[NANOG-announce] NANOG 72 Agenda is Published

[Ryan Woolley via NANOG-announce]

This message has been wrapped due to the DMARC policy setting to
prevent NANOG subscribers from being unsubscribed due to bounces.
--- Begin Message ---
NANOG Community,

The NANOG 72 Agenda  is published and
available as an iCal Feed.  The Program Committee has worked closely with
our speakers to develop a first-rate program and we encourage all attendees
to enjoy the whole conference.  As we review submitted content, minor
changes to the agenda are possible and will be updated in the online feed.
Thank you to the many members of our community who submitted interesting
content for NANOG 72!   Over 300 attendees have
registered
 and the NANOG family looks forward to
welcoming all of you to Atlanta.

Useful Information

   -

   NANOG 72 General Information and Registration
   
   -

   NANOG 72 Agenda 
   -

   Hotel Room  Block(s) Information
   -

   Reminder, registration is required to participate in the NANOG 72
   Hackathon 
   -

   Want to network with other attendees?  Use the meeting scheduler for
   NANOG 72
   

   -

   A welcome message that will be sent to registered NANOG 72 attendees
   shortly will provide more information on scheduled events and applications
   to help you to make the most of your time in Atlanta


Safe travels and see you in Atlanta.

Sincerely,

Ryan Woolley

NANOG PC
--- End Message ---
___
NANOG-announce mailing list
nanog-annou...@nanog.org
https://mailman.nanog.org/mailman/listinfo/nanog-announce

Re: Blockchain and Networking

[Saku Ytti]

On 11 January 2018 at 00:54, Filip Hruska  wrote:

> You can't just run normal software on ASICs. It's not a computer. They're
> literally hard-wired to do one thing - and do it well.
> Switch ASICs, for example, are good for switching network packets around.
> Though (I would assume) they
> can't do any kind of hashing, much less Bitcoin-specific stuff.

You're probably right that Switch ASICs would not be able to do any
BTC stuff. But most of them do use hashes, MACs are in hash tables
often, balancing requires hashes. There is also somewhat hazy
definition what is ASIC and what is NPU, NPU boxes are large
collection of same chips, and chips are executing same software to
completion, they can do anything, just matter of how long it takes and
how effective they are doing it, they certainly wouldn't be practical
BTC miners..


-- 
  ++ytti

Re: Cisco switch recommendations

[JASON BOTHE]

They work pretty well as do the new 9300s. They handle ldp nicely as well and 
are a tad bit shallower than the 3850 of that’s of benefit. 

Jason



> On Jan 10, 2018, at 17:06, Jorg Bielak  wrote:
> 
> I’m currently using Cisco WS-4948-10GE switches on some of my sites.
> Since I need more than two 10-Gig ports the Cisco 4948s have reached 
> end-of-life for me. 
> The current Cisco 4948 switches are doing my OSPF and BGP routing, as well as 
> some ACLs and actually work great for us….
> I’m looking to replace the current switches with Cisco 3850-12XS switches 
> (using the IP Services OS).
> Does anyone have experience with the Cisco 3850-12XS switches, in comparison 
> to the WS-4948-10GE switches? Or other recommendations?
> 
> Thanks
> JB

Cisco switch recommendations

[Jorg Bielak]

I’m currently using Cisco WS-4948-10GE switches on some of my sites.
Since I need more than two 10-Gig ports the Cisco 4948s have reached 
end-of-life for me. 
The current Cisco 4948 switches are doing my OSPF and BGP routing, as well as 
some ACLs and actually work great for us….
I’m looking to replace the current switches with Cisco 3850-12XS switches 
(using the IP Services OS).
Does anyone have experience with the Cisco 3850-12XS switches, in comparison to 
the WS-4948-10GE switches? Or other recommendations?

Thanks
JB

Re: Blockchain and Networking

[Filip Hruska]

Application Specific Integrated Circuit. It's even in the name!

You can't just run normal software on ASICs. It's not a computer. 
They're literally hard-wired to do one thing - and do it well.
Switch ASICs, for example, are good for switching network packets 
around. Though (I would assume) they

can't do any kind of hashing, much less Bitcoin-specific stuff.

Trying to mine Bitcoin on switch ASICs would be like trying to transfer
water through a 2.4GHz WiFi connection - both are absolutely 
preposterous ideas.



