussbacher wrote:<br>
<br>
> Thanks for the posting.  How come they couldn't access their routers via<br>
> their OOB access?<br>
<br>
My speculative guess would be that OOB access to a few outbound-facing<br>
routers per DC does not help much if a configuration error withdraws the<br>
infrastructure prefixes down to the rack level while dedicated OOB to<br>
each RSW would be prohibitive.<br>
<br>
<a href="https://research.fb.com/wp-content/uploads/2021/03/Running-BGP-in-Data-Centers-at-Scale_final.pdf" rel="noreferrer" target="_blank">https://research.fb.com/wp-content/uploads/2021/03/Running-BGP-in-Data-Centers-at-Scale_final.pdf</a><br>
</blockquote></div>