Regards

--
Filip Hruska
Linux System Administrator

Dne 1/9/18 v 17:02 Michael Crapse napsal(a):

The definition of an ASIC is that it has only one use. Just because half of
a 100gb switch is not in use doesn't mean that you can mine bitcoin, or run
a blockchain with the asics not in use..

On 9 January 2018 at 08:49, Jean | ddostest.me via NANOG 
wrote:


BTC miners use asics. Big switches/routers use 100Gb asics. Some
switches have multiple 100 Gb asics and sometimes only half is use or
even less.

I guess it could be nice for some smaller telcos to generate some profit
during off peak period. I don't know how feasible and I fully understand
that the vendor warranty should be instantly void.

Also, sometimes telcos have off the shelves spare that gather dust for
years... It could be interesting to also generate few coins.

Jean

On 18-01-09 10:31 AM, Naslund, Steve wrote:

Sure but there are lots of blockchains other than bitcoin.  A lot of

real smart people do not even suspect that bitcoin is a long term survivor
due to its long transaction times.  Which blockchains do you want to
support?  150GB may not seem like a lot (although a lot of my gear does not
have the memory to cache that) but 10 of those is beyond the memory on the
vast majority of network gear I am aware of.  That sure looks like a
slippery slope to me.   Now that a lot of network switching and routers can
support applications, you could just host all of your apps on them just
like you could do all of your routing in your servers.   The question for
you is what responsibilities do you want to take on.   That probably
depends on what business you are in.

There is absolutely no reason that the networking equipment itself

can't both operate the blockchain and keep a full copy.  It's a pretty good
bet that your own routers will probably be online;  if not, you have bigger
problems.

The storage requirements aren't particularly onerous.  The entire

Bitcoin blockchain is around 150GB, with several orders of magnitude more
transactions (read: config changes) than you're likely to see even on a
very large network.  SSDs are small >enough and reliable enough now that
the physical space requirements are quite small.

Steven Naslund
Chicago IL

GAO Report: FCC Should Improve Monitoring of Industry Efforts to Strengthen Wireless Network Resiliency

[Sean Donelan]

https://www.gao.gov/products/gao-18-198
FCC Should Improve Monitoring of Industry Efforts to Strengthen Wireless 
Network Resiliency


What GAO Found
The number of wireless outages attributed to a physical incident—a natural 
disaster, accident, or other manmade event, such as vandalism—increased 
from 2009 to 2016, as reported to the Federal Communications Commission 
(FCC). During this time, the number of outages substantially increased 
from 189 to 1,079 outages, with most of the increase occurring from 2009 
to 2011. FCC officials said this increase was due in part to growth in 
wireless customers and wireless infrastructure. Almost all outages 
attributed to a physical incident were due to an accident, such as damage 
to a cable due to a digging error (74 percent) or a natural disaster (25 
percent). However, outages due to a natural disaster had a longer median 
duration (ranging from 19 to 36 hours), which was more than twice as long 
as outages caused by an accident. Power failures and failures in other 
providers' networks also play a role in wireless outages attributed to 
physical incidents. For instance, carriers reported that 87 percent of 
wireless outages attributed to a physical incident were due to a failure 
in another provider's network on which they rely.

ICANN 61 San Juan - TechDay Call for Presentations

[Jacques Latour]

   Call for Presentations
  TechDay
at ICANN 61
  in San Juan, PR

The ICANN Tech Working Group is again planning a technical workshop at
the ICANN 61 meeting on Monday 2018-03-12 in San Juan, Puerto Rico.

The TechDay workshop has been a part of ICANN meetings for several
years and has provided a forum for both experienced and new people to
meet, present and discuss technical topics related to registry and DNS
work and security.

We are specially interested in:

1. Continuity of Operations

   Preparedness and Mitigation

2. In addition, we welcome suggestions for additional topics, such as;

* Registry security, services and systems
* DNS & anycast security, services and systems
* Big Data & associated data analysis
* IETF DNS protocol updates (Privacy, Encryption)
* (Recent) DDOS Attacks and Mitigation

If you are interested in presenting, please email ccnso-tech...@icann.org

We hope that you can join us and request that you disseminate this
Request to other lists where it might of interest.

Jacques

Regards,

On behalf of the TechDay Planning Committee